Date 29.3.2024
Location: AIXC, VUB, Brussels
Time: 09:00 – 17:00
Organised by The Brussels Institute for Advanced Studies and FARI
General Chair : Prof. Bram Vanderborght (VUB)
Program Chair: Dr. Michel Joop van der Schoor (FARI/VUB), Dr. Fatma Demir (VUB)
Location: VUB AI Experience Centre – Pleinlaan 9, 1050 Ixelles
Description
Robots are becoming an integral part of our economy, society and planet, assisting us in various tasks. However, the rapid growth in robotics usage has also given rise to relevant concerns regarding sustainability. At present, urgent issues such as rare-earth material usage, e-waste, energy efficiency and ethical concerns hinders the sustainable development of robotics. As we increasingly rely on robotics, it becomes increasingly critical and urgent to ensure that these technological advancements are sustainable and eco-friendly. Since robotics is the integration of many technologies, we will discuss in this Forum sustainable robotics technologies ranging from material science, material processing and metal, electronics with sensors and processing, actuators, energy, power and batteries and control and artificial intelligence
08:30 -09:00 Registration and Coffee
09:00 – 09:30 Opening speech
BrIAS directors and Programme Directors
9:30 – 10:45 Session 1: Responsible Innovation and Artificial Intelligence (Development)
Michel Joop van der Schoor (Vrije Universiteit Brussel, Belgium)
Ward van der Tempel (Voxelsensors, Belgium)
Sophia Falk (Universität Bonn, Germany)
10:45 – 11:15 Coffee Break
11:15 – 12:45 Session 2: Energy Efficiency and Applications (Use Phase)
Goldar Davila Alejandro (Université libre de Bruxelles, Belgium)
Abhishek Gupta (Technische Universität Berlin, Germany)
Kris Verdonck (A Two Dogs Company, Belgium)
12:45 – 14:00 Lunch
14:00 – 15:15 Session 3: Sustainable Materials and Applications (Production)
Alix Partridge (University of Bristol)
Joost Brancart (Vrije Universiteit Brussel, Belgium)
Seppe Terryn (Vrije Universiteit Brussel, Belgium)
15:15 – 15:45 Coffee Break
15:45 – 16:45 Session 4: Circular Economy and Applications (End of life)
Jelle Saldien (Universiteit Gent, Belgium)
Ales Ude (Jožef Stefan Institute, Slovenia)
E e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e
E e e e e e e e e e e e e e e e e e e e e one okay good morning welcome on this forum on sustainable robotic Technologies I’m BR VOR Professor robotics at the V and affiliated with imic and I’m one of the directors of
This uh events together with Professor gon and Carl Mo and an yobs on sustainable robots and this uh Forum will be on sustainable robotic Technologies but maybe I give the floor to Professor J to introduce a bit Bri thank you uh Braham uh dear dear colleagues on uh
Behalf of uh Bri it’s a great pleasure to welcome all of you today to this Symposium the Brussels Studies Institute for advanced studies is a wonderful collaboration between the two Brussels universities University lib Brussel and V University Brussel as you might know Bri was the institute for Advan studies
Was created in Princeton in 1930 and since then several Institute for Advanced Studies have been established worldwide today there are about 30 Institute for advanced studies in Europe alone hosting a total of more than 500 high ranking researchers The Experience shows that those Institute uh are one of the most dynamic
Internationalization actors in the field and are a very effective instrument for the appeal and reputation of the regions and universities which host them based on submitted project Bri offer fellowships and housing to a number of high level researchers the Bri fellows for periods up to 10 months it is located in US
Square the prestigious renovated listed uh listed military barracks in Excel situated within walking distance of campuses of both the ulb and V Bri is not simply be just another Institute for advanced studies the traditions of ulb and V has funding universities highlight some unique niches such as subjects with important potential soci societal
Impacts residency for researchers who have had to flee their countries or collaboration ations with countries where Society faces critical challenges with respect to sustainability Bas will stimulate interaction between science and other fields of creativity which is essential for an harmonious and sustainable future and which will contribute to The Prestige of the
Brussels universities Bri selected the topic of sustainable robotics for its 2023 24 editions has robotic and automated devices become an integral part of our Lives it has become critical to understand if they can be truly durable and how they could optimally contribute to sustainability as a world
The Bri project allow our experts in robotics control automation to go further and achieve a dream bring new Partnerships will emerge from the conference I wish everyone a productive and inspiring conference experience thank you very much so thank you uh so indeed as mentioned Robotics and Technologies in
General has the power to completely transform societies and also Robotics and automations has such an enormous impact but of course we great power comes great responsibility so we need of course the digital transformation because we need to reduce costs and also improve productivity but we also need a
Sustainability uh transition and we need to Target that in the three dimensions economy Society uh and planets and at robotics we aim to bring the two together in a robotics uh context so we in the AI experience uh Center and so uh a robot is an integration of the best
Technology say you need new batteries new sensors Advanced processing AI algorithms and so on but of course you need to collaborate with humans so also the Human Social and medical Sciences are super uh important and Robotics encap plates uh these different disciplines uh together and this is also
Where you see in this room different applications and of course to create impact we collaborate a lot with companies but also um we are founding a lot of uh spin-off compan so one is the Brussels company axilus bionics which make bionic feed to to help amperes to
Regain a Mobility we’re also in the process to develop a company to make high efficiency uh drives gears uh because Motors consume a lot of energy and that creates because the low energy efficiency of especially the gears makes that they very heavy need a lot of
Batteries and that has a impact on you need bigger batteries and so one to carry it so that’s Innovative spin-off company we want to found we’re also working on a company for Sports Injury detection so with sensorized garments and also an exoskeleton uh spin-off company to help people to work uh with
Their hands for example above the head and to lift heavy loads and the last one but this is also supported this forum by the European projects uh Shiro and uh sorry Shinto and smart but sea and G will talk about more about our selfhealing and sustainable robots that
Can uh repair itself after damage but of course it’s also important to have the social impact and I think five six years ago we started with homo roboticus and the question was how we can preserve the human values on equality freedom and solidarity how they can stand ground in
A robotized and AI uh world and 56 academics from the VB from very different disciplines collaborated and we wrote uh this book but we also had a series of events when was in the uh the and we have a testimony of victaria Modesta which she is the Bic pop artist
And we will listen to her testimony it’s not like a pet it’s not like a human but it’s a robot [Applause] I’m really excited that an event like this is happening it’s great that the awareness of the kind of the responsibility about Innovation and future is kind of becoming much more at the Forefront and I feel like my job is to like Inspire the mind and the [Applause]
Imagination one of the saddest things I find is when people think that thinking about the future isn’t for them that they are not good enough or smart enough to think about the future I think that everybody should be engaged with how the future of their life and the environment
Is going to be thank you very much so uh out of this event and now in collaboration with theb fari uh started which is the AI Institute for the common good and which is heavily supported by the Brussels government and with a number of experts including Carol we
Wrote an uh Journal paper on robot otics enabling or innovator role of them in sustainable development goals and we saw that the results indicated that robotics has the potential to enable 46 of the targets particularly for the industry and environmental related SGS while on the other hand also 90% of the SG Target
Could in be inhibited mainly through uh exertion of inequalities and tensions in the sdgs and of course there is a whole series of uh Robots coming towards us this is uh Andre K is the famous Silicon Valley robotics uh director in startups and she distributed you have the famous
Hype guard hype cycle of Gardner where you have the plateau of productivity where currently already a lot of robots are present in industry for milking cows in surgical uh hospitals and so on then you have the the slope of Enlightenment where robots are coming which go bit more to unstructured environments you
Have also the thir of disillusion where you have for example autonomous cars which are not yet delivering the promises they made but also exoskeletons and cobots and quadrupeds are technologies that are not yet there and of course 2023 was really the year that a lot of humanoid companies were uh
Presented and so this is really the peak of inflated uh expectations and so during these events on sustainable robots we talked a lot on sustainable use of robots but the robot itself has also a lot of challenges because a human myself and you are mostly built by water
Uh carbon and so on but robots are using very rare a lot of special materials rare earth materials like gold uh the batteries use lithium Eon the uh the motors use neodium a lot of materials for which also ORS are being fought moreover the robots are very complex uh
So that means when they’re broken uh you have or costly repairs or they end up on the growing pile of E-Waste and that’s already a huge environmental challenge moreover robots have a lot of Motors so they consume energy moreover they often start more and more to use also AI
Algorithms which need to be trained which also consumes a lot of energy and so that’s why we have the this forum also to investigate how we can make the robot itself the body how it’s being processing and so on how we can there do more sustainability and this is because
I we think that it’s the right thing to do but on the other hand we also need to see it as a new business advantage and I think the GTP was very famous everyone complains ah all these buttons you need to say yes I accept or reject the
Cookies and it annoyed us but on the other hand several companies made it a business advantage and for example an apple which is a very famous MB company had the the idea is turning privacy into a business Advantage not just the marketing uh slogan so they gained a lot
Of money by implementing uh privacy in their products and now is the same happening with the European green deal which is a legislation towards a Greener future and so instead of objecting uh this trend also Apple for example in the beginning uh did did everything so you do couldn’t repair your smartphones but
Now they made a huge U-turn and now they sell self-repair uh kid so you can prolong the lifespan of your uh phone so I think it’s not only a matter of doing the writing but also thinking how we can make that a European business advantage
And so uh also Fatima she she wrote a new European project that is being submitted on sustainable robust am Mar Innovative training Network doal Network that we hope we will win we will hopefully soon have the answer where it starts from the core sustainable values on increased Lipan circularity resource
Conservation and so on and that as values and how we can develop Technologies which make up a robot on Lifetime extension material processing sensing processing activation and so on and so this was also the idea of this uh Forum to start discussing those different asp aspects so that means that
We build up the schedule like this so we have four different blocks so the first session is on responsible design and artificial intelligence ched by car Mo as s we have for example Michelle who is also now postto in the lab on life cycle engineering W it’s a very Innovative uh
Brussel startup company that also want to make more energy efficient and also low latency sensors very cool uh technology Sophia also uh wants to talk on digitalization and artificial intelligence how you can include sustainability and the boundaries there then we have a coffee break then we go to Energy Efficiency and applications as
So with gold there on fast charging and how you use control algorithms in it had to improve the charging and the batteries uh there uh then ABD Gupta on designing service robots and infrastructure systems for responsible operation uh chrisopher donk uh on robots without batteries very impressive
Artwork also and then after every uh session we also have a small panel uh discussion then we will have lunch then about sustainable materials and applications uh so Alex patri will talk about growing your own robot and then y will talk on our sustainable polymers for soft robot Technologies and then the
Final session on circular economy and applications with uh but y saline couldn’t come today because it was also interesting because he wants to include how we can make in the design robots that they can for example also be easier disassembled but the last minute couldn’t come anymore so it’s a bit pity
But Alish UD will talk about that you can also use robots for recycling of electronic devices so a very uh hopefully interesting uh day today in order to be interactive uh there is slid so you have the QR code uh there so where you can type your question and
Then they will be raised but now I uh ask Carol uh to introduce uh Stephanie hello everybody I just wanted to start by saying all the uh thanks to all the people involved in making this day possible from Michelle Anda Andres the directors uh also of the Bri program
And all the fellows uh present to today so thank you very much I also now need to and want to uh say thank you to uh one person Stephanie lipinsky who has been very instrumental in the development of fari one of the institutes behind uh the this year
Thematic uh and it’s Stephanie lepinski she is a policy advisor uh for scientific research European Affairs uh with the state Secretary barbarat trash here uh in the Brussels Capital Region uh they’re here also really following up closely on all the scientific Affairs uh with inovis if you don’t know it in
Brussels uh it’s the agency in charge of funding research and development and Innovation and there is a very very strong emphasis on sustainability and so Stephanie I wanted to say thank you and the floor is yours thank you thank you Carl um hi everyone and welcome to Brussels uh it’s great that
Um Bri yes and F can actually gather all all you um uh here it’s it’s really a pleasure and I hope you’ll have a fruitful day today um yeah sorry I also wanted to say um I’m really delighted to be here I’m delighted with the topic uh of today uh
Because um I must confess that I kind of have an obsession with the environmental Footprints of digital so um and and Carl would be would be here to testify that the first time they came came to the cabinet to pitch forari I was already raising that question Barbara attracted
My Minister she was as well so it’s really great to see today that you are spending all your um energy and brain and expertise to uh think about that uh that topic um I also wanted to um stress that this is even maybe uh even more important today because all of this
Happened like four something uh years ago but uh there was no generative AI at that moment uh for example at least not developed uh as it is today so this impact is increasing uh by the minute actually so it’s it’s really important to to think of how we can have an impact
On how to decrease it and I would like to stress today two lever I think uh we have first um and this is very personal at the individual level um just wanted to illustrate with an experience an illustration with what happened to me yesterday um I brought my uh
8-year-old 8-year-old son uh to the dentist and I was seeking advice on how he could brush his teeth in a more gentle way because otherwise I have to buy him a tou breath every two weeks which is not really sustainable and it’s no good there’s a sign that there’s
Something not going wrong going wrong and um believe it or not but the first advice that the dentist gave me was to actually for him to use nothing less than a little robot um she told me he should use an electric toothbrush which would actually make a noise when he when
Your son um pushes too much on his teeth and it was actually um disappointment for me because I thought are we now to the point where we actually have to rely on the machine to feel something inside of our body is it really a need that we have
Now I don’t think so I really want my son to be able to feel when he he brushes his teeth in a nice way which is not hurtful for his body uh the same uh once I was um on a on a trip with the state Secretary Barbara tra and we’ve
Been presented a connected uh water bottle and I was like what’s the point Sorry I want to be self-conscious of when I need to drink water I don’t want a machine to tell me that so I think the first lever to really reduce that environmental footprint is really for us
To reconnect with ourself and to think to question the need which is behind the robot that is in front of us do we really really need that or is it something that us as human we can manage second level uh I want to put uh
In in front of us and it’s because I don’t think this is only a question of the the individ individual uh themselves this is also bigger Collective question so that’s why at the governmental level in the Brussels Capital Region we act and actually Barbara TR she is not here
Today because she has to be at the the Brussels Parliament to make sure that a big uh change in the law that we have uh proposed to the parliament is going to be accepted today um we propos to change the law granting I mean giving the rules for granting uh subsidies to research
And Innovation projects um within the Brussels Capital Region what is the main change that we propose in the law this is to actually include a criteria about having a positive environmental and social impact in the project that we will finance as a Public Authority this is completely this is
Really a a big uh paradigm shift I think because so far we were just saying I mean Everyone likes Innovation you know especially in the political spheres uh and so no one really questions how we uh how we subsidy research and Innovation with this change in the law we really
Give a direction and a vision for which projects we want to uh we think are good for society and uh which we really be helpful uh to change big societal um um problems and when I look at the artist that you’ve just shown I think think
This is one of the collective issues we have you know how to make people more uh how to live in a society which is more inclusive for example this is really an an example where technology has something to uh to add so with this um I
Just uh wish you a very uh good day very productive and I thank you very much for coming to Brussels [Applause] thanks thank you very much for these beautiful words and also for setting the scene right for the for the hall uh you want to switch yes okay please I will so uh
For setting the scene right and I think it’s indeed the main question that has been driving us uh to create this year’s thematic but also to do a lot of activities with Emanuel Bram uh anob who is not here today uh and the goal is really to question uh the purpose of
What we do with Technologies uh what is the context behind it and also we hear a lot that the technologies that we work on every day can contribute to sustainability but we also wanted to have a strong critical look at uh is it really benefiting uh Society is it
Innovation at uh the expans of maybe more sustainable uh sustainability Dimensions so to set the scene right uh I want to first introduce uh for this sh this session on responsib Innovation and artificial intelligence the brain behind the operation of today and the program uh alongside brah Michel y bonders he is
Uh a postdoctoral researcher here at biotics and at fari uh he will be talking about the life cycle engineering for sustainability responsible design choices so Michelle it’s my pleasure to say that the floor is yours and please don’t forget that you have the slido where you can ask uh the questions and I
Will try to relay them uh after and I will keep time as [Applause] well oops Yeah just have to set up everything again okay there we go mik’s good perfect okay yeah um thanks theal for the introduction I will as you said or pointed out talk about the life cycle engineering for
Sustainability uh in terms of responsible design choices and I want to start with a little uh quote that was really motivated me or describes really um how I came to my work or how I changed my point of view in the technical area um so it goes if a discourse on technological developments
Were to remain solely at the application Level EG the question of technical feasibility it would lose sight of the profound influence the technological developments have on the constitution of a society and how they also change the Human Condition and the entire social arrangement of people um I think that
Puts it very well um that we need to look into technology not only from the application Level but also from all the other dimensions of sustainability and Bram already introduced us a bit to what those impacts can be I have a few pictures trying to depict really what is
Happening in the um extraction oh I think okay it’s hard to see that one there I will do it like this um the extraction of raw materials the energy consumption and of course end of life which will impact um not only environment but also as you can see the social life of
Humans um also a very big impact will come from robots going also from industry into the public so um those are some future Visions but uh like this one is a Future Vision but those are already existing so we have robots intruding the public space more and more so the
Impacts that they will have on um Society will grow and there is um very nice uh picture of the SGS in uh in order for the three different dimensions and um robots will have as Bram also pointed out in their paper have impacts on all of those
Stgs and the question is how can we estimate those impacts and one answer or one solution can be uh especially for developers uh to look at the life cycle so the life cycle if you haven’t heard of it is uh what you have for a product from the extraction of raw material
The production of intermediates production and then the use phase and finally the end of life for the product but then it goes on to recycling recovery but also to disposal those are all different life cycles that you need to consider in terms of impacts for um a
Product or the impact that a product has and a product like a robot um this will influence I just put a few of the SGS here um but not all of them but as you know it will have impacts on on all of them so for instance um production can
Influence the work um can influence the climate action the production and also the use phase can uh into impact the energy consumption and of course you have um life on land life on water for raw materials uh extraction in disposal in terms of um water um acidification or
Toxification um and of course there will be during the use phase concern about equality and consumption so um the question now is what can we do as designers or developers and um I think we can do a lot uh it especially in the beginning of a product in the um Inception of a
Product so there was a study that showed um that about 60 to 80% of the ecological impacts are determined in the early phase of product development um but the problem there is we have a dilemma of product design which is as follows we have this little graph here
And we can see this is the product life cycle I introduced the the phases already and then we have here um knowledge information and cost in terms of money and time and influence and the influence that we have as developers will go down over the time but the
Knowledge information will go up and also the cost of change and what that means is in the planning and development we have a lot of influence on the product how it will shape out uh how the impacts will be in the end during the use phase and the production but we also
Have not a lot of knowledge so it’s really hard to change we don’t have a lot of information and data and as soon as we get to a stage where the product is more um finished I would say of course the cost of change will increase
A lot so it’s harder to change or people are not that much likely to change the product in the end um so this is the phase where we as developers or also researchers I will um talk about that a bit um can do something and have influence so this
Product development phase um has four different phases and the more or the most interesting ones are the first two so in this conceptual design you have to determine functions search for Solutions evaluate them and then structure them into models and also the clarification of the problem or task is a really
Important thing to actually investigate more in the problem than the solution in the beginning and what you can do is you can add goals here you can add requirements that will really shape the outcome of the product and um have those really think about the impacts that it
Has in the determining functions you can also estimate or start to estimate impacts you can have models and tools to try to look for Solutions and an assessment during the evaluation to actually look how the impacts might turn out but always this is on the estimation level we don’t have that much
Information as I said um the product development process is um or can be a complex thing I will break it down for you a bit it is like an itative process of looking for results by the requirements that you have or the goals that you have so it’s always like an
Iterative cycle of synthesis and Analysis and um the thing is you want to really find an Optimum or you want to find the solution that’s not the solution there will be a lot of solutions but you will have also um quite a few Solutions doing that uh doing those activities and the
Evaluation is to actually make out which product concept or solution concept will be the best or the optimum um but there will always be trade-offs so the question is who defines the optimum and um there I have another quote that I bring again um that I think
Is really interesting in that regard The View that technology just changes either following science or of its own accord promotes a passive attitude to technological change it focuses our minds on how to adapt to technological change not on how to shape it it removes a vital aspect of how we live from the
Sphere of public discussion choice and politics so I think really interesting is this um this part not to technology change but shape it themselves and it’s not only designers and the developers or the researchers but I think it really be uh an effort of a lot of um stakeholders
So I put this slide here with politics society industry and Academia and I think all of those should really work together to create responsible values that we can then Implement into not only production or development but also research and uh thanks a lot Stephanie for your words I think you are one of
Those tiles and and really doing a a great job there um from the point of view um that I’m sitting in like the the methods of product development I also ask myself how to integrate those responsible values and there’s like a plor of um different methods and tools which can be
Individually really complex and its entirety very hard to overlook and oversee so we cannot really expect one designer to know it all or to have like all of that in mind and I think it’s really the interdisciplinary and the team that makes a good project or a good research
Or a good product um development and it was also I think um suggested in the standard the i e 7000 standard to really have a a composition of a team where there’s different roles and different people who can actually account for Society for Ecology ecological Dimension and the economic
Dimension and we have that in the industry and the companies we have always a lot of business developers that are taking care of the economic Dimension and we need to add the other two Dimensions into those teams into those research and development groups to actually account for those responsible
Values that we as a society um create or agree on and um I will finish uh my presentation with a little um oh I have this yes you can see this um with a little advertisement for um the work I did back during my PhD which was exactly
Looking at this picture and thinking how could we really do something about the goals the requirements and the assessment and um it’s a bit too much toh talk to you about it uh in five minutes but um there is a few things that I I started to think about or
Suggest as methods like a poiny social preliminary social assessment um I have some tables that uh suggest goals for you to start um your production or your development I have checklists or madeup checklists that um will help you gather requirements and really evaluate your product in the process so if you want to
Have a look at that um you can go here and uh see for yourselves that’s the the paper where everything is in its entirety and detail written down so um yeah yeah if you want to take a picture I’ll just hold on for a second one last there’s more okay uh and
You can always ask me afterwards um so yeah that’s it for me uh thank you very much for your attention and uh call introduce the next speaker please thank [Applause] you thank you thank you Michelle uh I think this really shows that if you want to do this exercise of assessing
Sustainability and Robotics in integrating the social Dimension uh to do a good job you also need to be very thorough and I think this shows that there are already methodologies and approaches that can really help you uh getting this right so uh now it is my
Pleasure uh to maybe first ask are there any questions for Michelle all right any questions for Michelle I’ve looked at the slido but no question sorry so then we can go straight uh and cut to the Chase and also go to uh wart from Temple Vander
Temple uh he is uh very uh involved in voxal sensors uh an organization and Company also uh subsidized by the cap Brussels Capital Region and inovis and uh present in the Brussels Center uh and he will be talking about the future of contextual Computing in sustainable
Robotics and as Bram said uh they have worked also uh together and uh we are very happy to have you today so the floor is yours thank you Mr chairman let’s see if I can raise bit outside of the bell curve yes this would be better okay so
My name is wer Temple I’m the CTO at voxel sensors it’s a small Brussels uh company startup company we found in 2020 we’re working on spatial sensing uh and also gaze tracking for contextual Computing and I will have some some more words on that later on so it’s my
Pleasure to be here but I come from outside of the the domain of of Robotics my background is in semiconductors sensing sensing systems so the robotics field while it’s very interesting it’s not my day-to-day uh world let’s say so I have had to catch up a little bit and
Refresh my mind and I’m sorry BR I didn’t read the homo roboticus book so I should have read that uh but instead of reading that I actually asked a robot so I asked it okay what is robotics again can you tell me what robotics is and it
Just came up with this very simple definition it’s a it’s a machine can be mechanical but can also be ethereal or purely digital that is capable of carrying out a function or a task in an autonomous way or semi-autonomous way so okay fresh my mind on what is Robotics
And then I asked it what is sustainable Robotics and then the answer was much longer seemed like it’s very conscious about what is sustainable so I had to summarize it and I picked out out this uh this line um what’s interesting is it presented me with a balance and it
Presented me with a balance of minimizing negative effects environmental effects and probably also so societal effect and maximizing the benefit of the robot robot function uh while keeping in mind is economic viability so it’s a very interesting definition as well it gave me then a list of bullet points a vast list of
Bullet points of every aspect in which a robot could be sustainable yes or no and I summarize it again in the in those three domains that also came back in in bram’s presentation uh so it seems like I have a lot of catching up to do and some
Reading to do the three domains are were very clear to me as well social environmental impact the life cycle of the product resources uh and also its energy consumption and then I tried to merge this view on robotics with my very very first impressions of Robotics uh and
That was when I was very young I was really Nam azimov who knows azimov probably a lot of you do right so science fiction writer and back in 1942 uh he he he put his first law in robotics in a very short short story and his first law robotics is uh very simple
A robot may not injure a human or through inaction allow a human to be hurt uh and while this is a very individualistic view of what a robot should and should not do I think everybody can subscribe to to this view I don’t want to be hurt by a robot my
Question was can we replace human here in here by humanity and does the law still hold and it turns out this law is much more strict than the sustainable view that we we had in in in the first page because if I say here that a hum that a
Robot should not hurt Humanity or allow Humanity to be hurt then this is a very very strict rule right and if I look at the threats that human Humanity uh has today again we can recognize these three domains that we put in the sustainability uh idea right we have a
Climate threat to humanity we have threats to our natural resources they are finite what we do with these resources and societal stability I think everybody uh also realizes these days with the upcoming elections that societal stability is also a big threat or challenge so I found a very simplistic but very
Interesting exercise to do to understand why why we’re here today because these questions about the the vagueness of the sustainability how much negative impact do I allow in order to reach a certain positive effect of of of of a technology that’s uh that justifies actually this
This session today I think to to make that balance and understand uh what do we allow okay so and the connection to voxel sensors I will pick the energy domain because in my view energy will be probably the biggest challenge in robotics in the future um when we think about the old robotics
The industrial robots right most of the energy consumption in those robots sits in the physical domain moving stuff around uh uh navigation and and auto automated movement while the the brain itself and the senses is very small part of the energy consumption of a robot but the
More modern robots and even just pure digital robots like chbt their consumption sits entirely in the digital domain and the Brain in the intelligence and in the senses and the physical power consumption becomes less and less dominant in the operation of of these uh these machines and while already today we can
See a significant footprint of this digital domain the brain it is still very light because these intelligences they are kind of autistic they live in their own world and they re to prompts which are very limited in scope but once you add much more contextual awareness aess imagine that I
I have a live video stream which pushes information like a bot like chat GPT imagine the power consumption that these kind of generalistic AIS will have if everybody starts using or every robot start using it in this way so there’s a challenge to give contextual information
To these kind of brains in a very very power efficient way and to illustrate this again I have these very short slides this is a traditional robot it senses in a very uh constrained domain right it parses what it sens and it knows what it’s looking for then it
Needs to think about what to do how to act and it acts that’s s very simply what this industrial typically industrial robots do now if you if you go to a generalized vision of this I have a robot in a very cluttered environment a very wide domain with a
Lot of objects a lot of a lot of uh experiences and a lot of different tasks that they will need to do it needs to sense much more it needs to parse much more and think much more so the brain or the the the the the thinking will need
To scale much more than what we have today and the challenge is to divide and conquer I think and to create um to create a contextual awareness very close to the robot in order to for him to understand what is relevant in my environment what should I focus on in my environment
And then to think and act and that is my link to voxal sensors because what we are trying to enable is this contextual Computing in a very uh very uh efficient way allow for autonomous uh entities to focus on the things that are relevant and not having
To to think about everything around them right so foxal sensus is really about contextual Computing what we we want to do is we want to give awareness of the environment the 3D World and awareness of the human to an application or an entity or robotic entity for them to
Better understand and with more efficiency understand what is going on and act upon what is going on now contextual sensing is a longstanding topic it’s been around for for a long time there are uh different applications in robotics like slam and odometry which require in environmental um uh information object identification
Segmentation Etc all that requires already uh uh contextual uh sensing however there’s a challenge and these challenges are mainly on the power efficiency how do we how do we continue to give this contextual awareness in a very power efficient way and especially in the domain where voxal sensor is
Active that’s mobile and augmented reality latency is also very critical it means that we don’t want to wait for the for the senses to sense before we can act upon it so the latency the time the time delay between sensing and being being able to act needs to be as short
As possible so that’s where the latency come in and typically latency if we want to achieve very low latency and high frame rate sensors will consume a lot so there’s a uh yeah how do you call it there’s a pairing that we need to break and voxal sensors tries to break this
Pairing by by delivering very power efficient sensing with low latency and we do this in two ways one we’re working on the sensing itself and secondly we’re working on the Computing side of the sensing data is being generated by the sensors and I have now some very simple slides on the technology itself
Um for the spatial sensing what we do is very simple we have this is a kind of technical slide I will go go through quickly but we have a laser beam scanner which is a very simple device which scans a dot in the scene we developed specific sensors uh which are tracking
This dots with very high temporal accuracy and with a minimum of energy and then once we have the projection Vector of where we’re scanning the dot and we we have the the vision of where the dot is from the camera we can triangulate this point and produce a 3D Point stream very
Simple and what you get is for example In you promised me it would play anyway I didn’t know where my mouse pointer was yeah anyway what you would what you get is then a Serial stream of 3D Point points which are which are created as we scan and as we scan we can build up the
3D information and by doing it that way we can create a very power efficient way of sampling the environment I will skip this video It Doesn’t never mind the benefit is that it’s very scalable we can choose with which density to scale SC we can choose how long to scan so
This flexibility allows to tailor your sensing system to the needs of the application and to the needs of what you want to sense and in doing it this way you can create an additional uh benefit in your power consumption of the system imagine if I’m looking at this scene I
Have a variable and I want to sense the environment in three dimensions typically the further I want to sense the more Optical power I will need with an active system right if I’m wearing my on my head then the floor typically is isn’t further away than 2 meters in my case 2
Meters in your case maybe a little bit less but it doesn’t make sense to allocate power budget to see six meters to those parts of the scene right so in serializing the acquisition of each point I can choose where to put Power where I need it and also I can change
The the the density with which I inquire the scene uh when when I need it so that’s from a sensing point of view and then from a I hope this video video runs from a processing point of view this is work that we did in the context with the IND project oh that’s
Too bad can I I don’t know it’s a good idea thanks way yes okay thank you so this is work that we did with the very Unity Brussel it shows uh the way to leverage this sequential information this data stream of sequential information and making your algorithms more efficient by also
Working on the Progressive data as it comes in so what we have is on the left hand side we have the scanned data which is being generated by by the system in the middle we have a traditional algorithm which will wait for the scan to be complete to have a 3D Point cloud
And then do for example a segmentation and on the right hand side we have an algorithm which already works on a progressive data as it is being built up and so you can see that uh I will run it again even in the beginning when we already have just sparse information we
Can already start thank you we can already already start to uh to segment this data and provide an already course estimate of for example segmentation uh and in the process we have an algorithm which is which can do the same thing in a shorter amount of time or in the same
Time do the do the algorithm with much less uh energy so there’s again using the benefit of the sequential nature of the of the of the sensing using this to uh reduce the foot the power footprint or the latency footprint of the algorithms that come F
Okay yeah I will skip it so in in a nutshell what what we we are doing is we’re working on the sensing side power efficiency and latency efficiency of the sensing to allow for low power spatial sensing which is very important to give contextual awareness to your uh
Autonomous entity we are working on the uh early classification of what we are seeing in order to for the subsequent algorithms to be able to focus well on what is relevant in y scene and what is not typically I don’t want to focus on my background if the task at hand is
Right in front of me right so I don’t need to process all these things in the background I just need to focus on what is relevant for my uh activity and this allows also to only provide the information relevant to that activity to for example the artificial intelligence
Uh Network that is running uh afterwards so this is how voxal sensors tries to put a stone in the river in order to reach some more sustainable goals in for for example robotics I also want to mention that we are part of a a European project uh
Horizon Project Bram team is also um a part of that project where we’re working on spatial perception and research and embodiments uh specifically for autonomous navigation or autonomous exploration um with also a positive societal footprint for example in disaster management Etc um and with this I would like to end
The talk and I welcome questions thank you any questions yes surely hello okay so maybe maybe you said it or May maybe I miss a shot I’m wondering so you make physical sensors so like do you sell your sensors to companies that make robots for instance or do you SE like
What is your or do you sell let’s say your software as a service to people who want to see things like I I’m trying to understand like where you are in in the market if I yeah so are we actually you’re welcome to join us in that exercise of thinking where we should
Position um today V sensor really is positioning itself in the XR augmented reality mobile consumer electronics World actually and while at the core we are building these sensors to enable this serialized scanning we’re also building sensors for gaze tracking so the core is really the the product is
The sensor but in making the sensor we also are creating subm modules which are more system level implementations as well as software layers in order to showcase the benefit of our sensors right so depending on the company that we interact with and depending on their interest of where
What they consider their their product that they would like to have there is room to to navigate between okay just the sensor the sub modules or systems which provide awareness um so there that’s a bit uh in discussion with different different customers okay thank you uh thank you for a nice presentation
Um I’m going to ask this question from the perspective of a I’m a father of a 29y old woman with autism okay and when I think about the difference between me and her being in this context sitting in this room so what I have learned or the way my senses
Have adapted from through culture and training is is I’m paying attention to the things that are important in this context I’m listening to you talk I’m not listening to other other things that are going on in the room as as much but my perception of her experience and sensation
Is she can’t do that all of the all this information is coming into her brain at the same level of intensity and what’s difficult for her to do is to incorporate the the the cultural things that we identify so of of being important to pay attention to so when I
Think about the creation of of a sustainable robot with uh lots of different sensory systems um it is is I think it I can I can understand how you can design a sensor for a specific condition that would do a specific thing right by by providing context to it but but but what
I wonder in terms of again creating a sustainable sensory platform for a robot that can do many different things you need to be able to um downplay the sensory array with one sense area increase it with another and I need to be able to change that depending on the context of what I’m
Asking that robot to do and I need to have like 16 different sensory systems running at the same time and I need to be able to to transition from one to the other so is it is is that is that maybe I’m presenting the context wrong but is that the real contextual
Challenge for maybe not for you or for somebody to to to figure out how to um provide maybe it’s a programming platform to change which centy system I’m paying most attention to in this particular context because this this is this is what’s relevant at the moment it’s it’s a very interesting line
Of thought I think the the way I’m I’m thinking about it is I’m I’m imagining a system where the senses are always on and I want to have a low power always on sensing regardless of the modality I don’t want to turn off things because I might miss things but
What I want to do is very locally with a with also a low energy footprint locally what they call Edge computer locally I want to isolate in in in all all of these triggers of my environment in all of these senses what is relevant for my
Context and what I want to do today or at this moment and from that moment on I can feed it into higher level intelligences which can then process the right senses in the right location at the right time but if you feed all of these senses to the higher level
Intelligences I will need a lot of it will consume a lot of power to try to figure out what what it needs to do right so I think this divide and conquer approach where very locally you will allow what we are able to do and your
Daughter is not able to do is our senses are your senses are not turned off right they’re always sensing but you’re already filtering very early on what are you going to feed into your brain or what are you going to pay attention to and that is I think what we would like
To achieve in in in this line of thought in this aware context awareness and I think one of the POS posibilities we we are thinking about these uh AI assistants uh the companions that you that you wear on your head imagine that we would be able to help these people
With with issues of filtering out the information that that a device filters the information and is intelligent enough to know what context is relevant for her at that moment maybe these people who are suffering from autism would would be helped with a system that does that job for them sometimes she’s
Right identifies a sensory issue that when she identifies a sensory issue that she thinks is so important that she needs to make us aware of it she’s 100% right that that is the most important sensory information that we should be getting at that time wow okay interesting thank you there was another question
Yeah uh I have a short question maybe a follow up on that uh first of all thank you very much for the talk I really enjoyed it uh I’m interested in finding out how reconfigurable is your system so you have different setups with you’re talking about robotics so you have
Different uh uh you know scenarios so how can you easily reconfigure your sensor systems have you thought about this and how can one one think about it uh I’m trying to imagine your concept of reconfigurability what what are the different dimensions that you have in
Mind um I can just say what I think about reconfigurability yes so the sensing system is really adaptable you can choose where to position your different sensing units and illumination units it doesn’t matter whether you have a a baseline of a meter or 12 cmers depending on the application that you
Want to have within a certain geometric configuration right once you fix that there is again reconfigurability or flexibility in what kind of density do I want to create so how how dense do I want to scan my scene what kind of power do I allocate to this sensor modality uh in and in
What kind of time frame do I want to have certain uh information so there’s a lot of flexibility which makes it probably for some people hard to use as well because with flexibility comes a lot of optionality as well that you need to manage um but yeah the system is very
Reconfigured we can talk after in in the in in the in the pause what you have in mind one question in front of hi uh thank you for the yeah okay okay good so um um so if I understood correctly you’re designing the semiconductors in a way that they’re
Very energy efficient so you’re already taking some sustainability aspects into account but are you also is there a department that is maybe measuring the water consumption of production because semiconductors are very water intensive and you can only use use drinkable water because like small fragments in the
Water might already um scratch the wers so I was wondering if you also have like a different department that maybe looks at other environmental impacts that are um yeah produced by the semiconductor production in general mhm the short answer is no okay uh as a small startup that we don’t we don’t
Have uh uh those those those different functions in in the company although we working with um with fari on uh and the V on on our e on understanding our ecological uh footprint but we don’t go as deep into our different suppliers because we are fabulous right The
Foundry makes the semiconductors for us that kind of impact as for many things in our lives it’s shielded from us I don’t I don’t know or I don’t have to care well maybe I have to care but I don’t have I don’t have to see daily
What the impact is of the things that I’m I’m using so it’s probably something we we should consider um maybe I have to add that um while the energy footprint is very important the reason why we set out to do was not about the energy footprint as such that is not the
Business goal that we have it is by Magic maybe that those business goals of low power which makes our technology very suited for wearable devices aligns with a more General uh goal of low power devices which are then uh relevant in the robotics context but we didn’t found
Ad voxal sensors to reduce the ecological footprint of sensing I just want to clarify that because in a business that is not the it’s not a viable goal economical goal thank you I have one question on the slid but I think your point also makes me think of there is this
Initiative in the region also financed by nois called sustain uh that we host also hello to Arian in the back uh that also provides to because the size of the company will also dictate also how much can you put efforts money into investigating all that and so to address
This issue the region has this free service for smmes in the region uh called sustain so I also encourage you to to check it out so now that the publicity is a bit over I will ask you are your customers asking for footprint analysis and do you know what your
Customers use uh sorry and do you know what your customers use for sensors is it only positive impact uh do you investigate a bit the type of use that they will make of um okay on the first question do our customers ask about our our environmental impact of the sensors no
The primary concern of using lowow sensors is to be able to enable wearable lightweight always on devices so function is a primary concern uh I guess that the footprint analysis will come at a later stage when we enter supply chain uh considerations uh which where we are
Not there yet as a startup right so I think it’s a bit early for these uh these footprint analysis but likely it will come it it is important thank you very much a round of applause for w thank you again for your time so you mentioned quite a lot of
Things things that are a very nice segue for our next presenter uh Sophia FAL from bone University uh in Germany uh she is a PhD researcher there representing proudly one of the spare heading Labs on AI and sustainability uh in Europe she will talk about digitalization and artificial intelligence transgressing planetary
Boundaries so really looking at this critical uh macro look at uh what do we put within some robotics systems so the floor is yours good morning and thank you for the introduction and also thank you for having me here today really happy to present my research and yeah as um Carl
Said my um part of the research that I will present today is digitalization and AI transgressing planetary boundaries and I’m from the bond sustainable AI lab I was also asked to briefly introduce it so these are my beautiful colleagues we are sitting in Bon at the Bon University
And the lab was founded 2021 by Professor Amy uh from weinburg and she has a uh background in Robo ethics care ethics value centered design and together in the lab we combinedly cover the topics of um bioethics um extractivism Ai and the labor market how is AI changing the labor market how is
Labor or the work uh workers changing with AI and Robotics then we are covering the colonial ai ai from a majority World perspective AI from indigenous perspective and um then also AI from a a sustainability perspective from environmental perspective which is then um my part I’m a resource Economist
Working interdisciplinary and a team full of um philosophers and ethicist so I’m bringing in the environmental aspect so we have already talked a little bit about sustainability what where can we put it we have two buzzword colliding here we have sustainability one huge password we have artificial intelligence
Another huge buw word um I don’t have time to go into details for both to describe what they are but um to roughly um picture frame we have social uh sustainability environmental and also economic sustainability and while they’re interacting and interfering a lot with each other I will try today to
Really focus on environmental sustainability but when we have environmental impacts we should not forget that they always have consequences for nearby communities for the population living nearby and then this also has effects on their economic um situation so it’s really hard to look at it in like a isolated way but um yeah
Today we’ll try to really look in environmental aspects and this I’m doing with the help of the planetary boundary framework maybe a few of you already know it but I’ll briefly describe it it was developed 2009 by Johan rrom and his team and since then it has been um
Criticized updated and re-evaluated and you can see here how it changed over time that we had three boundaries crossed out of 9ine in 2009 then four crossed in 2015 and now we are at the limit of six boundaries crossed um last year so what are we looking at here we
Have the climate change boundary which is the most popular one it’s measured in carbon emissions so atmospheric carbon emissions measured in particles per million and each boundary has a threshold like a control variable uh with a maximum limmit so the maximum limit for carbon emissions were 350 PPM
But they were already transgressed in the 80s so a long time ago at the moment we are at 420 particles per million but this is only one part we have nine earth systems that are important to stay on a planet that is hospitable for human and non-human life on the planet which is
Called the safe operating space this is the ball in the middle so this is an area where we know the planet is in a state that supports human life animal life and we can live here so it’s a holen like State um the further and each of the boundaries have
A threshold and a control variable that we transgress and the further we transgress it it’s not a Tipping Point it’s more a regime shift that we enter and we don’t know how likely it is to support human life as we know it so it is a zone of increasing risk it’s not a
Transgressing not a Tipping Point which is really important and um today today yeah one too much uh here’s um the one from 2023 and a bit bigger scale so you can really see what nine Earth Systems we have here today uh we have novel entities which is measured in uh
Chemical substances that are hazardous and toxic to the environment but are released to the environment without proper safety testing then we have a stratospheric ozone depletion which is currently not a problem uh since we had the Montreal protocol in the 19 you can look that up again but um this is a
Proof that if you have a drastic change in an earth system policy can actually really quickly interfere and change and then the boundary was immediately um back to a safe space and we don’t have uh huge oone holes anymore um then atmospheric Arizone loadings are particulate meta em emitted by vehicular
Emissions or stone crushing and Mining um this is another one then ocean acidification is um when the ocean takes up a lot of carbon emissions the acidity state of the ocean is changing and this is then as we heard yesterday very um problematic especially for coral reef
And um calenic species um so these are very interrelated with a climate change boundary then we have biochemical uh flows which are phosphate and nitrogen um freshw change which is measured in blue and green water so one is freshwater consumption and the other one is um root moist uh
Availability for plants then the land system changed is deforestation of land cover changed or land system changed into a different cover biosphere Integrity is genetic diversity so Extinction rate of animals and then we’re back at the climate change boundary and to have like two sub um control variables measuring the same um
Earth system so the question is now now how are Ai and digitalization contributing to transgressing these planetary boundaries um we know the popular climate change boundary we are measuring energy consumptions um but how are AI other than that impacting these other eight planetary boundaries so the question
Really is how do I bring together AI which is kind of like an intangible thing that we always know is we’re surrounded by but how is this very intangible concept imp perfecting this very hard physical planet that we have here and um I’m looking at the AI Hardware from a data center perspective
So data centers can be the heart the nervous center the backbone of artificial intelligence without data center as we also just heard semiconductors are necessary to process the large amounts of data that we have um without the training of the AI it would not be
Existent um so on the top left you can see a Facebook data center in Ohio you can already see it is very spacious it is large buildings then we have uh what it would typically look from the inside huge server um boards and then we have a
Server blade that is um set up with multiple gpus graphic processing units and then you can see a individual like this is the state-of-the-art a100 Nidia um graphic processing unit without this um it is very hard to train um AI systems but I’m always talking about digitalization in AI because you cannot
Really say what part and what um amount from the data center is actually used for training AI especially when we’re looking at cloud services you can rent like parts of a data center while one is used for AI the other one might be used for a streaming service or something
Else so it’s very hard to distinguish and I go into more detail uh in a paper but not today but um so AI is part of this problem but not the only one um and when researchers usually look at um sustainability of AI from this perspective of a data center they come
Up with um AI training and the en energy amount that is used to train a data center or to train a AI system in the top you see the very popular gpt3 it consumed in one training round 52 tons of CO2 but this is just one training round so usually especially commercial systems
Are trained multiple times retrained when there’s like an exploit in the system they’re trained again again um and this also takes only scope one and two maybe emissions into account but not larger scope emissions so we can compare this to the average human life you can see the average American and I would
Like to point out I’m pretty sure it’s North American life is 18 um tons in a year of the average life in one year and then the average human life on the planet would be 5 and a half tons um so one training round of jet GPT is equal
To the average human living 91 years on the planet so if you really want to train and retrain and retrain you build a whole new set of population on the planet in other words um but we already have a lot of humans here so this is the usual approach that um researchers look
At but when I look at AI Hardware I see something else I look at this so we’ve maybe been we’re familiar with a few of these pictures from today already I think especially this one Michelle used to also um but yeah on the top left you see uh gold mining in Western Australia
Then here you see lithium extraction in the atakama desert so lithium extraction is necessary for lithium batteries and you can see it’s through evaporation ponds and lithium takes to produce one ton of lithium we have to evaporate 800 square Kil square meters of water and this is done in the desert that should
Be counterintuitive for anyone to evaporate water in a desert but we’re doing this for our batteries or electric anything that we have in electric and um this has larger impacts on the environment there on the biodiversity also on the population then we have on the left we see for Semiconductor Fabrication in
Southern and Central Taiwan it is the most uh popular region for or the most the largest preps for semiconductors are located there it is a huge water footprint to produce semiconductors so here you can see in the rain season the water reservoir only has 10 to 20% % and
Usually at the rain season it should be at 80% and then in the end we see the disposal of e-w um because all of the products that we’re developing have to go somewhere and this is um how I try to analyze the impact of AI on the different planetary boundary system you
Can already see that multiple systems are affected by this but we can go into a little bit more detail um we read oh no has a Time okay so we can already also look at it from a life cycle perspective that we start with extractions that are necessary minerals and materials that go
Into the AI Hardware then we have the manufacturing and the processing of the materials then the training and the use which is already quite uh we are quite familiar with and in the end we get to the disposal so looking at the first um oh okay sometimes when I send PowerPoint
Presentation they get good with mac and not Mac um but yeah so the climate change boundary is uh data center energy consumption processing of minerals and materials also semiconductor processing has a lot of um energy consumption uh related to it and then we have the manfacturing of electronic and micro
Electronic prods but this is not the focus of today so it’s not a problem that this looks like this so the now we’re going to into novel entities so novel entities I said is um the release of hazardous and toxic materials into the environment without proper uh safety
Testing and this is for many materials that we’re using and they are released during multiple steps in a life cycle we have them during the mining process we have them through Wastewater from manufacturing processes again semiconductors use a lot of chemicals to be produced and that goes through the
Wastewaters just straight into the environment and the groundwater then we have through informal E-Waste disposal e electronic devices themselves contain a lot of materials they themselves already detoxic these can be um lead Mercury platinum gold neodymium and they all with our e-w disposal go right into um
The soils the groundwater and also the air and ew is the fastest growing waste stream that we already or that we are facing currently in 2019 it was almost 54 million tons and 83% of this were unaccounted for we don’t officially know where they went but um there are already
Some um um yeah some approaches to test where our E-Waste goes and it’s usually southeast Asia Africa and also um Latin America and they’re just illegally disposed and then through informal um informal recycling methods they try to still gain some of the valuables from the electronic waste then um it is
Estimated that in 2030 we have almost 75 million tons of ewaste so it’s a growing problem and we’re we have to find a solution for this uh by for example designing products differently then we have as I said a steros feric ozone depletion is not really a problem so I’ll just skrip over
It then anist this um atmospheric orizo loading is measured in the difference from northern and southern hemispheric difference in particul at meta which was inconclusive because we are mining and around the globe northern and southern hamis and also through trans a it’s also a global interactive um supply chain so
That was very inconclusive how this is supporting or changing the inter hemispheric difference then we have uh the ocean acidification which is largely um interdependent and dependent on the climate change boundary since the oceans take up 30% of the atmospheric CO2 that we push into the atmosphere and then
Changing uh the acidity state of the ocean for the uh bi geochemical flows it’s also inconclusive it’s also mostly impacted by the agricultural sector so um we’ll neglect this at this state then we have fresh water consumption and this is largely affected by Hardware production especially semiconductor industry we have energy
Production and this is a huge trade-off that we also have to take into consideration we always trying to go more into renewable energy but producing renewable energy has also largely a high water footprint um biofuels are very water intensive hydrogen is very water intensive so we have to make some
Decisions here if we want to maybe um help uh the climate change boundary and reduce it or do we say like no it’s more important that we um secure the fresh water change and drinking water so these are trade-offs that we always have to take into consideration and then another
Point is data center cooling um which also takes up a few um yeah fresh water then coming to the land system change land systems are changed for multiple reasons within um the life cycle we have mineral extractions as I already said for structural elements such as housing for the computer
Hardware would be um aluminum and steel and then functional elements um to like build the little semiconductors would then be um copper Gallum germanium indium lithium talum tum and so on there’s many many minerals going into this um and then we have the infrastructure development for network connectivity and data center you could
See the data centers are very spacious so they take up a lot of um room and this has a direct impact not only on the land system but also on the biodiversity and the biosphere Integrity because all of these network and infrastructure buildings you need um roads public sewer
Lines you have graywater infrastructure power lines and they all lead to Habitat defragmentation so it’s more difficult for animal species to uh cross and they become like separated in ever smaller patches and then it’s very hard to maintain species um population size because they’re separate and then as I
Said before the mineral extraction for lithium for example in the atakama desert is impacting um flamingos andian and James uh Flamingo species they used to live on the surface water but now there’s reduced surface water availability and you can already see that the population sizes are thinking to a very
Um low level that it’s hard to maintain um sustainable or long-term um population size to yeah not fall below the um yeah size that would maintain dises in the long run so we can see um six out of nine boundaries are definitely impacted by the examples that
I brought and we’re always talking about one of the boundaries but there are five others that are really highly impacted and there are studies showing that um mineral extractions for example have a larger impact on Earth Systems than the climate change boundary or the carbon emissions so we can say for my research
Question um how are AI in digitalization contributing to transgressing the planetary boundaries in multiple ways um the next question that follows up to this is of course to what extent I’m working on that right now so hopefully I could tell you more about that soon um and now for some take-home messages I
Would uh conclude with our daily online activities and AI development undeniably contribute to pushing us further into the zones of increasing risk uh we are facing ecological implications far beyond energy use and carbon emissions and environmental impacts of develop AI development stages are geographically unevenly distributed the majority world
Is mostly suffering from the development and the minority world is uh benefiting from all the great things that we get from these very necessary tools um and out of habit a question that I like to ask um is sustainable AI illusion or reality um for me at the moment is more
Illusion but I saw the agenda for today and it looks like we have many cool alternative approaches to develop these systems I’m very happy and um to hear some optimistic point of you towards this uh topic so yeah um thank you thank you sopia uh you have uh thank
You to all the speakers for having been also very efficient we have plenty of time for questions so is there a question in the audience yes please Shirley thank you for your very interesting presentation um I was not aware I mean I I on the general I was
Aware that there were problems but you pointed out in more detail so thank you for that um can you share a bit more on how you are planning to answer your research question the to what extent well uh yeah so the one on the um so I’m looking into life cycle
Analysis assessments from products that we have assessed which are semiconductors and then we have the output as global warming potential or water use and I’m working on conversion functions to adopt them to the control variables that are used in a planetary boundary system and to get rid of uncertainties I use statistical methods
To replicat uh replicate the numbers that I have to to come up with a distribution of data to see how it which likely or like in a high likely case in a low likely case but yeah so it’s more of a conversion from life cycle assessments to put them onto planetary
Boundaries I cannot assess all of them so far I think I have five that I can take into account but it’s already a start okay thank you and maybe a suggestion that it would be good that after you what you found that you find a way to communicate it also to people
Already like working at the policy level so you can influence them with your research I would love that thank you another question okay we have one H yes thank you great thank great talk thank you have you thought about what what are the allowable limits so if we talk about
Resources mining for example and all these uh these sectors of impact that you thought about suppose that we cap them on the Green Boundary have you thought about what is available uh well we have scars resources and then it comes to a question like the most sustainable
Resource is the one that stays on the ground right so it comes to the question how do we want to allocate these resources what are necessary tools or products that we need to develop because there’s just a skar like there’s a limit what we can produce if we really look at
All the deposits underground what we have um it will not it will not be L right so we at some point need to make decisions um and up we’re here okay so um I don’t have an answer for you to say like okay we have to stop at this amount
But I can tell you that we need to consider what we want to use it for because it has an end the resources are not Limitless so we will come to a natural end at some point so yeah but like like with oil people have have always thought
That there there is an end but in the end there’s more and more drilling there’s more and more oil fields found Etc so while while everybody knows that at end is the actual reality is that people are searching for more and keep on finding more and I think the same is
True with the scars elements so if you if you don’t cap it that people will still find lithium Fields somewhere which are minable or find some other deposit on on the deep seas or something so we are very very resourceful in finding other resources so maybe
Y yeah yeah we are very resourceful in finding them um the question then is to find these new sources there’s a lot of investment going in and um then the impact will just shift to then you have more land system changes then you have more water consumptions
Maybe we find more sources on the ground but maybe we don’t have the means to extract them anymore at that time we never know like I don’t have an clear answer for you right now but uh it’s a it’s an interesting question yeah I have a question maybe in
Continuation with the previous um uh question so uh I I I’m um curious about your comment how far are we to make a U holistic uh assessment of the product that we are using in terms of life cycle assessments for example the microphone that I’m holding now what is the
Sustainability assessment score of this maybe we should if we could be able to kind of score every product that we purchase that can make it impactful social effect I mean at the end so what do you think how far are we to make this assessment for every product that we so
I think it’s very important but we’re not very far in doing this this has multiple limitations one limitation is it’s not open source data available so people would need to pay to do these assessments and as we live in a capitalistic system you don’t really gain much from labeling if just one
Microphone is labeled and it’s not a policy standard you have to label them probably the company will not choose to make this extra cost to evaluate the system then the next point is that the data sets that we’re using are often a little bit outdated so you will never get really the
State-of-the-art answer that you’re looking for um and the last point is that um some data sets I also try to find life cycle assessments it’s very hard you can find a few that’s why I’m trying to simulate data as well based on the one that I found to increase the number of data
Points that I have because it’s very limited that you what you can find um and then I read a critique that data sets for micro micro electronic products are very outdated or not even taken into the data set because the harm from small product is not that bad so we don’t
Consider the quantity of what what we produce it’s more like that’s such a small part we don’t really need to assess it we need to only assess the big items like cars or huge um industry halls or our buildings but we don’t look into the smaller parts because it’s not
Really realized that the quantity also makes a huge difference with with the scaling yeah which is not correct actually I mean if you make the quantity huge for example exactly but we’re just looking at this one part like oh but this one little pin it doesn’t matter so
This is a huge limitation in that work yeah thank you so is there another question here yes I’m just curious to see if um you have or you imagine a contradictory issue from the perspective of um the countryes getting the E-Waste uh for example if they get advantage and maybe
They do of getting e-w could it create a demand and um yeah a demand for ew so there was um there used to be export to uh countries to they would get money and the E waste uh but that was banned so now it’s illegally disos there
Often hidden as like um a donation so I’m donating 20 computers to a school in Ghana and one works the rest is damaged but when it arrives they say like oh it must have been damaged during the transportation so we can’t take it back and um the valuable of this is there was
A huge issue that called AB ABB blosi and Ghana it’s near Akra is the ew site that I had on my picture and there was this huge informal industry built around the ew so they really adopted to E-Waste being sent there and extracted the valuables that you could find there’s
Still some copper there’s still some gold there are valuable materials inside and through informal e-w recycling through open open AET burning through um AET bath they tried or the population they tried as much as possible to regain the valuable materials um which is then there was like a lot of
Studies conducted the remaining toxins seep in the groundwater in a one kilomet radius um the air there’s pollutants in the air people breathe it there’s a lot of carcinome uh detection and um then from one day to the other two years ago the government decided oh this is a bad
Image for the city we just going to erase the whole e site and they just came with bulldozers erased the whole site and said we have to relocate outside further outside from the capital and everyone lost their homes they lost their business they lost the little that
They had and this is also not a solution right it’s not a solution to bring the E there but it’s also not a solution overnight to just without a warning demolish the whole site and take away your livelihoods your collection of materials your houses um and saying like
Oh but this is good for you because now you’re not getting sick anymore but what happened it was collected in one dumb site people moved into more into the city into individual living rooms and doing this now in their backyard in the living room open acid burning where the
Kids are right there so they just changed a huge image of ABB blosi we have a bad image in the country for this we need to raise it but then it just came a whole different new problem so there’s no easy solution for this as
Well like yeah I don’t know if I really answered your question but another input thank you um is there another question in the audience otherwise we have on slido uh a first one uh we strive for Innovation AI Etc and likely can’t stop the developments your presentation shows
How urgently we need to change what is the solution in your view I would love to know the answer to that um so my solution would be to really think about what systems are necessary what do we need to develop what should be developed and not just
Bring anything on the market and be like oh here’s a tool that you need for two days and then you get bored with it and it’s uh becoming obsolete so we are striving for Innovation yes but maybe we should reconsider a bit uh what we want
To see being produced and um create them it’s very connected to designing for sustainability and the design of the systems how do we design them what do we design should we design it I think it we should reflect a bit more in what we’re producing and not just like in this
Consumption Society just bring everything out there and then see how the market reacts and then just dispose of it if we don’t need it so maybe have some more insight what is really needed yeah thank you um a follow-up question here maybe just as a as a thought would it be interesting
To to remove when I sell a product it seems like I just sell away the responsibility right so is is there has have people thought about keeping the responsibility at the producer side even even though you’ve sold your physical Goods uh yeah there are approaches that try to make producers responsible um but
They just pass it on as a on an increased price on the consumer what happens often uh every time they gets like a carbon tax or something on the producer it just reflected in higher consumer prices so it’s something that we always uh divert and it would be
Great if we have for example what fat um Fatima also related who is saying like um stamps on the product itself when you buy it that you really have to be accountable and that you have to implement for example ice cycle Assessments in the TR uh product
Development and then the cons like we have a more and more um sustainability oriented Society then maybe the consumer would make a more sustainable Choice and then the manufacturer also needs to follow up because this product is not being sold that much on the market
Anymore um but for that we would need to start with the maybe like the labels and then see how the market reacts yeah all right thank you so we have a multi-entry question for multiple speakers so we will have this as a last question so Michelle said to include
Sustainability at the start but W mention it’s difficult for startups and often not the main concern how can we change this so easy question for you to start with Sophia then we will go to Michelle and to word yeah um um yeah um I don’t know if it’s really
Difficult for startups maybe it would be I’m not in a startup so I don’t know but I would think that maybe if you start on a smaller scale that is easier to implement right from the beginning some sustainability criteria or assessments then after if it’s like a huge company
Already and we have everything said and done and then afterwards Implement some sustainability criterias is for me person Al seems to be more difficult than doing it right from the start but that’s just a personal opinion yeah Michelle come here on the stage um yeah that’s a very good
Question but very hard to answer easily um I think it’s always uh also the question that people had at conferences when we talked about this issue and uh in my point of view it’s really the the motivation and the incentive like what would motivate um startups or Industries
To change their behavior and I think it’s um a bit coming back to the slide that I had with the different parts like Society politics uh researchers developers um there will be or you could say there’s maybe some bottom up and some top down um things that can happen
But I think it’s a lot of um politics law regulations um it’s consumer Behavior it will be also a change maybe of the values of the designers themselves um I think one really important thing is education uh it will take some time but um as we can see
Sustainability is a growing interest a growing topic it’s getting a bit bigger in different uh conferences so there might be uh a generation of designers Engineers that grow up with a bit more of Education in the topic itself and that could also change I think the view
If if the companies cannot employ any more um designers that are willing to do whatever it takes to to earn and gain more money um it will change and also I think the the question that I think you had um internalizing um the the costs like not um like CO2 taxes that’s
Something or a good example for internalizing the cost to make um the company really pay for its product or its impact um yeah thanks so after two answers from the academic perspective now the startup also uh response I had time to think about it I didn’t come up with an answer
Uh I I would say the sustainability question is is probably inherent in in everything we do but the the business goals are or a primary concern at first right when you’re a startup you you you have to think what do I do first and how do I position my company so that
My products are appealing in the market and that product Market fit today is not around sustainability for voxo sensors we are a technology platform a technology provider so our product Market fit really is oriented around the low power low latency proposition and it just so happens that this low power
Concern aligns with a lot of low power concerns in sustainable goals um um but it’s not our primary driver to service these sustainable goals our primary driver is to create low power low latency technology which can be used in the market uh and and create exciting
Applications thank you so a big round of applause to the three speakers I hope you you got to have a bit of a better understanding Al also of uh when we integrate uh contextual Computing AI what do what are also the challenges that come with these different aspects uh do we integrate in
Robotics and so yeah robotics being a complex field where you have multi- uh disciplines working together at all times this was a good example of how you need to have this uh intersectional view looking at uh all the challenges coming from different disciplines so thank you
Blah blah is over coffee is ready uh for you e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e
E e e e e e e e e e e e e e e e e e e okay sorry uh welcome back to the Forum um we will be uh heading into our second session which will be on the use phase uh battery and energy consumption and our first speaker will be
Alejandro yes um and he is or his background is a mechanical engineer and he joins us from the University lib brel and um he found his way into uh the field of control and Automation and into um batteries and is one of his research topics or interests is batteries and
Battery management especially fast charging and he wants to elaborate that for us now a bit more in the tradeoffs of this fast charging so give it up for [Applause] aleandro thanks Michelle for the kind introduction as he said I am Alejandro and I promise you three things by the
End of this talk first of all you will know a bit better how batteries or lithium ion batteries are commercially charged second you will know why we have to improve the charging uh protocols and the charging mechanisms to make them more sustainable and three of all uh my
Last promise it would be that uh we will know how the commercial charg Char charging mechanisms can put in Jeopardy the life of the battery and also we can age faster the batteries but why batteries and Robotics maybe you are wondering well we need to increase the
Autonomy of drones and robots and to do that we have to get rate of cables and to do so we need to plug batteries to provide all the power that such devices need but the main problem is that when using batteries we have a tradeoff between the usage time and the charging
Time of the device why because the charging time is higher than the time that we can use theot or own so that leads us to two things either we do a battery swap we rep place our battery to continue working or we charge faster our batteries and why we want to charge
Faster the batteries because we are human beings and we are not so patient actually surveys since 2017 show that we don’t want to wait more than one hour for the full charge of our devices and when we are working with robots and with drones that is a kind of charer that we
Are facing this is a bit more fancy than the charger of your phones but at the end it’s more or less the same we have to decide how to charge the the battery so there are two questions that arise at the moment how a charge works and second
In these kind of charges and how I can choose those parameters for the charge well first of all let’s solve the first question commercially charges use uh an algorithm that is called the constant corrent constant voltage algorithm in simple terms we charge the battery with a constant current to will reach a
Voltage and that corrent is defined as a c rate I mean a relationship between the current that we are feeding to the battery and the capacity of the battery that means that if we are using one C we will charge the battery at least in this stage in 1 hour and if we
Increase the CATE we will decrease the charging time open reaching a constant or I mean that voltage we will keep that voltage constant while decreasing the charging current so the end sorry I jump uh a bit further there will be an end of charge condition that will stop completely the
The charge either a minimum current or a charging time so the second question is how can we choose those values the voltage and the current that we provide to the charger well first I wouldn’t say what is the best value I would say what is the safest
Value in terms of the voltage for those reducing the voltage can be uh can be uh only um a problem of how fast I charge but in terms of the current as I said before if I increase the current I am reducing the charging time 1 C is 1 Hour
2 C is half an hour and 4 C is 7.5 minutes but is this safe how can we ensure that this is safe for the battery well we conducted some experiments with commercial batteries to test the life cycle of the batteries and to do that let’s take into account a simple single
Cell a single battery that is a Sony battery that you see there 3,000 milliamps this is more or less what you can have in your phone in a flatter way and we use high Precision equipment and we charge and discharge the battery charging with the cccv that I just
Described and discharging with a conservative approach to neglect the effect of the discharge I am focusing everything on the charge because the discharge depends on the device that we are using okay and what happens then what you can see is that if we run 100 cycles of charge and discharge at the
End of those 100 Cycles instead of decreasing the the charging time when we increase the charging current what we are finding is a wall we are hitting a wall that is increasing the Aging of the battery in terms instead of reducing the charging time and perhaps you are wondering this
Is only with the Sony batteries with this manufacturer no actually you can find even weirder WS of this limit of this wall that you’re hitting this is a case of LG batteries that it happens if it a low current and also in the case of Samsung batteries it happens the same so
When you are using a battery and you are saying oh my battery is not working or I am charging it and it’s aging or is not working anymore after a few days is because you are using it in not the best way and why perhaps you are say okay
After 100 Cycles I still have 10% or 30% of the of the charge like in this case this is the normal plot of the capacity over the time of a battery and you can see okay at the end of The 100 Cycles I still have
50% and in fact the end of life of a battery is when the capacity is reduced below 80% that means that at that point you have to replace the battery because otherwise you are promoting fight hassles you are promoting that your device is in danger due to the battery
And as uh Sophia said before at this point we have to replace the battery and we are creating two scenarios first of all we are generating the demand that come from overseas because the materials come from Asia a Africa and South America and also we are generating waste
Because replacing a battery it doesn’t have a second life after the 80% we cannot reuse that battery for any other purpose so how can we improve that well some companies as opo BBK they propose some modifications of this cccv to contact this agent phenomena and that is a reason why in
Academic in the academic field you can find over or I would say more than 100,000 Publications on battery management system F fast charging at strategies and health aware approaches but the main problem is okay we know how the Char charges the battery but what happens inside of the battery what is the
Problem of increasing the the charging corrent so how can we proove the CH the the charge of the battery extending the lifetime that means that also extending the sustainability of using that battery well this is a battery if you see it from a microscopic Viewpoint we have a
Negative a negative electrode that is the anode and a positive electrode that is a cathode this cathode is composed of lithium ion lithium materials that when the current passes through generate lithium ions oh you cannot okay okay sry for the um for technical details okay so coming back this is a
Battery and we have a positive and a negative electrode the positive electrode is composed of lithium materials and also we have Cobalt we have cium we have aluminium and the negative one is composed of graphite what happens when we pass a current to charge our battery is that we generate
Lithium ions and those lithium ions go from the positive electrode to the negative electrode we can think of the negative electrode as a sponge that is accepting the electrons sorry the lithium ions that are generated in the positive electr the main problem is that if we impose a charging current
That is really high we’re increasing the flow rate of lithium ions and that uh well actually this is how that happens graphically and what happens is that when we increase the charging currents we are increasing or we are promoting also the appearance the urrent of parasitic reactions in the negative
Electrode what we are promoting is the plaum that means that the lithium ions are becoming metallic lithium and we are Lo losing that lithium within the cell and also we are clogging the spots where into that sponge the lithium ions can go and from the Viewpoint of the Cod also
If we improv high voltages high currents and so on WE promote the solvent oxidation we have an electrolyte that is surrounding the electrodes and the solvent of that electrolyte is being oxidated by um oxidized by the by the the lack of lithium ions so we can model this and that
Happens in many battery management system as a simple electrical model and actually that is okay for the microscopic Viewpoint because we can characterize the voltage and the current but we are not characterizing what happens inside the cell so the other approach is to do it from an electrochemical and thermal
Viewpoint in that case we can accurately characterize what happens inside the cell and also we can limit that degradation uh um phenomenon that we saw before so by doing that what we can do first of all or characterizing the battery in terms of electrochemical and thermal uh models is to monitor what
Happens inside the battery what happens at the core of the battery what happens at the electrod level and according to that we can characterize also where is safe to charge the the battery meaning that we can find a relationship between the electrochemical state of the battery and the charging
Corrents to find where the charge is possible without accelerating or without promoting the side um I mean the degradation phenomena the main problem of this is that uh this is chemistry dependent this is temperature dependent and this also depends on the core temperature of the battery why I say chemistry dependent
Because not all the batteries they have the same chemical components within second this depends on the temperature because the plating for instance is more prone to happen or more accelerated at 5 degrees that and 4 that at 40° so that means that we have to be aware of how we
Are charging our batter is not only from the microscopic viewpoint but also from the environmental Viewpoint if we are in the middle of the winter outside and we are charging the batteries for sure we are promoting degradation mechanisms even at low conr however that allows us to create
The map and to develop a healthware uh Charing strategy that is able to contact the degradation mechanism or at least to limit it because we will always have calendar AG I mean batteries will degrade anyhow um and what we can find is that if we come back to the case of the sunic
Cells if you remember that was the plot that I showed before for the case of the cccb we can reduce the charging time in 50% while having the same degradation of conservative and also the commercial I mean the most conservative commercial protocols that is what the manufacturer
Recommends to put you in other words this um or to give you another perspective what we are doing is that we are decreasing the idle time of the battery due to the charge we are increasing the efficiency while also increasing the sustainability because the battery will last the same number of
Cycles that it should be uh lasting so to sum up a bit I’ve shown you how commercial strategies can put in jeopard the life of the battery if we are not choosing wisely or accurately uh the parameters of the charge also I show you that there is uh
Uh there are degradation phenomena that can be triggered or can be accelera during the charge of the battery and with a regular uh commercial uh charger we cannot see this or we cannot control this so that is why we developed a health aware fast charging strategy able
To uh contact or to limit the uh occurrence of this kind of phenomena and I would like uh to thank you for being here in this uh to conclude this presentation but also to thank inovis and also the Brussels region that have supported financially all of this research along the last three
Years as uh I had at the beginning uh we listen to to representative of B region the main problem is not to solve B’s problem is also to solve the problems that we are generating outside and overseas so thanks a lot and okay thank you very much Alejandro
And now um for questions is there any questions from the audience yeah you have one yeah thank you very much for the nice talk um a question from the material scientists so you showed uh that the the performance is dependent on many things of course including the anodes cathodes and
Electrolytes H and the temperature and the chosen chemistry in between and Etc there are so many researches are also going on about the um polymer electrolytes and polymer sorry polymer um uh polymeric cathodes for example so how do you think we are um uh kind of evolving to um um material
Dependent uh cathod uh selection in terms of going away from the metal dependent uh battery Cod but in that case you have first of all uh as um Michelle said at the beginning my expertise Al is on chemical engineering not from the materials uh science perspective um the selection of material that you
Have uh in the batteries so within a cell for instance depends on the different of potential okay equilibria to create that voltage on the power that you want to deliver obviously you can improve the materials you will have always a let’s say a a metallic compound
There that you have the lithium is a metallic compound let’s say also you have now uh sodium uh batteries or they are developing sodium batteries but at the end the the the definition of poly IC materials is also in terms of safety MH and how to contact from a physical
Viewpoint the degradation mechanism that I just describe because sometimes it’s uh it’s it’s an effort that should be uh jointly done by the designers of the battery by the electrochemist but also by us that we are operating the batteries so to be honest uh because because it’s not only always
Uh how efficient to use the batteries in terms of sustainability but how we design the batteries in terms of sustainability which sustainable or which more sustainable materials we should use to design the uh batteries U as a material scientist I was always approaching from both ways because
Metals are always the limitation for the batteries as well that’s why I wanted to um draw the attention from our material view point I can I can question you back what is the best design that will ensure a good uh recycling process because so far we have
A few recycling processes not so well developed I would say or well implemented so the batteries that are wasted are batteries that are lost so as you are the expert on the field how can we improve the design to improve the after life let’s say for example there
Was a a remark in uh yesterday’s talks I think there was six hours of time spent to kind of disassemble the battery mhm am I right I do I remember right so um for One battery six hours I mean if you imagine the time and effort and expertise to put
On starting to recycle the batteries is is huge so sometimes I think chemists uh could do much better um I mean effort and put some better uh products to switch from material to polymer that’s what I is also kind of a research U of course uh topic for myself but I just would
Like to raise the concern here also uh to make a discussion points in fact lithium ion batteries are quite young they have less than 30 years in the market so and and the first development is back in the 80s so that doesn’t mean that scientists are not trying to do I
Mean many of us we are scientist so we are not trying to change the the perspective and the usage of certain uh uh materials the main problem is how how Keen the companies are are are are to to implement such uh developments because uh changing a manufacturing line
Of batteries is not something that cost penuts so that is a main problem in and and to us as users of batteries something that I didn’t point out is that at battery cost a few Euros I mean those batteries that you saw there are seven five for
The user of a drone that is the lowest cost for the user of a robot that is the lowest cost cost so for most of the companies that use batteries batteries are legible it’s a cost that they can afford and that is why we have to improve what I mean the perspective of
The company that is using the to their own and not the the how to improve the batteries uh physically I would say because that that change will take uh I would say a couple of decades to change manufacturing lines okay and we have uh two questions from
Slido I think the first one might be very quick um so Jo just to clarify is there a way to consume less energy by charging batteries in a different way or speed charging energy in which sense uh if I mean the energy stored within the battery will be the same all the time
The way that we are changing is uh how we are charging the battery if we are using 10 amps for 10 hours or if you are you are using 300 amps in one hour the the amount of power that we are feeding is exactly the
Same uh I mean the efficiency uh that we have to consider but uh at the end the amount of energy the amount of power that we are feeding to the battery is exactly the same and the amount of power that the battery will take I mean the
The Drone or the will take from the battery will be exactly the same so we are not changing the consump consumption of uh of the power of the current what we can do is using a stationary B um batteries or storing facilities based on batteries to uh let’s say to store
Energy coming from the wind from the Sun and so on that is different but if we are seeing it from the perspective of the Drone is just the same amount of energy in different time frames okay thanks and uh last question there is a lot of research on this topic
But do consumers know about these issues EG uh and of Life at under 80% and how could consumers be educated better on this I don’t know if that falls into your uh expertise but maybe you have a opinion on that um I hope that at least for the attendees here present um this
This my talk helps a to understand the batteries and to help them um knowing how to charge the batteries is it’s true that with some devices some charges we are not able to uh control or to regulate how the charge is done or what are the limits that we are imposing
For sure but uh is it’s really nice to to to create awareness about how batteries are being charged and how batteries can impact also the the the efficiency of a device because that that should be done from the once again the industry view point and we are uh just consumers of the
Products that come and the algorithms uh the energy management algorithms that come within our our our devices so but yeah it could be a nice way okay thanks a lot I give it a round of applause again for [Applause] Alejandro and then I will introduce our
Next speaker it is um abishek Gupta from uh coming from and uh we know each other for the last four years I think um four and a half okay thanks and uh yeah another mechanical engineer and we got caught up in robotics yeah I will move here a bit
More um and uh his research or his interest is also on robotics but um he’s taking a bit uh a step back and trying to look at the the whole system like what is surrounding the robot what is um around the robot the what is the infrastructure what do we need to plan
For the robot for the service to implement the robot itself and in that way trying to think of benefits and advantages to really um yeah have this knowledge and improve the service improve the sustainability so um give it up for Abby and I will just manage to
Have your presentation on the screen and you can um install [Applause] yourself okay thank you Michelle for the I hope I hope I’m right on the mic right yeah okay good thank you Michelle for the quick and uh concise introduction good afternoon or maybe still good morning uh
Everybody in the next 14 and a half minutes my aim is to present you how it’s possible to rethink how we design service robots not as standalone products but maybe as part of service systems so that as the um as the title of the presentation says for responsible operation
Management every research work has a motivation well most of them have and I hope every and our motivation when I say our I mean I speak for Michelle for myself and the department at tuberin our motivation is if that somebody thinks by just incorporating service robots into conventional Services everything is
Going to change and everything is going to be perfect optimized and efficient I think we are wrong and that’s why the motivation is humans not robots are responsible agents robots are tools just designed to achieve human goals you wouldn’t believe me just because I’m saying it but there are many
Other organizations such as the i e Automation and um robotics um society which believe in this and luckily this quote is not from for me coincidentally I’m in Belgium the home of uh European commission this quote comes directly from European commission Horizon the EU research and from they published it in
The Innovation magazine and they also believe that robots are just tools uh designed to achieve human goals now I don’t want to give a lecture here what robots are etc etc just a small important information for the next slides robots can be simp very simply defined as uh mechatronic wait just a
Second so yeah robots can be very simply defined as mechatronics products combination of feedback system sensors actuators Motors Etc and then they uh transform into intelligent mechatronic product cyber physical systems and smart products now smart products are basically when such robots are combined with uh internet Based Services or smart
Services michell he gave a really nice presentation on the product development uh system and the product life cycle I wouldn’t dive deep into it but just to understand a very simple way to understand the product life cycle is the holistic product life cycle starts from uh strategic planning uh goes to product
Development where basically the prototypes are developed realization and production where the serial production takes place and then product use and end of life now the problem is I said something about internet Based Services my mind says inter internet Based Services should also be started to be de uh developed in the initial uh
Strategic planning phase sadly that’s not the case they’re actually starting they actually started to be developed at the end of the realization of production phase now just to the mathematics 1 plus one if I’m developing a service robot which is using a service to fulfill certain kinds of goal and then I realize
In realization or production phase ah I need to change the robot now I need to reconfigure the robot maybe it’s just a prototype but the amount of materials that have been used spent and resources that have been invested they all might just go to waste so the idea is to come
Up with a methodology where the life cycles of the product and the services are coexisting or at least uh they run simultaneously to go ahead and to make you understand and make you believe in the methodology it’s easier for me at least to come up uh to explain
Everything with a use case and the use case again uh back and forth to Michelle and ABI uh is the marble the functional prototype uh from tubin which we developed uh for autonomously emptying the litter bins that is why the full form is mobile aut autonomous robot for
Litter EMP uh litter emptying uh as you can see it has a platform several sensors uh a robot arm to open the litter bin collect the garbage and compress the garbage the process is basically more or less very simple um starts with autonomous drive a to b b
Being the litter bin opens the litter bin uh the garbage gets emptied U uh it’s transferred to the compressing machine gets compressed and then needs to be extracted why because we couldn’t achieve uh that the robot can empty all the 26,000 lit bins
In dust bin at in Berlin at one go uh so after four uh emps it needs to uh transfer the garbage now seems pretty straightforward and easy but uh luckily we ourselves um the developing team and the potential end user we identified some kind of shortcomings for example limited speed
Has to drive on The Pedestrian path can’t drive more than 5 5 km per hour Dimensions constraints expensive robot arm and then also time intensive operation um Sophia gave a very nice speech on how the time um time consuming programs can also have an impact on the energy consumption
Then uh limited garbage storing capacity and the need for garbage transfer now to overcome these things the question can be for optimized usage should we reconsider the design I would say yes why not but uh I would I would hardly s uh and strongly suggest not just to reconsider the
Design but reconsider the design approach as robot infrastructure and service based design approach approach to just to give you a short um view uh or impulse what I exactly mean imagine a smart product combined with a service then we are speaking uh in the field of smart product Service Solutions helps to
Meet indidual goals combine it with um intelligent systems and maybe integrate uh when automatically the needs of stakeholders can be changed and other product Service Solutions we are talking about the system level from smart product service systems can optimize overall results move it a little bit
More ahead it uh and um combine all the other Associated smart PSS or product service systems we are we are speaking from an ecosystem they speak of to have the greatest impact on all on total system I’m of another opinion uh they cannot achieve the greatest impact but possibly
One of the best impacts and that’s why the methodology that we developed is called claps methodology coexisting life cycle initiation and management for autonomous robot product infrastructure service I had another name in my mind it was slaps uh synchronized life cycle and Etc thankfully I decid decided the name
Together with my professor and he said uh yeah Mr Gupta let’s work to get claps in the end not slaps so that’s the name claps methodology and over here as you can see I hope the pointer is visible the life cycles uh of the product as well as the service they have been
Synchronized since the beginning of the modified uh strategic planning phase which I showed you in the previous slides we start with a socio technical analysis that is just to see what exactly is missing in the current uh actual service system uh being performed by humans so that these um shortcomings
Must not be faced by the by the service robots we move on to generate the requirements uh and if need be uh also incorporate additional assistant infrastructure and uh carry on with the digital prototypes based on a specific operation management system now operation management system for me can
Mean everything and nothing but uh and for you guys uh from different fields it can be different goal for some for somebody it can be the ecological perspective somebody economic persp perspective and uh and for somebody social a dream world will be when everything combines together for one researcher um unfortunately it’s not
Possible and then if you’re satisfied that okay everything is more or less fine we we work again on the requirements and provide the specific requirements in the product and development phase where the product the assistance infrastructure and the service they are developed in a combined way so that if inter inter inter related
Changes are needed they can be carried on simultaneously and then obviously we go to the production phase uh the usage phase and the end of life now don’t be scared of this um no questions will be asked it’s part of my doctor thesis but every phase need a method every phase
Needs a method I can’t just expect um without methods to work ahead but these methods are more or less uh should be in are defined in such a way that every developer can choose their own methods um it’s more or less like uh use case friendly um domain expertise friendly
Resources friendly um and you can choose your own uh methods but one of the most important thing over here is sp or um as I like to call Standard physical prototype development it’s like although the idea is to use this methodology for new service products or new service
Robots if luckily there are existing smart uh existing standard physical prototypes for example in AR the Prototype is already there why not use the results and the experience gained from the functionality tests uh for for providing uh the new service system that we are aiming to provide and
If need be for example in this case uh the new the requirement of additional assistant infrastructure and then we focus on the operation management system to analyze the whole service service process and see if we are working in the right or the wrong direction for marble
It was uh it was pretty clear we need to overcome the problems to infrastructure assistance infrastructure one a mother ship helps to transport the robots and um take away the garbage after the marble has reached its maximum capacity and then the second one maybe a filling level predictor uh from the litter bin
So that we know how much full the lit bins are uh and not to just get lost with overloaded marbles or over floating streets with uh garbage when marble is not able to empty it to make you a little bit believe more we simulated
This all in a real world um uh Park um example it’s uh those of you don’t know it’s in Berlin James Simon bju Park a really beautiful park with 51 lit bins uh scattered overall uh the problem is even the municipality doesn’t know where exactly the Lans are
And just imagine telling a robot hey um congratulations you are automat go and the L bin uh the results um let’s not discuss over here I’m still getting paid uh and then we develop a operation management system as Michelle said basically my background is a mechanical engineer I’m still striving in this uh
Physical space cyber space but luckily I have gained some knowledge uh how to combine all of them and especially the stakeholders and combine them in route and operation Management Service that is why the focus for the next four and a half minutes that might have been left is the operation management system you
Take it into a little bit more different uh and um in-depth level and then you can have a information flow and the information flow is again physical layer platform layer and app player and then my concern is how to manage the operation Management in uh routing you start with different
Parameters you have different algorithms I wouldn’t say one algorithm is the best never possible then you have the operation goals uh and then you see how how it it functions for us the parameters were pretty much straightforward energy consumption for marble from from mother ship from litter
Bin uh and the filling level with respect to their positions for the conventional one we assume that every little bin is 50% full and for the smart letter bin one it was like um we give them randomly the number between 0 100% in 10% steps we come to the results with the
Help of root planning algorithms and we see over here the first one with u marble without any Mother Ship we are way higher uh in time consumption little bit better than the diesel uh vehicle that the municipality is currently using but I have been told electricity or
Electric is the future so we also compared with the electric vehicles and in those aspects we are way bad uh or yeah we are bad we are way bad uh uh in the energy consumption and also uh in the time in the time consumption we tried something different
Like with the fleet getting a little bit better tried with the smart dust bins getting more better and then comes a hurdle and it is in Berlin vandalism you can’t just convert all the 26,000 lit beans and one go it will cost Millions nobody wants it but I like a
Very I like a very specific proverb in German language it’s like K Zak Aon it means simply if one hurdle comes it doesn’t mean the the the jewels are falling from the crown and you can’t use the crown we took it as a another opportunity to add feather in the cap
And what we did was we took the device that we developed for measuring the filling level of the Dust bin modified it a little bit when two months in December and January um in the park gathered the data at 7:00 before the workers of the municipality were there used it for predicting the
Filling levels uh in the future uh over different classified uh classes such as days weekend special events we also worked on Christmas and New Year by the way um and then we realized 85% of accuracy if we skip the dust bins or the litter bins which are less than
25% we might be able to save more energy and uh achieve it quickly and this is how we get a little bit better but still not the so close to the electric vehicle variant of the municipality so we redesigned we so we almost done Miss so we redesigned we
Redesigned we took away the robot arm and we uh we redesigned the compressor in such a way that the capacity has been increased and luckily we got a little bit more closer to the energy consumption and the operation time of the as compared to the electric vehicle now the moment of truth
These are the two benchmarks for us uh Diesel and electric from the municipality we started over here we landed somewhere over here Solutions with least possible operational time then we landed somewhere over here Solutions with least possible energy consumption and now everybody that is present over here and online please took
Please take just two seconds I didn’t promise you that I’ll provide you with the minimum solution or with an Optimum solution I said responsible operation management so with the help of different solutions we are trying to get there where we want to without trying to actually start developing things physically but first
Let’s stay stick in the in the simulative world and when we are confident that the results might be so good let’s start in the uh in the Prototype actual development and then later on think about this serial production and in the end it’s more or less uh having a developed system
Instead of Standalone product as a product and service system or smart product and service system as you want to say and combining it with the other product service systems that was my time there is a reason why I have my picture I’m standing alone over
Here and my goal over here was also to see and to win some people who can actually help me criticize my own methodology and make me show what the plus points are you all are developers in the field of Robotics you all are users and maybe as sopia said yesterday
Also many people said it’s time to collaborate and it’s time to see how one thing affects the others and um maybe it might be a chance to sit together one once again for one and a half hour and see how things good or bad have done and
So far yeah that was my time thank you very much and last thing Michelle last thing a big thanks to all of you for organizing the event everything is more than perfect yeah thank you okay thank you very much um I will postpone questions from the audience I
Think uh for the Break um I will just go through the slido questions we have to run a bit with the time so we have enough time for the discussion afterwards so uh first one for you is um thank you for the talk I enjoyed it my
Question is about the motivation as to why did you need a robot to empd recycling bins really nice question question um I didn’t feel the need um I didn’t feel the need that we need a robot but there was a need from the municipality they’re having Workforce problems if you
Come to Berlin in Winter where there is snow unfortunately the people that are working on emptying the litter bins they don’t have the time to empty the litter bins they need some kind of help and providing a robot was one possible solution we are still not there we’re
Still we’re still working but we don’t think that robot is the only solution if need be if need be in future we are there to help to provide a responsible solution thanks um the second question is you considered environmental and social aspects what about the economics
The robot is um expensive does your root planning help to show the business case solvings um it’s an amazing question but I’ll divide it into three parts and very quickly environmental and social aspects yes we did actually I didn’t speak anything about social so thank you still for the compliment
Economics Yes Energy in the end is money uh if I’m spending energy if I’m having energy consumption I need batteries I need to charge them it’s in the end economics but if I uh I don’t want to go back quickly to the slides but there’s a
Part where it says analysis and there be it the life cycle analysis for the ecological aspect or life cycle cost analysis for the economic aspect I said you can choose your own models uh you can choose your own own tools to evaluate and I’ll be more than happy if
Somebody can merge together and say how it’s in the how effective is it in in economics perspective or cost uh cost effective perspective second one uh show the business case of savings um I really don’t uh like planning helped to show the business case or saving Savings in terms of money yes we
We we will be saving money if we are saving time and energy but business cases um small information product service systems or smart product service systems are just a fancy name for business business cases and it it is a kind of a business case uh but to
Develop more depth into business cases I might have to change my background from mechanical to software to manager background it might take some time and maybe in few years I can answer the question properly thank you thank you Abby insightful and is always very entertaining as well thanks a [Applause]
Lot okay so the third speaker on this session will be um Chris Fon please come up to the stage he’s a artist from Belgium and has over the last years I think uh you can say decades um um had a lot of uh projects between installation performances and he’s is uh particularly
Intrigued with uh machines and to uh include machines in the performances as I’m as I’m right and he will show us a few of his uh works and then right after his talk we will jump into the discussion with the three speakers from this panel so I will just plug it in and
You can start ahead thanks yes so um thank you um yes here the mic there yeah so yeah so um uh I would love to uh start with um to explain you a little bit how the visual arts uh work so poor artists uh no money
So that means uh all the machines we invent are prototypes to get from prototype to Something Real most of the time there’s no money nor time so that’s not happening but uh curators always fun they absolutely need machines that work for three months or plus um machines that have to be
Interesting uh stable uh all of this uh audience uh children dogs um these machines have to survive all of it um and everybody needs something Innovative uh Etc so the list of what is needed and what is provided there’s nothing provided um you come in in the morning
Because there’s no budget and in the evening your thing has to work um there’s another plus uh in those three months um even a projector even just a normal projection against the wall is difficult because uh every morning the thing has to be installed I mean it has
To put on um if you have a a electric breakdown everything has to restart but itself because you can actually not demand of the with all the respect for the people who work in Musea um that they work that that that they can handle
And to to put it again on let alone if your machine is even more difficult uh to really set it up I will even go further when your machine breaks down and uh the exhibition is in Hong Kong uh there it is so if you go to an exhib
Where there is multimedia please go the first week because the second week there will be nothing to nothing anymore uh and I really mean it it’s not even a joke um so uh so then I I from there on um I thought okay so I have to come with
With and I did it for years and years this kind of high-tech interactive maxmsp or whatever machinery and indeed mine was also failing always um so so so then I thought okay let’s go uh so I need machines that or or objects uh that are completely self-sufficient
Completely uh they even I don’t want even to use the the electricity of the of the the um Museum because that can fail also so from there on okay my next step was why not having solar panels and as a as a solar panels as a um uh as my
Own as my own battery or as my own production of uh of of electricity um and then my next step was how to go as low Tech as possible because everything else if that fails uh is pro is problematic and to get completely something stable why not go in DAV Vinci
Mode and really go complete low Tech as much as possible um and from then on I uh The Next Step was that I was really intrigued by how the electricity of uh day and night and Everything Changes by uh with nature together um it follows
The the the even if there’s a cloud in front of the sun you have less energy so so then I thought let’s try to give form to this um uh to to the fluctuations um so there was my first one it’s a uh Rock uh 650 kilos pure
Marble um that turns around uh of course completely on one point there it is um on one point turning around and uh when there’s a lot of sun it turns fast when there’s no sun of course it stops uh and it’s even as sensitive for the moment
When there’s a cloud when there’s a cloud in front of the sun it goes a bit slow and so it’s completely working with uh with the elements instead of against um and the fun part I thought was that the engineer was working with he during this production he begged me almost like
Shall we not use a battery or can I please plug in or uh just something to control the DC engine uh because it’s so difficult to get it oneon-one with the with nature instead of against with the elements um so that was also quite intriguing uh to so this is the the
Explanation um it was quite intriguing for me to to start thinking about um why doing exercises or or thinking to develop uh machines that are really with the thing and canceling out the battery uh to go low Tech as low te as possible and while following all these uh the the beautiful speeches There because the everything has to be in Perfect Balance not to uh have too much um to demand too much energy uh from the solar panels or the the the electro engines or the DC engines um so all if everything is imbalanced those were three uh Motors that were also uh
Completely working um on by the their themselves so I come into the to the museum I put the thing down and it works for months because it just uh uh uh is mechanical and not uh so it skips quite a lot of of uh uh
Problems um and so this is one of the one of the things that I was really uh intriguing and it’s also more a question for everybody here um is is it possible without a battery also for example I see these these wonderful exoskeletons or um uh or these uh these
These well yeah these hands and and and stuff A friend of mine who was born without a without a hand and uh he just quite cute when he was a young when he was a child he always had a fake arm and with this arm he could he lived with it
And he could do everything he was just if he says I I call the big fish he says it’s this big um he does everything with it and then uh I think 10 years ago or so he had a more complex uh machine that could also do other movements and it was
Always breaking down there were always problems it was more expensive than ever um uh he was while we are talking suddenly the ant opens because incidentally he uh he touched the sensor um and after a while he said you know what uh it uh I’m going to I’m
Going to throw this thing away I’m I’m much more um I when I’m just having a a stupid stick instead of something uh robotized um and so it’s it’s uh one of the the another example um and that goes way too far I know but I will uh I take the the
Liberty here um at one point I was in Cuba and there was a a huge concert um and it was not even five minutes when the electricity broke down nothing anymore um and then suddenly people searched all instruments themselves and it was a great night because uh everybody was just partying along
Without without the the the the concert what I want to say is that the question is also uh the question here and that was also I I love the the question there um does it need to be a robot that collects the the garbage and I think
That’s one of the the first questions we should pose um and on a design level uh I I’m I’m I’m I’m quite convinced that actually [Applause] That uh that it’s a it’s an impossibility etc etc uh the world doesn’t run uh low Tech anymore um um sure as long as as it is completely stable and uh uh that can survive uh for in my sense the in my case the a three
Months of Museum uh uh without any any mistake yes of course yeah so uh there are these uh toys that uh sort of can rotate forever just under um light pressure uh that one side is shiny the other side is dark and then they can start like turning around
Forever yeah beautiful have you explored that uh working on maybe larger size objects than the tiny things that we can purchase yeah I I I know them but it’s almost impossible they’re so beautifully in glass uh in vacuum most of the time uh because there can not be a single
Little bit of friction because the one is black and one is white and so it’s really just uh it is too delicate to to uh uh make in a big version of it but I did yes I have another question which is uh or more like a a comment uh where we
Where we find like an interesting fight between complexity or Simplicity is uh when uh we work on uh self uh self Of course robots that construct robots are uh I don’t know what to answer in this one uh I don’t know yeah thank you okay uh so this question is from Patrick uh from climate robotics network do the speakers Point um just now the use of soft robotics bioinspired
Robotics and morphology enables for more efficient energy consumption and nature powered Locomotion for example he post a link um also I’m a big fan no I’m a big of minimalist approaches big fan sorry a big fan of minimalist approaches to designing and building robotics so we can check out the link later all
Right thank you I think we will get to that Soft Robotics will be a topic of the third session so we will get to that patient Patrick and I think we’ll just go into the discussion panel now um can maybe Alejandro and ABI join us on the stage
Um I will move the into the frame okay uh I hope we have it everything on screen um no we don’t need no slide okay so um I can start to introduce the discussion we just heard uh three different views on batteries the use of batteries or um not using batteries and
We heard in the first session of course all the impacts that can come from batteries as the environmental the societal impacts and um so in general there’s a big impact coming from the batteries and of course we would be better off if we had no batteries but is this possible um uh
That’s something we will want to discuss here in the panel and I will just give the mic to um you and start with the question why don’t we get rid of the batteries who wants to maybe start I think a lot of you will have uh the
Thought in mind that’s not possible but we’re just here to to explore and um see what alternative approaches there might be uh yeah in my in these last in these last works I made um the the biggest I I also see there’s a lot of challenges uh on a on a
On a on a development side uh that somehow for example just DC engines are now also uh directly controlled by uh processors etc etc um so we we at the moment we’re even thinking about uh constructing our own um DC engines that are able to really have also that can
Take the the hits of fluctuations of the sun Etc um so so it’s I think there’s there’s a lot of challenges uh that that would uh um that are necessary if we would if we think about doing things without uh without batteries um uh so yeah a lot of mechanical um goals are
Are are there that could be interesting um and challenging oops yep just go ahead thanks um I think that we cannot get rate of batteries I mean they have been with us for 150 years with the the case of the lead acid batteries but we should
Or we would um I mean I think that we should use them in the rightful amount it’s not that we should power every device with the battery we have cables we have connections cre show that actually you can have a structure that depends on cables and PV panel
So that doesn’t mean that everything has to be powered uh with batteries uh we have also to cultivate patience because most of the things that uh we have shown today is about uh being patient and waiting for for the the charging time of a device or waiting to see the the the
Work of creas moving in the right speed due to the Sun so everything is about patience and also about uh picking the the the lers of the the gabbage is like patience about when it will be done no it’s that it should be done right now so it’s about efficiency but also about
Patience I would say thank you little bit confusing now one wants to get rid of the battery one doesn’t want to get I don’t know whose side to take but I would say um I completely agree if there’s a possible solution to work around without batteries it’ll be the best possible
Thing but uh we would we should also a little bit uh think like first we wanted now we want to get rid of the know of the diesel vehicles then sometimes we want to get rid of the batteries etc etc we really need to see how to achieve the
Solutions because uh when we speak of sustainability it might be possible over here uh in Brussels to get rid of the batteries or in maybe in Germany we also need to think how it’s possible all over the world and and then see what kind of impact and how the technology uh goes
Further but yes uh if I can live in a world without batteries uh I think I’ll be a very happy person because mostly I don’t even understand how they work so but uh I know that their impact is not that good um on our system yeah okay
Thank you um the discussion is also open to the to the audience and uh we received the question on slido so I will read the question out loud um asking whether we need a robot uh at all seems to be a theme of almost all presentations today I agree but how do
We convince roboticists and robot companies so anyone of you have an idea as maybe um presenting a robot assist how to convince us of using less and going more into the direction of sufficiency can you repeat the question yes uh it’s more about how do we convince roboticists or robot companies
To use less robots yeah do you yes am I good yeah yes um I think if we want to invent stuff um because that’s what we love to do um uh I think there’s quite a lot of challenges it’s just a mindset to
Try to uh as a as a first step to to make something sustainable and from then uh uh you go on and I think the the mechanical challenges are as big if you go for the high-tech stuff in IE or in the low Tech stuff um so the the the the
The the fun of making things that’s why what roboticist probably why they do things uh because it’s fun to do uh but the fun is also to do with low Tech um so in that sense the the we just have to convince them that for me in my case to
Go low Tech is also very very chenging thanks uh do you have another opinion on that maybe I completely agree okay okay I see um I have a question maybe um it’s a bit about uh back to the battery theme um do you think taking away the battery would make a robot more
Autonomous or less so and why uh you can also pass if you want to more uh absolutely because it um it uh it’s it’s uh the mechanical parts are are are so reliable that it needs less human beings around to uh survive uh so so much more sustainable and less less
Uh less human beings that are needed yeah well as a as K said um it’s matter of autonomy it’s uh what you want to do if if you are attached to a cable or if you are or if the robot or drone can go independently I mean if you
Can go around without any issue um it’s also true that you can profit from uh wi energy solar energy and so on but uh when there are flu fluctations of that kind of energy how can you do it so it’s about uh autonomy and uh I would say
What is the propose of of of the technology know just a small addition to it um if anything is happening on its own irrespective of the fact how and if the mechanical components with help of maybe solar energy or etc etc working on their own I think we are increasing the level
Of autonomy over there that’s my personal opinion okay thanks um another question I I thought of uh in this topic is uh maybe going back to bit more technical um standpoint what trade-offs would you expect from trying to work around the battery so taking away a battery what technical try trade-offs would you
Expect you can do okay um I would say um it brings me back to um um my my my topic or my focus if you’re getting rid of rid of batteries then you need to think how you’re going to charge and then coming back again it’s an additional additional component that you
Need to provide uh the service be it from robots or be from any other thing and it’s the same no wonder electric vehicles have been for so long uh amongst us but still we are afraid to use them because lack of charging infrastructure range anxiety and no wonder when robots would be having
Batteries um we might be afraid hey uh how exactly is the robot going to charge itself will there be enough sun or wind and I I think um it brings me back to the same conclusion that we have to see uh how the whole system works without
That to give a concrete answer might be foolish of me right now it’s difficult to add on top of that um how to replace the battery will be about how to replace the the the power system of the device if we have the connectivity the electrification uh if we can find
Another source of uh energy to to power the device now um thinking about uh the the the that you are developing uh if you’re not using batteries you are using fossil fuels and we are coming back to once again to the to the I mean if you’re not using
Obviously solar energy and wind energy uh some sort if you are not using batteries you are using fuel fossil fuels and so we are coming back to the same to the auding of the problem about uh waste energy and um carbon um the carbonation and these kind
Of topics so what is the solution if we are not using uh electricity if we if if we don’t use batteries from your Viewpoint what kind of energies we can use I think um irrespective of the fact coming back to fossil fuels batteries they get charged right the it’s not necessary
That the electricity that we are using is 100% uh emissions free for ex example uh I know in Germany we call it like Str mix uh stro mix and uh the the amount of energy that we are using just to charge the battery even if it’s electric we are
Having a very high amount of CO2 emissions so the only the only idea which I really like from his presentation is getting rid of batteries is really nice because then we can just switch to um natural or renewable energy providing sources solar energy wind energy Etc it’s just we have
To see how to get uh get the get the whole robot or system running that’s my opinion because fossil fuels irrespective even batteries they are not uh 100% emissions free it’s just that we are used to selling this uh or we are used to Green wash that hey
Electric CO2 emissions free do whatever you want yeah I I’m sorry I I I don’t agree on that point it’s not about the um I mean it’s I’m not saying that all electricity is uh purely uh produced um what they said is that it’s from the Viewpoint of
Autonomy if you don’t have a battery to store electricity you have to depend on a of uh maybe hyrogen or a that can be an option or uh fossil fuels to move your vehicle so it’s from a the Viewpoint of autonomy not from the Viewpoint of
Generation of the of the power of the electricity that they were consuming what is the option uh if we don’t install for instance on roof full of PB panels on your device how can we power it during the operation some remarks from the audience I think it Al I think it also
Depends on the how much Fidelity we want that the system is up and running for example uh in Belgium there was at a certain moment iot devices okay it’s not a robot but a kind of measurement devices in your garden to to know the how much humidity is in the
Ground because they want to see the new climate with long periods of dry period Peri and wet periods and that were all battery power devices because they had to send it to the cloud so it was 10,000 of batteries being used then the question is maybe they can go to solar
Panels or they were also thinking for the temperature difference between ground and uh the upper part to use that as a source of electricity and then to agree that not always the device is switched on because it doesn’t matter maybe that in your measurement you don’t have all measurement points because the
Return is that you can save a lot of batteries in it the same for my grass mowing robots imagine that solar panels will be efficient enough yeah if there is not a lot of sun probably my grass doesn’t also grow a lot so it’s okay that for a few days it doesn’t work
Anyway when it rains it shouldn’t drive because it’s too muddy than uh the garden so of course when it’s Som out health related maybe you want of course redundancy and at always so I think depending on the application but we need to get maybe rid of the need that always
All robots should work uh and to see what is the tradeoff between availability of energy the resources you want to put in and the up time of the robot thank you very much and another one I think you’re on the back from Steven uh thank you very much and thank
You for um three great presentations um and perhaps my question or idea is is consistent with everything you shared it seems to me that conceptually what we’re trying to do is to create a new technology related to Robotics and exoskeletons and applied industrially um in the household in the
Community but this this discussion right now about batteries is limiting the the part of our conception of these this new technology on the current technology we have available related to energy because I think it’s going to be impossible to come up with robotics or exoskeletons that don’t rely on energy in some
Capacity to do their job so should we be collaborating with people who are working on new Energy Technologies to to to develop robotics that will be sustainable over the next several decades Century um for example is is there new Battery Technology on the horizon that will all allow a battery to
Be charged a greater capacity for a longer period of time are there ways to get energy from our environment that are more efficient more sustainable than uh say a a a solar charge but I I I don’t know I’m just asking that question yeah thanks for the question I’ll just bring back the
Mic wants to start uh or maybe I can try to uh I have a you uh similar question I think that goes into Direction which I wanted to post at the end and maybe we can uh conclude the discussion with with that question um what Innovations or advancment advancements uh advancements
Yeah sorry could change the approach of how we perceive robots and their structural composition I think it goes a bit into that direction maybe you can uh elaborate on that for or in a few sentences and then we’ll conclude thanks um do I think to answer your
Question first um I have no idea uh which technology what kind of batteries should we do if you want to move in this in this direction the only answer from my side as a very honest answer will be let’s do our own work on our own then we
Don’t need robots we don’t need machines happy uh to answer your question um if I understood it correct it’s more or less like the changing the constraints uh the dimensional constraints right yeah What technological innovations could maybe change the way we we see for instance I don’t know if there was suddenly Fusion
Energy and we have unlimited energy would we change our perception or construction of robots I would just say um if any new innovation comes um um just uh use it uh responsibly because even it won’t last in then maybe not us but in 10 decades some people might be sitting again in
Modified room version of this room and thinking how to move ahead because everything is uh as she as Sophia also said it’s limited yes uh humans are capable enough of finding new things but the question real question is till when that’s my conclusion um to elaborate uh about the
Question um the new trends uh on batteries are mostly on selfhealing batteries uh batteries that are are generating themselves during the the operation to avoid this kind of degradation that they suffer also we we know that there are uh techniques based on on including sens within the
Batteries to know what happens also for the thermal management for the degradation management and so on and from the other from another perspective in that case also we can use a stationary variable um battery sorry to to be a as buffers in the production of hydrogen I
I know some uh works that use uh wind power to um to Esto a bit within a battery and after to generate hydrogen uh and that way they can store energy in a in a in a higher capacity I would say so we have Trends uh uh about how to
Improve the batteries and how to use them to generate other uh storage of of energy from the the question again uh about rots I’m not an expert on the field of Robotics to to elaborate on that so I think that he uh he deliver yeah sure thanks maybe Chris
With our last words last words for the for the Forum the last the last words goodbye everybody uh finally uh relieved um um no I got a I got a little um um so I’m a I’m a hobby wine maker and um I’m in my second year
Um biodynamic and natural stuff um and it’s it’s I was just I’m just new in this uh wow I’m almost in I’m almost dead maybe the battery is going maybe the battery is gone so maybe I just have to speak louder um uh and so so the the the crazy thing
About biodynamic stuff is that the whole relationship while producing this kind of wine is only it can only happen on on perac culture um and so so I was so the last two years I’m so astonished by the wine makers who who rely who have trust
On their plants that they will uh uh if they it’s dry so the The Roots have to go deeper down and so you cannot Add Water etc etc so the the whole relationship with uh with with uh with the planet is for them completely different and uh and what I want to say
Is that also there uh but they’re not scared of any uh Innovation uh they would love to have little robots that that help them out uh but in a complete different way uh so the Innovation is still possible but uh just as a basic thinking uh it’s it’s completely
Different thanks okay yeah I think I understood as um being not so techno optimistic but uh if it comes if we have new Innovations we can use them but we shouldn’t rely and say uh the the technology of the future will save us so we don’t have to worry about it now I
Guess that’s what you were saying okay thanks a a lot um we will conclude this discussion uh thanks again for the three speakers and we’ll move on to lunch um where you just had coffee there will be now sandwiches and also I want to um point out that we have uh the other
Artist in our Bri Fellowship has um exhibited or is exhibiting a few of his foldings and you can maybe go over and look at it during lunchtime thanks a lot e e e e e e e e e e e e e e e e e e e e e e e e
E e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e
E e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e
E e e e e e e e e e e e e e e e e e e e e e e e e e hello everyone uh welcome back to the third session for today’s um Bri uh Forum um we are going to first welcome
Alex uh from um National nitorum in in from uh Edinburg he is working on Soft Robotics especially for um extreme environments um uh in their applications so his talk is very appealing today grow your own robot um so the the floor is yours thanks hello um yes so I’m Alex page from
National habitarium um i’ like to make my talks as inter with as possible I usually give out dog toy squeakers so instead of raising a hand people just squeak at me um I don’t have enough for all of you so if you want to just scream
At me Heckle me in any way shape or form please feel free it’s after lunch I want energy don’t you falling asleep um yeah so gr and robot I’ve been here for a few days I haven’t had enough time to grab my robot yet but I’ve got a single slide
Where we can talk about a bit of that okay so from ration robotarium the hell is that it is this Lely building it’s 2 years old um this is it when it was fresh it no longer looks like this it’s now a nice Rat’s Nest of technology
Which is as all robot Labs should be um yeah come find us there’ll be a link at the end where you can scan a QR code and find out some of the work that we do but essentially we sit between Academia and Industry and we kind of solve problems
For industry with academic um Solutions um who so I’m Alex I work in Soft Robotics my PhD was in reflex response robotics where I essentially tries to take the brain and electronics out of robots because he needs that and give something that could instead respond to a dynamic environment so
Things like a robot where you could punch it and it would grab you without a brain um currently working on tendon driven robots for subc um pipe inspection also responsible for a rubit cube that rolls around the floor by inflating colorful tumors on its body you’re
Welcome um but today I’m going to talk a little bit about a recycling process that I created for silicon which is one of the main materials that we use in Soft Robotics so as with all stics talks I’m going to start with a question and a fish what’s wrong with this fish hands
Please I mean if you know the answer sh okay yeah it’s a dead fish I can’t believe you didn’t see it straight away so yeah you got this dead fish um and because of the body of a dead fish it has been designed or it’s been evolved
Through many many years to basically an a foil but with an elastic tail right hello um so as the stream goes over it you end up with this um ability for it to then like be forced up the stream it’s kicking around so it also has this
Nice little swimming motion so this dead creature has these incredible properties that are inherent to its being which is beautiful and that’s the kind of mentality that we trying to adopt with Soft Robotics there’s so much we can exploit without having to resort to powered electronics and
Like um proof that it’s dead there’s a YouTube video you can see a dead fish starting to swim enjoy that at your leisure or now apparently okay let’s switch it cool um so yeah in sof robotics we try and exploit a lot of this so we you might see us mostly in gripping
Technologies so this is the Harvard lab um gripper it’s just a fancy balloon you can see examples of it over there um yeah you and I know I wasn’t going to speak until it’s back so as you inflate it the device um conforms and grips but because of the stretchability and
Conformability of the material I see you um it means that you can have lots of different objects you can pick up with the same input so the input is on you inflate the gripper it conforms to an object it could be an egg it could be
A raspberry it could be a dog whatever it’s going to conform to that that shape which is beautiful can’t take much weight but you know we can also do wonderful things like make your soft toys walk around we can put out fires with plastic bags grab subsy creatures with little cages that
Are pneumatically triggered combine chemicals together to make little chemical reactions that make a robot just do that and apparently that’s useful for some reason you find a reason we make Belgian celebrities like healable little um devices it’s essentially a fancy balloon again but it heals itself and we also have recently
Branched into taking frog spawn and turning it into a new life form but we won’t talk about that because it’s terrifying okay so sof robotics is growing we are here to stay I hope otherwise I’m out of a job um we’re expected to reach about 5 billion by
2025 um 6 billion by 2026 and then we dip to 3 billion by 2027 so I’m assuming there some kind of Scandal that means that we have our worth but we’re still billionaires so you know we’re still here it’s all good but with massive growth comes increased waste so as we
Start to have these devices being developed and introduced into homes and different parts of society we need to start thinking right now about how we’re actually going to deal with this you know potentially quite hazardous thing that we’re going to be creating so Global waste right now is expect grow by
70% by 2050 in the UK alone it’s currently 20 it’s about 200 million tons and our waste recycling is not great we only recyc about 45% of household waste um so in 2022 robosoft the biggest conference in Soft Robotics proposed um soft robots for the planet and I took
This on and thought about how we can actually start making changes now that will affect the future of Soft Robotics so one method is to reduce um the amount of material that we’re using one way to do that been mentioned before um by Abby was we can use
Simulation tools with Soft Robotics it’s not so easy the control mechanisms that we use for soft robots are pretty complicated um they don’t behave as you would hope often um another way you can do it is by using other materials in the design process so rather than going
Straight for that final material that you’re going to use you can start playing with things which are recyclable and endlessy usable like Clays or um cardboard papers um it’s a really nice way to actually just figure out the shape the form the size of the thing before you actually go down that design
Route another thing that we can do with Soft Robotics so reuse or repurpose so as I said a lot of things that we do with silicon this is a two-part elasta often that you will mix together and if you’re good you will mix exactly the right amount that you need for your
Robot most of the time you will not do this you’ll make too much so if you have other molds lying around that will make useful products for yourself or for others for example little like scrunchies you can use for um washing dishes heat pads that you can use for
Bing pans down on just take that what would otherwise have been rubbish and start to repurpose it in ways that can be used for you know for others and it’s super easy just throw some CAD designs together and get that going final um method you can do is to recycle this is
One I chose for the paper I submitted and it’s for the essentially the granulation of silicon into tiny little particles that you then flood with pristine elasa to make new parts that you would hope would then have pretty pretty similar properties to um the pristine elasta
Process is super easy um I tried to make it as accessible for everyone is as cheap as possible so I just took some scissors and a coffee grinder and I start to cut that silicon down into small pieces that I then throw into the coffee grinder initial tests weren’t
Great I put into the coffee grinder and apparently when you put two rubbery things together and you try to make them move around a lot it generates a ton of heat and I melted a coffee grinder and the little blade spun off like a little angry bee and embedded itself into wall
So don’t do that but with the powers of icing sugar you coating you’re your particles in ice and sugar you reduce that surface friction it grinds it down no worries and it’s water soluble as well so you get particle size you want you pan for your new robot wash off the
Sugar mix it around a bit Dr an oven and you’ve got a nice new part and then you take that granule you put it into a mold you flood it with a little bit of elasta so this is 30% recycled and then 70% um Rubber and you blow up let’s see if it
Blows up it just keeps going going going it’s pretty good I think it works so we found that the max stress and strain of the material is reduced but that’s to be expected so when you’re doing this kind of recycling you just kind of have to keep that in mind so the
More you recycle it the more you’re going to reduce that um stress strain and you just kind of plan for that in your parts so maybe the first part you have needs to have massive expansion maybe the next part you do just has small expansion but it’s still usable and even with gripping
Technologies it’s still absolutely functional which is great can still lift those Ducks all good it’s nice it’s always Ducks there still a duck over there being lifted around as well I don’t know what it is with us I think we’re just children that like playing with toys soft
Robots yeah so that brings me to fun a bit so been a few days I’m here to play with mushrooms which is great um there is a lab that is growing tons of mushrooms um there’s a PhD student that went out into the woods and found a
Single sample he’s now spread that into his own culture that he makes leather handbags out of which is super cool and I’m here to try and play and see if we can make robots out of it as well so it’s called mym sv1 after Simon the
Creator um it looks like a tortilla or it can look like a piece of artwork I genuinely think it’s gorgeous I love the patterns you can get on the material um so some ideas I’m trying to play with the moment um are cable actuated systems so it’s actually got quite nice
Elastomeric properties in the material so it’s currently grown in a bath of nutrients and you have like a layer that’s formed on top about 1 to 2 mm thick you then put that into a fabric backing and so depending on the fabric backing that you use you can then
Generate different properties within the material which is um really cool turn go again um yeah so everything is sustainable the adhesives that they use are sustainable and biodegradable the fabric they use are linens and hemps so everything can break down after use so we’re kind of aiming for like
Shortterm robotic Solutions so maybe you go into a a woodness setting you’re gripping things and moving things around but then you want it to kind of like bgrade into the soil afterwards so we have these wonderful materials um I wanted to see if we could make it
Um like airtight as well so I put some silicon onto it to see if it come off and unfortunately it did which is a shame but actually I did a test um earlier this week where I just kind of folded it together sealed the edge and it was actually able to hold um
Pneumatic pressure as well so there’s actually possibility of making fancy balloons with mushrooms as well so we’ll see what happens over the next three weeks um one thing that I’m re pursuing is use of kugami techniques and folding um to try and create gripping systems so this is a really simple
Design but by pinching it in a certain way you actually create a gripper just for making a few Cuts in some material doesn’t get it the first time doesn’t get it second time nice well done Alex um but unfortunately the stress of the material like this particular sample
Was very old it was basically taken out of their rubbish bin and it cracked a lot but I’ve been assured by Simon that we can make materials that will actually be able to withstand this kind of compression so we’ll be able to make little griffing systems so I’m currently
Just sitting in an office cutting paper and mushrooms until I can find something that works just right my goldilux mushroom but yeah there’s lots of possibilities I think and if you have any suggestions please throw them at me me so thank you very much if you want to
Look at the recycling process it’s this one if you want to look at my lab it’s this one but yeah [Applause] cheers thank you very much it was indeed interesting and appealing um any questions for her for him sorry as you like thank you for your uh beautiful talk the
Thing is uh that there are some challenges in the Soft Robotics is number one is The Locomotion okay so what do you think that your uh mushrooms inspired Soft Robotics will come up to that at least some of the limits of The Locomotion still yeah because in the Soft Robotics
The prime challenge is The Locomotion how you are going to move the robots uh if generally people use it for different different techniques and different obstructions so how do you think that uh your mushroom inspired Soft Robotics is going to meet those kind of challenges um so with sof robotics I
Feel like we we should be moving in a direction where we’re starting to actually mix around with Hardware as well like I’d like to see it use a lot more as like a skin or as like a end effector for for hard robots obviously there’s the issue there of the soft to
Hard interface and that’s kind of where we should be focusing our attention I think so they can actually start moving in that direction um yeah I wouldn’t maybe see this being like a full mushroom based system I think it would just be a part of a larger thing but there’s other
Materials that you could use as well so like um kombucha Bots have been shown during pandemic when scientists had very little to do but lots of time on their hands um MIT pression has created little Kucha Bots which were essentially sealed but they had the ability to self-inflate
Through um respiration so you can make some robots are able to do quite big actuation um and they’re all biological so I think with that you would be able to use like bias friction or something to create something that was actually able to move around as well although the
Environment of moveing would have to be very you know flat probably or maybe even on a slight incline so it can go down a hill great but yeah there’s possibilities but I really think the integration into to bigger hard assistance is probably the way forwards any more any other questions over teams I
Don’t think so yes no uh yeah um so I don’t know a lot about your field and we talked earlier I think yesterday or hello like this okay um we we talked I think also on Thursday and you you said that uh they were already growing mushrooms for you and I was wondering
Now you’re going to stay here for the next three weeks mhm um how how far I I know it’s a lot of experimenting but like how far can you get in those three weeks with regard to like designing and making iteration Etc like how yeah so um as with my
Like uh what I say yeah I’m essentially at the moment playing with materials that aren’t in the mushroom so that I can use as little mushroom as possible um so I’m yeah cutting a lot of paper to try and get those kyami grippers working I think viably by the end of my stay
Here I could have a grip like a functional gripping system that we could actually Implement and have you know useful for the world I’m not sure but I mean I’m here for a collaboration that’s going to last beyond the month as well so I’m not too worried about like what’s
Actually achieved in my time here it’s more about setting things up so we can move forwards together okay so you will continue with this oh yeah definitely it’s way too much fun not to late thank you that’s all right thanks a lot for a nice presentation I was wondering so the
Actuation is usually external so in your case I think it’s tendon driven or you can have a pneumatic system are there people researching on making those external actuation units Also let’s say biodegradable or sustainable or so on because if you want to bring something in the environment that then biodegrades
I think those parts are need to be removed or need to also be biodegradable yeah absolutely um so I mean my PhD was done in Bristol where they had the robo Foods project and there’s a lot of movement there for creating different things there’s so many chemical
Reactions that we can create like you think about like the Coca-Cola Mentos kind of experiment where it just explodes there a lot of actuation that you can achieve and you know dramatic pressure you can generate so there definitely things or combustion is a way for as well um you can use like platinum
And you can use thatr of water to then create hydrogen which you can then spark and that can then create an explosion so you’ve got a little bit of platinum but otherwise not really too much there that’s actually going to do um harm to the environment so yeah I think there’s
Lots of ways that you can do it that haven’t really been explored okay thanks a lot there is Happy okay thank you very much for the nice talk again we can now uh I can introduce my dear colleagues while thank you thank you [Applause] while uh Mikel is um setting up the
Presentation I will introduce my dear colleagues um uh from the VB here uh sepin and um Yost branard my dear colleagues here from the VB they both are uh actually the Belgian celebrities uh that Alex has mentioned uh our self feeling soft robotic field actually emerged from their PHD Theses
Uh Soft Robotics and sustainable materials so these two guys have joined their powers to make this um a nice research field for so many researchers in our research group now we are uh Co collaborating a lot of uh projects together and they are already uh part-time professors here in the
V so um sepe is going to um uh talk about sustainable Robots part of it but first I think Yos is going to uh introduce the sustainable self feelable materials at first thank you uh so I need to adjust your microphone maybe I will also go there otherwise I feel too small
All right uh so thank you very much for the warm introduction and also for the invitation uh to speak here today especially together with uh sepa we’ve been a tandem already for a very long time actually collaborating very intensely between materials science and uh and Robotics engineering and so inde
Need we will talk about uh sustainable material Technologies uh for uh for Soft Robotics and we had a fantastic introduction already on uh Soft Robotics uh by Alex and what I want to focus on is the life cycle which has also already been uh introduced uh earlier this
Morning um and different steps in the life cycle where we can make improvements towards uh a more sustainable um Soft Robotics technology and so the first was was definitely mentioned a few times already um of course you need to extract materials and that’s where you’ll uh generate actually
Quite a large part of your uh of your impact negative impact uh on the environment uh so I’ll definitely U be looking at a lot of the um extraction or at least the materials uh that uh you can use for softare poy applications um then of course there is
The conversion of whatever you’ve extracted into um into really useful materials um then these materials need to be processed manufactured into the products that you want to use in this case uh Soft Robotics or robotics in general um then ideally there is a long use phase of these materials as as long
As possible but eventually um as already been mentioned a few times uh today also very often they uh still become waste um especially in Soft Robotics as was mentioned silicone rubbers are one of the most common materials the most easy uh to use materials for soft robotic applications they’re essentially almost
Uh non-recyclable we’ve seen just an example of of how at least you could reach 30% and and probably that was still um quite um quite humble probably 40 50% was also possible I don’t know um so you can recycle a certain amount definitely um but what we like to look
At is the different ways how we can U further improve the um the life cycle of these materials from something that looks somewhat linear with a curve here to fit on the slide um into something that is much more circular um and first to to increase the lifetime of materials
Of course you can do maintenance and repairs that’s that’s uh very obvious that’s very standard um but what we are focusing on here at the University in Brussels is to make materials that are self-healable materials that are able and and systems robotics that are able to recover their functional properties
Their functional performance um after they get damaged and after a healing action um and I was as was said um SEO will go a bit more into um how selfing is used inside Soft Robotics I’ll talk more about the materials themselves once they become kind of obsolete they can still be um ideally
Recycled either thermomechanically like typical thermoplastics like uh the plastic bags and so on that that we that we often use using a thermal method or using a chemical method if it’s um if it’s not easy enough or if the materials cannot be thermally recycled we can go to um to
More chemical Recycling and eventually of course we want to completely avoid um waste disposal but if it cannot be avoided that there is somehow loss to the environment or in some way um waste generated we would also like for these materials to be um biodegradable so that
If there is accidental loss um I don’t want to I don’t know any numbers um but every one of us has Plastics inside their body um I don’t want to be um very um apocalyptic here um but it’s it’s it’s very important even even even if
It’s not the whole intention to make a material that would be single use and and would be uh degradable at the end um still for our own health and the health of our environment it’s actually very important that we also consider how these materials um in the end will um will
Degrate so I want to look at a lot of different types of materials today so I won’t be able to go to into too much details but I’m already giving um quite a bit of an overview here uh so there’s different types of materials that we can get immediately from nature um that’s
What people have been doing uh for a long time then we eventually evolved to synthetic materials because they became available they were cheap and so on but actually there’s a good reason to um to continue to look at biological materials such as um polysaccharides um acid conson fatty acids um proteins for
Example um other types of of biomass that is actually generated that is renewable uh through nature um and I’ll dig into some of the um examples one of our PhD researchers um he used different types of biological biomass and converted this into uh functional materials uh so start with materials
Derived from from carbohydrates um I’m showing here the example of corn but you can from many um waste materials from food and and feed industry you can derive um actually very useful products that can be eventually turned into materials that um some of you that are
Present here today uh may have also seen here in the lab tour into materials that can be used for grippers we can make them into elasic materials that um have this recyclability that I mentioned before the selfhealing properties um that’s SE will um discuss a bit in more
Detail um so we can really make um biobased materials either uh completely biobased or at least to a very large extent U biobased materials and ideally also biodegradable another very cool example is is the use of wax wax is actually a natural uh thermoplastic material that
Can be uh shaped in in in different uh different structures and so on um in this case uh on the slides um I give the example of um the combination of uh simply the the the wax together with actually um the black threads which are um a conductive composite made out of of
The wax using conductive fillers and then you can actually make uh things like electronic circuits and so on using a fully natural material that eventually after use would also be able to be uh recycled wax is also biodegradable so that’s actually a um very cool uh example here the
Disadvantage definitely if you look at um lifetime of materials many of these natural materials they don’t have a very long lifetime so you need to choose your materials very well uh depending on the application the the performance that you expect from your uh materials then um so the mying materials
The mushrooms um have already been mentioned in the previous talk um so these are very interesting materials because they are grown by bacteria so from typical nutrients like like sugars water um air and so on you can actually grow materials um into large pieces of uh indeed something
That’s that’s letter like that can be used they’ve they’ve used this for for making shoes for making handbags um now um we’re glad to have Alex here who’s going to try to make U robotic systems out of these he already showed uh one of the gripper Concepts and what is very
Cool about these materials they are kind of living materials um the materials that Alex has been using um are are dead in a sense uh they’re they’re they’re killed so to speak uh to avoid they that they grow any further but actually they they can keep some kind of um viability
Which can be reactivated and what is shown here on the slide is that if you cut actually um a large piece out of these materials and if you give this materials nutrients again they can actually grow new material uh to to seal in this case the the hole that was made
In in this material um this is very very new research uh very fundamental but that’s something that here at the University we’re also trying to um exploit and if we can make uh that into U robots that can heal large damage that would be um that would be very
Cool there are other um biological materials um that can be used um BAC they um they generally consist of a lot of a large portion of of proteins up to 60 70% they’re made out of proteins proteins are are polymers um for uh synthetic purposes not usually the the
Best um properties but they can be combined with other materials to make um make elastomers make other um types of materials that can be useful for these kind of applications um if we get stressed um sometimes we start eating a lot we start accumulating a lot of fat in our body
And so on bacteria do something similar when they get stressed some of these bacteria they make polymers they make polymers to store them as an energy source then in a later stage they can use these polymers for um to convert them back into into energy so actually
Um they can be used as very small microscopic factories to make uh polymeric materials and these polymers can then be used to replace uh petroleum based polymers um I said they can convert them back into energy uh for for so that also means that these materials can quite
Easily be degraded eventually at the end of their their lifetime so these uh polymers they are generally also U biodegradable people have also been using really living materials uh for what they call under general terms uh robots so what people have done for example is to use real life bacteria
Load them with with whatever something that they might want to use to transport uh maybe a drug or something as drug carriers uh maybe some kind of sensor maybe something else that they want to activate and they use actually the um the bacteria themselves to move around uh certain certain objects certain
Functions um also some researchers have used um the muscles of worms to either either use them as as elastomers uh for um for well actuation let’s say or since these are um muscles are um electroactive materials uh they actually even stimulate these muscles to create a um a activation
Motion uh so that’s of course also uh maybe slightly scary um option out there then there are other materials that were much more common uh familiar with um like paper for example paper can be used for paper uh robotics um Alex also show some examples if you have
Origami Cami all these kind of things you can make actually very cool um uh deformations motions and so on using um uh let’s say a smart intelligent way of of folding and and and cutting um paper um into different kinds of shapes from the paper and Pulp industry
There’s also a lot of waste some waste streams can also be valorized to make materials uh so one of our colleagues M St from VTO she used um ligin which is a um a very common way stream from paper and Pulp industry to make um self-healing or um recyclable uh
Thermosets out of of these materials um in Gent University they fractionate this liin into smaller molecules which can then be used to replace other petroleum based aromatics so there’s a lot of opportunities there um from another waste uh source humans are a s side stream of uh the production
Of um poly eater Fano bef which is a biological replacement or biobased replacement for uh p p which we all knows from from pet bottles for for soda and so on from that waist stream also a colleague of ours in in Len made this kind of um finger grippers that are um
In this case it’s a tendon driven system um so many many of these um waste streams can still be valorized and uh you might even have kind of a negative um well a positive impact or lower negative impact on on the environment by actually um doing something with these
Waist streams especially um replacing other um sources the needs to extract from other sources then I want to dig a little bit into the the work that we are definitely doing here at the university which is on reversal pmer networks I don’t want to go into all the um chemical details I
Just want to say that what we’re trying to do is we try to make a class of material that is generally not recyclable to make it into something that is recyclable and along the way we’re actually also um in many cases uh giving these materials the property to
Uh to heal themselves and the way we do this is by incorporating different types of uh of reversible chemical bonds uh if anyone’s familiar with with uh with chemistry these are physical interaction physic chemical interactions bonds that are formed very easily but also broken very easily and that can reform very
Fast and that’s what we want for self-healing we want our material to break and be able to recover its structure its chemical structure and therefore also its properties um as as as fast and as easily as possible uh and you can use phys chemical interactions you can use uh reversible coent bonds
Again not going to the details here the difference is that one is is much stronger but generally um requires more heat for reprocessing or uh requires more time uh or even heating or some other kind of stimulus uh to activate the the healing function for the sake of completeness I
Also want to say a few things about um processing I don’t have the time to to really go into the details there but what is definitely obvious for for processing is that depending on a type of process you choose you may have um a certain amount of of waste that you’re generating there
Of course if you choose material that is recyclable you may be able to reuse uh some of that waste but you can also choose um within certain design constraints a process that um that prevents waste as much as possible um and you can use different kinds of formative methods where you use molding
In some way or another uh to create shapes you can use additive manua manufacturing techniques in general and in additive manufacturing you produce much less waste but you have larger lead times you have smaller volumes that you can create so there’s always a balance that you need to strike between the um
The processes that you uh that you choose there or the um requirements that you have uh from your process and uh we also reviewed that actually these self heing materials these reversible bonds that you create in these materials uh can also provide new opportunities in terms of
Manufacturing of U of structures uh so things that were um previously not possible uh like like welding of of thermostats welding of of um elastomers some kind of joining techniques um have now become available um Ellen who’s by the way also with us here she did um the most part of
This uh this review she looked at a lot of different um requirements but I want to stress here especially um it was also stressed in this morning’s presentation that from a very early point we need to look at sustainability um before it becomes uh too difficult to to make a change or um
Before the technology becomes too mature um to um and and and there’s a lot of resistance against um movement to more sustainable uh options and especially from a manufacturing point of view of course um waste reduction is very important um energy consumption uh is is definitely very
Important um in together with with all the other um requirements of course that you may have uh to create your products and with this uh I want to hand uh to my colleague sepa who will talk uh more about the application uh itself maybe I [Applause] to yeah cool I think that’s good
Yes all right cool thanks uh Yost for giving the first part of our talk and also thank you Alex for giving an introduction into Soft Robotics it gives uh it makes my life a bit more easy um so soft robots uh they’re like Alex says it’s a it’s a new field inside of
Robotics and they bring uh a lot of new opportunities um they’re basically robots in which you make large parts or the entire robotic system out of uh flexible materials and that’s interesting because you can give them intrinsic flexibility which is nice for yeah safe human robot interactions for
Example in in a wide variety of applications but also to generate uh grippers like the one we see there and also on the screen to uh handle very delicate uh objects like for example fruits and vegetables in uh Agri food but there is more uh I think because you use flexible materials polymeric
Materials you have also get opportunities in new manufacturing techniques that can be used to make these robots for example a lot of parts of the robots or or not the entire robot can be made out of uh additive manufacturing techniques which can be in many cases also a sustainable solution
Because not a lot of different steps are required and also there’s not a lot of waste of material then because you create robots which are activated by by also these soft actuators you can also downscale them so this means that this Soft Robotics field is also opening a
Lot of possibilities into new robots which are miniaturized and can for example work inside of the body or in other very small application so there’s a lot going on and a lot of new applications which are not replacing the traditional robotics applications but are just uh getting our robots into
Other um uh Fields uh in which they can assist us but there is a big problem oh this doesn’t work there’s a big problem that is that these soft materials they are safe but they’re quite vulnerable so uh they get damaged all the time and it’s a bit like us uh
Whenever they encounter a sharp object they’re easily cut also soft robots they consist out of a lot of moving parts so that means that throughout the body uh the materials is stressed a lot and uh with cyclic um loading you’ll get fatigue which are tiny micro cracks which eventually also propagate into
Larger cracks and uh the system fails also a lot of these systems are consisting out of multiple materials because you want to increase performance or you want to integrate sensors inside and if you have multiple materials in which the the the mechanical properties are different at the interface you have
A lot of problems with delamination so over time uh these weak interfaces they tend to um break and delaminate so this means that these robotic systems they are safe they’re adaptable and so on but they have really a short Lifetime and for example um this type of
Soft gripper which is a granular jamming um soft gripper it was uh there was a company that was trying to commercialize this uh this type of robot robotic end vector and they had troubles because the amount of grips they could achieve was only 50,000 uh grips before it failed
And if there were sharp objects inside of of this uh application field then uh it was reduced to even 5,000 CPS so that’s really not a lot if you go to an industrial application and this is really showing that the limited lifetime is actually um making these these soft
Robotic systems in many cases not economically competitive with with other Solutions which is of course a problem if you want to introduce them and then of course there’s the ecological uh um aspect if we are developing a lot of pic materials which are most cases like your me mentioned not recyclable or
Recyclable to limited extent then uh this also will generate an an an ecolog ecological impact and there is a lot of legislation going on that is also transforming this uh ecological um value values into economical values which means that this competitive competitive uh value will decrease even even more so
That’s a problem that we wanted to to tackle and together with Y and brah and he here at the University uh we um we found a solution by making them out of self-healing materials so the yellow materials that you can see on the slide they’re all made out these self-healing
Materials and as you can see we can damage them quite severely so it’s really a big um yeah knife that we put inside of the robot so we make a macroscopic damage recall and um these gers we could heal them entirely by putting them at 40 uh for 40 minutes at
80° and afterwards these damages were completely sealed and the the system was actually as good as new so it was almost magic or or I thought at the beginning of my PhD but then we started looking in depth and so this is the materials uh they’re they’re like an an elastom
Elastomeric or or Theros set materials so that means that they have like a network structure so everything is connected and the special thing is that these connection points which we call cross links they are reversible in our case it’s a deals Alder reaction but like your said there are many other
Reversible reactions and because it’s reversible we can heal and we can also reprocess it so because what happens if we um damage a material this is a video so I hope yes so when we damage then locally we will actually break these reversible bonds so these uh bonds are
Then locally uh broken and they are reactive so they want to um react back together however there’s like a gap so then we increase the temperature when we increase the temperature throughout the network we will break these reversible bonds we don’t break all of them we keep
A few inside of the network such that we we stay in a solid state so during the healing the material is really solid and not liquid but there is more mobility and because there is more Mobility slowly we can start to close uh these gaps now whenever these gaps closing
Closed we can actually reduce the temperature so H we cool down to room temperature and then these reversible bonds they tend to react back together throughout the network but also across this uh fracture surface and as such we can um yeah we can recover the entire material and if there are no
Misalignments uh on like a geometrical or microscopic structure the this healing can be performed multiple times so uh in fact in theory in infinite so that’s the healing uh ability but if you increase the temperature of our materials even uh more then eventually you break the entire network structure
And you will flow so that’s something that we don’t have with uh traditional uh Network polymers like for example the rubbers which we all know for ties and sadly we cannot uh recycle these tires because they are irreversibly cross-link or vulcanized and as such they end up a
Lot in fill or they’re like burnt and so on and it’s a it’s a real problem for which also reversible polymers can offer a solution so the materials that we make our robots of they can be formed in palletes then we can put them inside of
A mold we heat not too much temperature actually 130 150 and then we can mold them inside of these uh grippers we can chop them back up and uh we can do it again for multiple times and the uh detoriation of of the properties is actually quite quite limited I think
There’s like a recovery of 95% or so of the mechanical properties which is quite good to when you’re recycling entire material now because we can reprocess them we can also use them in new uh uh manufacturing techniques like the additive manufacturing technique the fuse filament fabrication so it’s an
Extrusion based uh printing technique so again we start with uh pellets or power ERS we extrude them in an extruder to make filament and this we Feit in the printer and then we can make this uh selfhealing or Ellen who is also present can make this nice uh robotic structure
And what is an extra feature of our materials is that during printing the layers they actually heal or bind also together so this is actually this leads to isotropic Mechanical properties inside of a print and um this is actually a problem uh the 3D printing industry is facing a lot is like how
Could do how can we get isotropic mechanical properties throughout our print because usually the prints they fail if you load them perpendicular to the printing Direction because there the layers are not very well connected there there’s only limited uh interaction usually during physical interactions in our case we print and the deals aler
Bonds or the reversible bonds they also coal bind these layers and as such we can get uh very good properties which eventually also leads to more robust components uh which um with isotropic properties uh Beyond uh prototyping now we can also chemically recycle our materials because if you in
If you put them in an appropriate solvent they start to swell now normal uh Network structures they just swell and then they then they they don’t dissolve because they’re irreversibly Crosslink in our case if you if they swell you decrease the concentration of these cross links and then they also
Tend to to break down and eventually uh dissolve completely after which we can um cast them and generate again uh new materials that we can reuse so we have two types of recycling mechanical recycling or thmo mechanical recycling which we do Extrusion for example and then also the chemical
Recycling now in our polymer group together with Y and the other researchers we can also t these mechanical properties of these materials so we can adapt the network structure we can take longer polymer chains as monomers or shorter and as such we play with the Crosslink density
And we can make materials which are very stiff and also materials that are very very flexible and everything that it’s in between so as such you get a lot of design freedom to develop soft robots which is very nice for me but also these two materials a stiff one and a and a
More Flex one they all have the same cross links inside of the network so that means if you put them together you do a heat cool cycle then they will also bind together and as such we get very high strength interfaces between materials with completely different uh stiffnesses and
Uh for example we have a very stiff one and a very flexible one and if we put them together then they don’t break at at the interface and this is important because I already mentioned that it’s a it’s really a problem in in Soft Robotics but also in many products in general that
The connection between a hard plastic and a soft PL plastic is is always something that that eventually breaks think about your uh I don’t know swimming goggles uh there’s always a hole there where the rubber is actually connected to the plastic gloss of your of your uh goggles and um this is not
The case with our materials because they are chemically binded together without using any any glue of course all the materials are s still self-healing so here we went a bit more radical we chopped it in half we put it back together and it also healed so that was quite
Nice but we can take it even one step further we can make also sort of modular uh soft robots and take the modularity to the extreme where we create voxal based uh robots and foxal based robots they consist out of little building box foxal which are 3D uh pixels let’s say
Cub cubic parts and we make them out of different materials all self-healing materials and some we make active and some we make uh passive active they can be inflated so they pratic and the passive they also uh they’re just solid blocks but they have different mechanical properties and what we can do
Is we can stack them together and then we can again perform a heat cool cycle and then everything is uh uh binded together and as such you can make simple robots like uh a gripping robot uh but the nice thing is that because of the selfhealing
Ability we can also chop it back up so we can mechanically break the interfaces get again these foxal and reconfigure them in another robot a small walking robot and as such you can create modular systems in which you do not recycle the materials but you actually start
Recycling the mod modules so you can create reconfigurable robots and if we yeah go towards a bit more complex structures having more voxal we eventually can make uh bigger systems with uh a lot of different functionalities which can evolve maybe inside uh during the the the use of this
Robotic system and as such you can generate let’s say um robots that adapt for a functionality rather that they are just designed for a single uh functionality now I said that we needed to use heat to to um heal uh these soft robots of course it’s not very energy
Efficient to always take an entire robot and put it in a in a furnace or or an oven so that’s why we looked at can we also integrate the heating inside of U this robotic fingers so we made selfhealing conductive uh materials which are uh our materials but just with
Uh carbon black fillers and others and uh we have a heater that we put at the bottom of of this finger and whenever you cut the entire finger as well as the as the heater you can put it back together and then if there’s a current part you can uh increase the temperature
Because of Jeweler Heating and then you will have an internal heating system that heals the the finger but also the heater that is inside and what is also nice is that if you cut um and put back an a conductive part locally you increase the resistance and that means
That if you then apply a current locally also the heating will be much larger so we can see that purely passively so we don’t use any control that the heat is concentrated there where we want it to be because uh it’s at the The Zone where
It’s it’s uh it’s damaged and that it can then completely heal but we try to find a solution also for this problem on the material Level so we’re continuously trying to improve our materials and that’s where where we found a material that can also heal at room temperature
So we changed the composition of these materials we also patented it and then we found a material where we don’t need to use any heat and we can just um heal it at room temperature which of course uh is is interesting because then we don’t need to have an additional in uh
Integrated system that provides Heating and so on and yeah we we also use it to make uh soft uh robotic fingers like this don’t know if the videoos always start and what is also interesting is that we always look at healing from a material point of view and a system
Point of view and from a material point of view we reach in this slide it’s 24 hours and then it’s completely healed we have already materials that heal at 5 hours but that doesn’t mean that the system cannot restart much sooner because healing on a system level requires sometimes not the full recovery
On the material Level that’s so that’s what the video is showing here so we have a damage and only after a few seconds it’s already um healed it’s not completely as it was before but it can already be activated and then after some time during the the the operation of a
Robot uh this can be uh healed further on the material Level nevertheless for many damages you do need to he uh weight so the system itself has to go into a sort of on offline healed state and therefore um it’s very important that uh the system knows itself when
It’s damaged because then it can take the appropriate action to to to rest a bit even if it’s for a few minutes so that’s why it’s very important for us that the system knows if it’s damaged and as such we we we investigated can we make sensors such that we can um measure
If if there is a damage and preferably also where it’s damaged so also the selfhealing material was used for this and and uh by adding conductive particles we could make pzo resistive materials so that means that if you strain them or deform them then you see
It in the resistive signal so that’s why you can use it for deformation or touch or Force tracking but when you cut it you also disrupt the uh conductive part and as such you can also see this uh as a damage detection sensor because you see it here you see a steep Peak
Whenever something is damaged and then when it’s re Rach when it’s reattached uh reattached sorry and um uh there is back contact then you will also uh see it because there is a current part which is again uh generated so as such we we play with uh damage location or
Multifunctional sensors that that measure touch as well as um deformation and and damage and of course you can go towards damage localization if you start to B multiple of these sensors in different patterns inside of a soft robotic skin so we envision a soft robotic skin which can be placed on a
Soft robot maybe even on a hard robot that eventually can also measure where it’s damaged the location the severity and then uh can also completely recover from it it’s also when you start to combine a lot of sensors like this in uh it starts to become quite difficult so
That’s why we use uh machine learning to uh um to um yeah transform these multiple inputs into an um yeah damage tracking map uh and uh we get help from the AI experience Center because there are a lot of people knowing a lot about uh AI here what is also interesting is
To start looking at how can Hardware selfhealing um be actually connected to uh software adaptation so sometimes the healing can for example not be completed but on a on a software level you can actually start to adapt your uh your your your behavior such that you can
Overcome uh a damage as well it’s a bit similar to how we behave so whenever I have a torn muscle uh I can still walk a bit but I just adapt my behavior as such and this can be temporarily and but uh or if it’s a severe damage uh it can be
Uh permanent so we try to to bring that ID also into soft robots so this is a bit the overview that y made and I added a lot of stuff so yeah I think I hope we could convince you that we’re trying to look at really different uh approaches to
Make bring new reuse and recirculation approaches into soft robots by looking at the system level reconfiguration and so on which is of course achieved by new materials but also on the material Level uh we see a lot of uh opportunities and then I also want to um I think it’s
Important to see that um soft robots or robots in general they’re like a platform which can actually be used to bring these type of Technologies materials but also different approaches into other applications so far beyond uh soft robots into just um typical uh products that we that we use daily so
With this I want to thank you all uh and of course if there are some questions then your and I can can address them yeah thank you very much um for covering both aspects from the Material Science and Robotics uh there are questions I will start with the inperson
Questions hey thanks for the talk um you’re actually one of the first images use in my PhD so it’s nice to meet you guys in person um this is all really cool so you’re currently making systems that are completely made of this material is a
Scope for you to embed this kind of like as a network within other materials so you can have selfhealing properties embedded into say elastos or maybe other bderived materials I think it’s um so the the chemistry itself um you can you can introduce this yeah do that
Uh so the chemistry itself uh in principle you can introduce it in any kind of um elastomeric material um however it’s it’s not something you can easily mix in you have to do chemical modification uh to achieve it but for example we can we can introduce we’re actually doing this introducing these
Bonds in uh for example uh silicone rubbers we’re introducing them into um in this case it’s it’s more um polyprop oxide based systems we can do this in in in many different types of materials so in principle you have a whole register of of possible uh polymeric backbones and possible uh properties available
Yeah yeah thank you for a nice presentation so you showing us some uh um materials from built from living organisms for creating grippers or something like this as I was wondering uh so living organism would actually change properties all the time I mean I realized that for Soft Robotics you need
Totally different control strategies but isn’t can this be a little bit annoying if you can’t be sure you know if the control method that worked yesterday still works the next day because the organism a little bit changed I think yeah it’s it’s definitely a valid point uh which is
Something we should address but in our research we actually took the yeah we we turn it around so we saw that it’s not for the living tissue but more for our materials the the reversible polymers they also behave quite different based on temperature for example and other um
Temperature time and so on and this in many application it’s it’s a problem but I think in robotics the advantage is that you do have sensors even in Soft Robotics they are integrating more and more sensors and also a control unit such that you can start to compensate
For these uh changes so I guess I’m not an expert in living tissue it might be more complex to make a sort of um estimation on how the material will behave but with our materials we can based on the temperature and so on we can estimate how the material can behave
And then the the robotics control can then compensate for it so maybe like this um it can be that you have materials living tissue that is derating uh little by little but that the the system itself can can account for it and if I could complement uh so in in
Engineered living materials typically people um use living materials because they are adaptable and then they often engineer them to have a certain adaptability and change their function um as a function of time um of course it’s it’s a bit similar to our our body actually we’re constantly calibrating
Based on on on movements that we’re making and so on so I think what’s also very important is a Time aspect how fast do these properties change and do you have some kind of a of a feedback loop uh to to help you control uh for
Instance if if you go running first your you feel your muscles are kind of stiff and you’re you’re adapting while you’re you’re your body is H changing physiologically so definitely uh control strategies will need to become uh more powerful um yeah thanks for the interesting talk um was wondering uh
First question is for the grippers those ones um what kind of actuators are you using to make them bent is there like a compound or another sort of material in it yeah maybe I relied a bit too much on Alex introduction no they’re pratically driven so we use different types of
Activate uh activating uh so you have like exter I call it externally activated principles like tendon driven and and pratic systems there the one that are there are also pratically different and for me we we select those because you can generate High forces and high speeds with them um um compared to
Other Soft Robotics activation and also because um in these platic systems they’re why they’re widely used in soft robots but whenever there’s a little hole Yeah it’s of course has a big uh influence on on the behavior so that’s why we targeted those first to make them
Uh selfhealing but there are a lot of different um Act uation principle shape memory polymers or or for example another one in which the material itself uh creates the the behavior okay um as I expected a compound but I saw you have uh compounds as well my follow-up
Question would be for the recycling um you said it’s quite easy with the material because you can just melt it down and uh form it back if you have a a compound or two different materials say the capacity thing you had in there for heating yeah um does it posee a problem
For the recycling or is it very easily separated yeah so it’s a challenge as in many uh in many M material components the challenge that we’re still facing is that now we focused on let’s say yeah binding the materials very well together such that they’re because if you bind
Them together these U these these sensors and the Matrix or the the soft robot in which you you put them they become more robust it’s also better for sensor properties that they they follow the material in which they are uh embedded however it becomes more challenging to to separate them of
Course you can just yeah recycle the entire part and you will have a little bit of um filler which is inside of of a material but then it’s difficult to estimate how your final um material will will look like or which properties it will have so I think there’s still
Challenges uh on how to separate the material so it’s always yeah it’s difficult to bring bring both very strong interfaces while having reversibility also on the on the but there’s a lot of work on more on the material Level that is done uh on that part I think maybe you are reversible
Adhesives maybe you know also some more works on so it’s a big big part okay um there’s uh there are some questions from slido uh one of them is to use a bit broad but it’s uh asking if uh with the topic it’s infancy could you give your
Opinion on what obstacles you see in a scale up uh which would be necessary for Meaningful impacts yeah so it depends there a bit how how we Define the the the topic uh the question was raised I think just after my talk so it’s it’s mainly on the
Material side it depends on the type of materials um protein based Plastics have been made since the the early 50s or so uh so this is this is scaled this is used already for for packaging application and so on so many of these materials are quite mature and they’re
Ready to be used in in Soft Robotics for example um other Technologies they are very novel um like AR uh the chemistry that s was talking about many of the other examples that I gave have only been tested in the lab so definitely there are a lot of um issues uh with
With scaling with maturity and so on these materials need to be tested um it’s a bit the same with Soft Robotics it’s also to a great extent still uh in the lab that these materials are being used and these systems are being used so I wouldn’t call it per se an obstacle I
Would call an opportunity an opportunity to learn how these systems how these materials behave when they’re actually used in real life environments in field tests and so on when they’re exposed to um their the environmental um and and other conditions in which they they should be used and then to learn how to
Adapt uh adapt these materials to to improve them um for instance for uh the materials that we are working with we have now a project granted to to see whether it’s uh whether this chemistry is scalable um because there are limitations in terms of the the temperature into which they can be used
Um whether in practice uh the healing can be uh can be achieved multiple times we know in the lab for example we’ve been able to repeatedly heal the same area about 50 times or more um which is close to let’s say theoretical infinite
If you look at a um a use user point of view uh but this might be entirely different once they’re exposed to u to their use environment if you think about horiculture for example and you start to to pick um I don’t know whatever food stuffs in in in a vertical farm for
Example then you’re exposed to a lot of soil a lot of other contaminants and so on um so I think it’s really the the scalability um and and and and and the maturity raising the the maturity okay another question uh maybe for SE uh this material seems to be the
Optimal Sol solution and solve so many issues H what is the downside in applications plenty I think y you already mentioned quite some uh no I think um there are still some questions to be answered uh depending on the application that uh we envision for example we are going to
Also start up to look at the biocompatibility and the safety of our materials we have a good estimation that it will be okay but yeah these are things that questions that we are still need to answer and the upscaling is something else and of course if you bring these
Types of materials into an actual application cost becomes also very very important and it’s not uh our materials they’re not super expensive but if you’re trying if you’re like comparing them to very cheap uh Plastics which are used uh today thermoplastics and so on then in many cases we cannot reach these
These very very low cost and sadly that’s also still the reason why we have this cons consumption uh society which uh brings another challenge is like how can we actually develop new business models that allows us to introduce components in inside our society that last much much longer because now uh
There’s a lot of um yeah parts and and and products replaced not because they are they fail but uh because they’re outdated or they’re not so interesting anymore uh and and I think that’s also another uh difficulty that that we face is um how we we cannot just introduce or
Go to companies at the moment and and say now U create these materials which or products that last last much longer because it goes a bit sadly against the the business models of today yeah okay uh you partially uh answered the question but I will still ask if you
Would like to add a few more things if needed the question is if the polymer becomes injured or cut multiple times is it fully heating healing sorry or does it previously cut area become more or prone to injuries or Cuts in the future um yeah there there is just one
Uh big limit or two big limitations so one limitation is whenever there’s no contact between the material then it doesn’t heal so that means that if the um the gap or the wound is is open yeah we don’t grow uh although you showed something very interesting and I still
Some questions we need to talk more but where it grew but for the other materials it doesn’t grow and that means so if there’s a little dirt part or something greasy then there’s also no contact at that area and then it doesn’t heal as well and then there we see also
In actual applications that um you also get misalignment so with these elastomeric material if you damage them they tend to snap back into their initial position making good contact but whenever there’s a misalignment even on a microscopic level you induce a little stress concentration and you your material is
At that location not as strong as before so that’s the limitation it’s not on the material Level but it’s more on uh the microscopic level how you put it back together another question for Yost uh the question is when the material is used uh to replace the tires I’m wondering if
It becomes a problem that the material is cell filling at 80° for example can the temperature limit be increased uh yeah that’s very good question so um that brings me back a little bit to do what SE said in terms of requirements it’s important that the
Material stays solid um at the at a use uh temperature and the use temperature can be very wide uh can very be very broad so if your tire is expected to be heated up until uh 80° or more um we would then need to design the material
Such that it remains solid until that temperature and actually uh as as as also mentioned to some extent by sepa is that if the the material heats up naturally this actually speeds up the the healing event also um so the material can be designed in such a way
Um at least to some extent uh many properties can be designed um quite individually quite independently so the temp temperature at which material becomes uh liquid uh can be can be changed um while um keeping for example the mechanical properties the same so there are um quite a lot of design uh
Parameters that we can use uh for for these kind of applications and it also brings me back actually um to the to the question earlier by uh by Michelle is that um we can actually also design uh for instance two materials to um actually become liquid at different temperatures and
Then you would for example be able to remove one of the two materials um at a at a lower temperature and then reprocess the second material for example at a higher temperature so there there are some some possibilities there some flexibility but these design parameters of the materials they’re also
Not completely independent um but there is a a good um playing Ground there yeah good point uh one more question uh but it emerged before you completed your presentation sep so you partially uh answered can you build a complete robot out of this material instead of just the
Fingers yes yeah that’s our Dream but um yeah in the beginning of or even at the middle of my my PhD I was working with Y who was supervising me and I said okay let’s make soft robots we we have a nice material and I said to him okay I need
50 grams and you were like that’s that’s crazy because I think we were working with grams at at that point uh but now we did the upscaling so we put a lot of effort in this and now we’re more on the kilogram scale so now uh we are uh
Evolving towards uh yeah going larger systems and I think we will’ll be able to indeed make uh bigger parts of the robot selfhealing although we are always looking at um yeah more from an academic point of view on like where do we put the self feeling because I think it’s
Not I think as as a it’s not good to make every part sing so let’s just start maybe with the skin or so so that we can make a soft cover for a robot which is which is flexible safe uh but can also heal so maybe we will go with a skin or
In the future for a bigger robotic system Qui question on the biological fungi and bacteria based materials from the first part of the talk so so you said you can just add a little bit of nutrition material or something like that to make it grow can you actually
Make it grow into the original shape or does it grow all over the place and like could this be the solution to fixing those cuts that don’t heal anymore after misalignment or something like that could there be a combination of the two material classes uh okay quite a few questions
There um to answer the first question uh so it’s not simply um adding nutrients um it it depends you need to design the material for it so in this case for example you need to uh create create a condition where there are spores inside the material that can be activated and
Then indeed by just spraying nutrient solution on it uh they they can they can start growing again uh second question can it be controlled well to a very small extent uh so in the lab it worked quite well in the lab you can control it uh if you imagine for example a leather
Jacket or or a purse or so it gets damaged it comes into contact with uh with the um let’s say the healing agent or so um either willing or not it might indeed start uh continue to grow as long as it has uh nutrients uh so we actually
Recently applied for a project to to gain in in in some ways that I cannot disclose control over this um over the the time the location and and the duration of of growth um combining the two materials uh that could definitely make sense uh so especially with synthetic materials but also with these
Biological materials often uh combining the the multiple advantages of each of the systems um makes a lot of sense uh so for instance um the the letter composits that were shown they actually are grown onto kind of a substrate which can be a textile or so um so if you
Replace a textile by something something else synthetic that has certain uh selfhealing properties uh for instance SE was talking of the order of of of of minutes or hours while if you think about biological materials you should rather think uh in the order of of days
Or weeks um so there a combination would make a lot of sense so that you can have early on uh damage closure and then maybe indeed regrowth um at a at a longer time scale yeah thank you for that presentation um I’m thinking okay let’s imagine that you’ve got a an industrial application
For your grasp technique what you’re doing now like with this soft robot over here and and an industry is is is has come up with a plan with it and they have I don’t know hundreds of robots or whatever within their Factory that use it what level of Education will the
Person need that’s going to be in charge of uh making sure that those robots are working and to repair them on a daily basis um well I think compared to well the intention if um if we would for example want to to commercialize this and um you want to um eventually for
People to adopt this kind of Technology we would need to make it in in such a way that it’s um as similar as possible to the the technology that they’re used to uh so if we were replacing a soft robotic technology we would try to make the system as as autonomous as possible
So that uh for instance using the the sensors using embedded heaters for example um and a control system it would be able to uh sense damage and um act act upon it depending on whatever is is programmed in terms of for example loss of performance um so that uh the the
User would need to be educated as as little as possible of course if you’re replacing a a hard robotic system in a in a automation line for example by something soft robotic then there is a learning curve of course uh for the operators okay thank you very much for
The uh very nice discussion and your um beautiful talk let’s uh Applause all all the uh speakers of this session and let’s enjoy the coffee break for half an hour before coming back e e e e e e e e e e e e e e e e e e e e e
E e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e yes uh welcome everyone to our very last session of the day just had a coffee
Break I hope the coffee kicked in already and uh you’re uh you’re ready for one more presentation our last session is on circular econom om and applications focusing on the end of life as we said earlier unfortunately the first Speaker couldn’t come so we only have one more presentation but that will
Give us a bit more time to ask questions and uh Network a bit in the end and um I have the pleasure to introduce our next speaker uh Alish UD uh Alish heads the department of automatics buer cybernetics and Robotics at the YF Stefan Institute in lubiana in Slovenia
And he’s also associated with the ATR computational Neuroscience Laboratories in Kyoto Japan uh his research focuses on various issues in robot learning especially imitation learning and learning by exploration where he combines statistical learning techniques and reinforcement learning to increase the efficiency and autonomy autonomy of the acquisition of new sensory motor
Behaviors he’s also interested in the applications of robotics especially in manufacturing we’ll hear more about that now and Alish has also LED several European and national projects as coordinator or principal investigator at y Stefan Institute and uh I personally also had the pleasure to work with him
In some of these projects and very much looking forward to what he’s up to now because it has been a while um and uh yeah he’s uh he’s going to present his latest project which is called Recon Cycle uh this initiative focuses on adaptive robotics for recycling electronic devices and aiming at a
Advancing the capabilities of robots in this area and with that let’s give some Applause to Alish thank you forisa for a nice introduction so let me just start my presentation so as Francisco already announced my presentation will be about adaptic robotics for recycling of electronic devices this is actually um
Part this is actually a European project called Recon Cycle which is led by my group and there are actually a few other groups collaborating with us there’s a um university of getan with Florentine V and the Technical University of Munich with Sam had in and uh um IIT from
Geneva the Antonio bis group plus two companies does QB robotics which is essentially a Soft Robotics spinoff of IIT and the electroc cycling company which is a company that is actually interested in recycling of electronic devices so what do we actually try to do in this project um so it’s about the
Recycling of electronic devices and this sort of devices are usually actually um recycled using some using methodology called crush and separate meaning that the devices are put into some sort of a Crusher and put into basically put into tiny particles and this this particles and then sort of physiochemically separated into reusable components
However there’s a big problems with this because uh many electronic devices actually contain batteries and you cannot put batteries or into the crusher at least if you do if you do this then you basically basically risk fires and actually fires are quite common in recycling companies and and they’re a
Big Pro problem so a project is actually is uh so it’s already difficult actually to know whether the device contains a battery or not and uh there are European projects that actually deal with these specific issues but uh that’s not our issue actually so we assume that somebody has already decided that this
Uh that uh that our device contains a battery and the problem is done how to remove it um so um there’s a um this is of course not uh not an easy problem like in uh a assembly or disassembly uh in the in manufacturing so there’s a
Huge there there’s a huge variety of the devices that need to be that need to be disassembled so just if you look at the devices of one type there are many of them there are here you can see many different smoke detectors here there are there
Are here there are many uh there are many heat cost allocators on the right side or left right side and basically uh basically um all of them contains batteries so besides uh besides this um the next problem is that also all these devices are in different conditions so
Even if they are the same type exactly the same model they still since these are used devices they will not be they they will not be the same so there are missing parts and there are and uh and this uh this missing parts or or used up parts they actually determine they
Actually hinders the recycling process so finally the device design itself is a problem so um some electronic devices specifically designed not to be disassembled so here you can see some smoke detectors so the batteries are often soldiers actually the covers are made so that they cannot be opened to
Guarantee that they work all the time and this is of course a problem for recycling because basically recycling at end of life is uh not considered during the design of the the design of the products um so um the solution is of course the manual operation but this is
Way too expensive actually to be practical as you can see here even a human worker has quite some troubles actually to get out the battery out of such devices so we’re trying to uh so our project is actually trying to look at how we can address this problem with
Robotic solution and for this we design a sort of a fre stage approach so in the first stage we are looking at how to disassemble um just uh one of the device of a certain type basically we are looking at the Then when we get uh get new devices then we need to basically look at the existing disassembly procedures select the correct ones and uh possibly adapt the parameters to actually to actually be able to disassemble this devices um so um obviously we need a lot of adaptability for this uh for this
Sort of task so first of all our cell itself needs to be configurable basically different devices will require different tools different fixtures the positioning of the robots needs to be different might differ and so on so our robotic cell needs to be actually quite uh quite modular and quite modular and
Enable us to quickly get from Vine cell design to another so here you can see then a few different cell designs that uh some made for heat cost ators some for smoke detectors and so on ideally of course all this would be um reconfigured totally automatically we are not yet
That far but we can actually Aid this reconfiguration process a lot so um so for example one we have actually quite a few quite a few ways of reconfiguration in the cell and we are basically developing our fixtures to be a configurable modular and the cell itself
To be modular so for example the cell consist done on of this of this small modules which we called uh archetypical module uh as a or a basic module of our cell and this this modules are connected through this plug and produce connectors which produce which actually provides
Power power um air and also also data to each to each of the modules so basically when we connect the cell together all these devices are powered and they can also be they can also start working together of course the um this sort of reconfigurable architecture Hardware architecture needs to be supported
Supported in software so here Ross is actually very good at this because it actually enforces modularity and in software so basically as we as we uh uh remove or add new devices into the into the cell they can be they can they can they they are interoperable immediately um so with this proposed
Design we were then able to to solve V various uh um disassembly problems so here you can see we’re also using actually some Soft Robotics in our cell so here you can see the soft hand to actually support grasping different objects with without knowing exactly what they are and uh basically the
Fiction is to help us actually to position the objects properly for uh to to ensure to ensure that we can actually do the accurate disassembly procedures and so on so the and the whole as we as we analyze the procedures then of course we need to also also reconfigure our
Programs based on what based on the state of each device so here you can see the example of the of the the first one was the example of the heat cost allocator here is the example of the of the smoke detectors so basically smoke detectors are quite a challenge so
Generators we can’t we couldn’t figure out any totally General procedure to actually disassemble this so besides having actually a CNC machine which is not very practical but sometimes you can do the screwing or some other operations as we you will see later so besides this General General cell Recon cell reconfiguration and
Adaptation we of course then also design different elements so we use for example Soft Robotics to to to generate appropriate picture for this assembly so here you can see for example an example of a compliant picture so some of these smoke detectors can actually be disassembled because if you can if you
Can get this rather comp rather rather accurate insertion procedure right but the the problem with the standard fixture would be that basically if you put the if you put the if you if you Gras the object to tightly then basically you cannot open it with the robot and basically the compliance in
The fixures then allows us to to allows us to to to to successfully lever out the upper part of the smoke detectors so another another tool we are looking at here is actually AI so this can be very helpful to actually put into the into the into the system some common
Sense knowledge so we basically have a vision pip P planine that generates some some um semantics in description first and plus the instruction of what should happen with uh with our disassembly procedure and this is then fed to the large language models which is of course
Trained with the data from the from the from the disassembly procedures that are used for example in repair and this then allows us to actually create some uh some prediction of what kind of action to apply in cases when we encounter the objects that that have not been considered yet
Before so um so besides uh so we need we need uh besides besides predicting the actions then of course such actions then need to be also adapted to be to be able to perform different operations so the control parameters needs to we change all the time every here you see an
Example of a Levering out operation which uh which is in some way similar for different hit cost allocators but uh of course due to the different gaps and different forces that needs to be applied to different objects this needs to be adapted all the time and
Online so and as a and as a final final procedure that we will looking at was also to besides having procedures that that are totally automatic that that are programmed and basically composed of known of known skills we also try to generate a totally automatic procedure
To find this assembly steps so this is possible for operations where we don’t need tools but we can but we can just uh we can we can just uh um apply applied robot motion to basically disassemble the object here you can see an example where the robot is actually trying to
Disassembly disassemble a car light so what happens here is that initially that the robot has some initial exploratory motions this motions explor we explo the we exploit here the compliance of the robot mechanisms to actually figure out in which direction the Motions are possible here we here we use the fact
That this assembly is much easier than assembly because basically you are you are Guided by the constraints of the objects you’re trying to disassemble so basically you only need to find out the correct correct actions that that are actually applicable you don’t need to explore the whole motor space of the
Robot so this this sort of procedure then basically allows us to to create automatically some sort of disassembly graph which is then basically basically um um um uh optimized by some sort of reinforcement learning procedure that I don’t have time to go into details here um so this is actually all for my
Presentation today so maybe what I would like to say if I go back to my original problems so which is disassembly of this electronic devices so in our project we basically get got to the stage where we can successfully disassemble different project different different devices and with a certain amount of adaptation but
Uh at the end of the day to actually get to the really practical system there needs to be done more about the design of these devices that uh that uh uh that need to be disass assembled basically um if the if the device if the disassembly
Steps are not considered in the in the design it is unlikely that we will never be able actually to generate generate really practical and economical recycling procedure so these are the contributors to this work and thank you for your attention great thanks Alish uh questions from the
Audience yeah we got one all the way down there hi thank you for the very interesting talk um yeah I was I was wondering uh so use now robotics mostly for for disassembly do you think that other parts of of the process can also be be
Um that can that can leverage the um the power of of robotics and do you think that um more advanced separation or less separation could be accommodated using let’s say smarter robotic techniques I mean in so you’re meaning there different recycling processes so yeah so along the the recycling process you have um
Separation steps sorting steps and so on uh do you think that robotics have an opportunity there in each part of the steps and and put allow for more advanced recycling or would uh would it be better to do a first separation step before um starting with more advanced
Robotics I mean it there to be totally general of course it depends on what you want to recycle and of course there are there are there are different procedures so sorting is something where robotics can help a lot and it’s a and it’s definitely an easier problem than this
Assembly so um so for for example if you uh since the batteries are covered you don’t know you don’t know if the battery is there or not so you can imagine you know having some sort of uh x-ray sensors like the output to actually detect the batteries and decide which uh
Which of the devices should actually can can go directly to the Crasher and which device should actually have some some other step steps that needs to be applied before you can uh before you can do the disassembly yeah um otherwise uh um I mean uh um um for electronics probably these two
Are actually the most the sorting and uh and the removal of dangerous components I would say at least from what I’ve seen are the most uh relevant uh relevant uh process is where BX can help but if you go to different other I mean the a lot
Of a lot of um a lot of products actually compare uh contain also valuable materials and there you can again use robotics of course to remove this valuable materials from the from the from the device before you start before you put it to the general recycling process procedure so this is
Another probably useful application yeah yeah thank you very much for the presentation and um some of us were at the European robotics Forum just two weeks ago and one of the recurring question was also the repurposing recycling uh of robots and also how uh you know recyclable are they how
Disposable are are they uh so you work also on electronic devices at large but have you ever thought of maybe um yeah working on robots themselves and how they could be dismantled recycled in all parts yeah um so unfortunately I can’t give you the answer that I know this
Already but we are trying to submit to get the proposal with the FMA actually getting through which would address exactly this topic so maybe wait for a few years before at least I can answer this question maybe there are other people who can already answer it you’ll come again uh to present your
Updates in two years um before we have always seen batteries or laptops and I never seen those two kind of electronic devices but it just shows like how many different electronic devices there are that we might even not think of and I was wondering did you uh had those in mind
With the project or was there actually someone from the industry from I don’t know producing those uh coming to you approaching you to actually ask for it like is it a a real application that is needed or was it more research based project it’s uh it’s let’s say 50/50 in
This Cas so we were actually working you know and uh uh um so not not my group but the partners in the project were work working with the electric cycling before on actually disassembly of some hard drives and basically they asked them what would be you know the next uh
The next step in recycling and So eventually it did come actually from the company that uh said that the removal of batteries from this small devices is a serious problem so it’s not uh it’s not our invention maybe uh I think maybe it’s not just a real real question maybe just a comment
Which uh come to my mind when you when you said it um the problem is probably when you or when I see this question as well like how to design for recyclability the the companies that approached you are not the companies that produced the the materials or the devices but the
Companies that have to recycle because they are the recycling companies I guess if it was the companies that produce those devices have to also like take them back and recycle them they would maybe start thinking a bit more about the design for recycling so I
Think you need to I mean I mean you really need the regulation probably first so I doubt that the market will force this actually so I mean unless there is regulation that forces the companies to actually recycle this products uh automatically they will not do it I would say
Yeah we still have some time so we get back to the slider question later we have one more question from the audience thank you uh congratulations on the project uh I’m curious if you have compared the cost uh of the manh hour and the person doing the job initially
And then with the robots what’s the difference and how do you find it if it’s applicable to the industry right um I mean I would like to say you know it’s not just uh it’s not just the cost here so this uh conditions uh conditions in the recycling plants are really
Unpleasant so this are this is really the jobs uh the job you don’t want to do so um you will have also troubles in actually getting people uh doing this sort of jobs so um yeah I’m I mean uh I if I am honest so I don’t think we got
You know to the stage where where we could seriously consider that our robot is uh is versatile enough you know it I mean of course cycle times are always a problem in industry but here it’s even more the adaptability so you really to be practical you really need to start
Being reliable enough on a on a significant amount of devices that uh uh that are disassembled and yeah unfortunately we didn’t get that far so but eventually yeah so then the next step would of course be the cycle times which needs to be which needs to be low
Enough and uh yeah and uh but uh as I said uh lack of Workforce also helps actually here why you want automation okay um thank you very much for for the very nice talk um not a question directly to you but maybe as a question
To to to have a small discussion at the very last speech of today’s Forum so everybody who has given talk today is always said that regulations you know policy makers and who are these regul regulations who who are these people who are actually putting the rules so that
The producers or manufacturers are also so kind of trying um to be um kind of following certain path so I had a small um research about are there some research groups who are working on the regulations and standardization there are some very small groups in the US as far as I’ve
Found but not in Europe maybe I couldn’t find do you have such a you do yeah there’s a lot of Civil Society like that microphone but not not the Civil Society group or civil groups but really research University or faculty research groups who are really putting efforts
And science behind it to to convince uh you know Regulators to to to put a foot on it I mean um as far as I know as far as I could find there are no regulations for robotic applications not in Europe so I me uh it’s great to hear that you found
Some in the US can you please go into detail about what they do maybe we can learn about it sorry so uh there are a few patterns uh that I could find from the US about um how the disassembly or assembly of those needs to be made and there are
Certain uh regulations I think uh according to the US law that is put uh I of course don’t remember anymore which number was it then Etc but I had zero input from the from the from the Europe so uh I think maybe that that’s a question mark that we need to keep in
Mind today to to to look for it if if one of us um could um I mean that that’s I think a research group that everyone would like to um you know make a collaboration maybe and and and uh start a a a research together comment on that topic
Yeah yeah but Ju Just to comment I mean patents don’t make regulation right patents certainly don’t make regulation and in Europe we have a lot of Regulation coming up right now the right to repair um standardization on power supplies things like that for robotics I think specifically it’s mostly about
Safety I think there is nothing about recyclability in the legal train of the European Union right now um but but for the cases where there is something coming up Electronics in general for example might be an interesting topic to lock into those processes are pretty open and as researchers we are often
Invited to cont contribute expert opinions or something like that on upcoming legislation and I think I mean I’ve not been involved in robotics but in other fields doing that and it’s something we researchers are kind of obliged to do and give our expertise I believe so there are some very general
Uh uh research groups or Publications in terms of mechanical uh engineering and generally electron Electronics but not specifically robotics you know what I mean I mean um maybe maybe uh maybe in the future that will will be done so I have read that there is a need for that but no one
Is really touching that group uh in terms of sustainability a lot of Ethics researchers are touching touching upon that they are saying that there are standardization is needed but that’s that’s it so maybe not specifically for recycling but uh in in the case of you know collaborative robotics the group of
Sami Hadad in Technical University of Munich is actually doing quite a lot of efforts to create the data that can actually generate the sensible regulations for collaborative robots you know this is because they’ve create they created this franka company that well I’ve seen a few frankers around here as
Well so they are actually actively working on putting uh um creating regulations that will make these robots that can actually work in contact with people uh more practical than according to the current regulation yeah just to just to understand better um because I know couple of researchers that are in the F
That are focusing more like on Law and digital let’s say digital technology in general and I know even couple of researchers that are really their focus is that so their background is in law but they’re actually doing really work on robotics so I couple of ERC like onec
Grantee that I know um my my my question is more on the I know in Europe you have this uh expert groups that the European commission is organizing and usually they try to come up with uh experts from Academia and Industry and I guess that intersection of people would be
According to me the most appropriate to probably to start something uh further I guess yeah um okay so uh maybe one thing to consider also is uh if you remember the slide from abish Gupta where he introduced the robot as a system of systems and uh so I’m I’m not sure if we
Really need because I always also thought about that um need something especially for robots or if robots are just electrical machines and uh you could sort them to be in the base a mechatronic product you could say it’s a cyber physical product but I think for those uh we have something coming up
With the right to repair and this everybody said so I’m not sure if it’s if it is a clear need for robots especially um if then we should do or work on it but uh for me robots would be just uh electrical machines or or applications that fall under the the
Guidelines I guess that will be installed in in the hopefully near future right we take that as a closing statement for this question and we have a new question um yeah thank you for your talk and I was wondering when you develop robots to dismantle e-w do you develop
Them in a way that they also easier to be dismantled or are they designed for reusability recyclability things like this I mean there are designing modular systems that can be easily changed with with the not consider the recyclability of the robots itself in this process you know I mean
The it’s uh I mean it’s not really it’s uh I mean the recycling of the robots for recycling of electronic devices are no different than other robots so I don’t think you know there’s any connection between the two if you’re working on recycling of robots then it’s for any robot regardless of the
Application so I don’t see you know here and anything specific on my robots that uh build so if you if I had a if I had a procedure to to recycle this robot I could probably recycle other robots as well um and do you have numbers on how
Much ew is recycled to what extent or how many robots are recycled in the last year no I yeah I don’t for E I do have some numbers but I don’t have it in my head sorry for the robots I don’t know yeah thank you U thank you very much for your
Interesting presentation and the Insight uh on the very interesting project I was just wondering um you showed a very nice example for dismantling the smoke detectors right um how difficult is it um to teach a robot to dismantle different kind of smoke detectors because like for example for humans it’s
More or less straightforward if they can’t find one way to dismantle different kind of smoke detectives they’ll try um is that also possible with the robots uh that you are developing for to test to dismantle different kind of smoke detectors I mean you need to basically identify the basic procedures that you
That you use for this manle which is primarily there are primarily two steps one is the opening of the cover so how to actually you know make the battery visible and the second is how to remove the battery itself and then you know and then there are options are there screws
Are there no screws is there some opening where you could apply Levering is there no opening and so on and uh then when you go to the batteries you have all sorts of shapes of batteries some are soldiers some are not and yeah so all this basically
Affects your pro your procedures and uh once you have enough of these procedures uh implemented then uh the next step is basically to take care of the adaptation of these procedures and the final step is then this prediction of actions so you know when you get to the device that
You actually never seen before and this will eventually happen yeah thank you I think that was all question from the audience I have one last question before we get to the slid one we answered I think we answered it yeah but uh we can we can double check um I actually I
Didn’t know that there are products that are not meant to be recycled or opened up so that was that was something meant to be not to be recy the idea is not to be open not to be open but it makes sense for smoke detectors for example I
Know that a big challenge for recycling electronic uh devices is that there’s no data so there’s a lot of different models versions devices and uh you could probably recycle these devices much more efficiently if you had the data on you mentioned if I had the procedure to uh
Recycle a robot I could probably do it the problem is often that there’s no procedure there’s no data because there’s so many uh manufacturers and so many devices uh there uh um have you experienced that and uh how how can could that be solved I mean there is
Some data so we were trying to use the um the disassembly procedures from my fixed site where where you actually have some a number of these disassembly procedures and we were using this you know to retrain this uh large language models to try to get something to something that is competent for
For our limited domain yeah but I mean you know the more specific if you had a more specific data set that deals with disassembly of the specific devices this would definitely be useful right but it’s not just the it’s not just the procedure it’s also the images of the
Devices you know how how they look like during the procedure yes yes so it’s not maybe talking about regulations and not only the uh the information and the design that for recycling but also the information that comes with along with it the data the pictures how to recycle
It in the end I can just agree with this yeah um yep uh then um the end question yeah you partly answered it already will the design for recycling happen uh what is needed from whom um my question also when can we expect it uh this is this is
Work in the lab now uh is it is it something that we can that you’ll take in your project out to the to the factories to the end users um yeah I mean you know there’s uh I really think you know that we need this uh that this uh this
Assembly uh design for this assembly needs to happen before before we can make uh commercial procedures for this I don’t think uh it will be practical before actually this devices designed a little bit differently right right so if we have the question what is needed from whom it’s collaboration from All actors in
The process that is needed here and that’s something quite challenging we noticed today in general so great then uh thank you very much let’s give a clap once more um ah yes and uh closing remarks the St so we are at the I use this or I use
This this one okay so we are at the end of another very interesting day of with several seminars with several interesting discussion because I have the feeling that we could have gone on for another hour with question and answer actually I refrained myself to make other two or three statement that
Would have let us staying here for the rest of the evening uh but the day is not over yet and actually at least for those that are physically here for the people online I’m sorry uh but it’s harder to join but we are going together to also enjoy a an evening dinner
Together and probably continue the discussion because this is what we want Bri to be place where we arrive with ideas where we discuss ideas and where new ideas are born so it was a really a pleasure to have you here and uh we look forward to have most of you in the next
Few uh events just as a reminder we have every Thursdays our seminars and we still have other two major events that are coming one in the end of April and the other at the very beginning of May uh so again good afternoon and uh goodbye to the people online and for the
Others let’s continue our discussions I will bug several of you and uh that’s it for [Applause] today e e e