logo

Maakt de mens plaats voor robotica?

AI en robotica hebben soms een negatieve bijklank, alsof het zaken zijn waar we schrik voor hebben. Waarom dat niet nodig is en we het volledig moeten omarmen in de juiste samenwerking, hoor je in deze aflevering.

Beluister deze aflevering van 'Scientists With A Cause' (in het Engels) op Spotify, Apple Podcasts of Google Podcasts.

Wetenschappers

Professoren An Jacobs en Bram Vanderborght zijn experten in sociologie, communicatiewetenschappen, AI en robotica. 

Wetenschapper in de sociologie en media-innovatie

An Jacobs

Wetenschapper in AI en robotica

Bram Vanderborght

Transcriptie

Bert: [00:00:11] Scientists With A Cause is a podcast in which you discover how and what VUB scientists contribute to the Sustainable Development Goals. Those are 17 goals agreed upon by the United Nations as a part of sustainable development and development coordination. How do they work? What goes on inside VUB? The scientists take their turn in this podcast to tell you about their cause.

An: [00:00:35] In our project that we scope, the focus is on how can we enhance humans without being blind for the potential negative consequences that are maybe by-effects.

Bert: [00:00:47] An Jacobs is a sociologist by education and has been working for VBE since 2005. With her team, she focuses on creating technologies that are society proof to fulfill our needs. Bram Vanderborght is [00:01:00] professor in robotics at VUB with a strong vision for collaboration between humans and robots. Instead of robots being a replacement.

Bram: [00:01:08] There is a lot of dystopia, a lot of things that can go wrong. AI systems that can learn everything. We need to give a good expectation management. What are the possibilities of robots and AI? And a lot which we see in movies are not.

Bert: [00:01:26] This episode puts the focus on the SDG industry, innovation and infrastructure.

An: [00:01:32] People are very afraid of losing their individuality and being controlled. But that's not what robotics and AI, they are very good in specific tasks in a rather controlled environment.

Bram: [00:01:44] And we hope that there is a vision where robots and AI and people work in synergy.

Bert: [00:01:49] Discussing the world of innovation, most of all AI and robotics and how they can help us move forward. In this episode of Scientists With the Cast.

Bram: [00:02:05] We [00:02:00] are in the AI Experience Center. It's a center where different research groups on robotics and AI collaborate and demonstrate their technology to work, especially companies like exoskeletons, prostheses, how their smart grids can improve energy problem, interactive robots, but also the ethical and social aspects of it.

An: [00:02:29] And the goal is really to give people an embodied experience of what AI and robotics can be in the future by bringing from the different research teams, different demonstrations of the advancements they are making, or also explaining in a very easy, accessible way what AI is really like. So it's easy to understand, like the basic principles of the AI that is used and developed.

Bert: [00:02:53] How often are you here in the center?

Bram: [00:02:55] I'm quite often here because my office is very close by. Moreover, we're producing [00:03:00] a new human augmentation lab funded by Flanders Make, in which we measure the impact of exoskeletons, cobots, and other human augmentation technologies on the impact on the human body and the worker. And so a lot of our researchers are actually also performing their research in the center. So there is always something new to see here.

Bert: [00:03:23] A very lively center. That's a good sign as well I think. There have been a couple of terms said already. I'd like to especially focus on AI and robotics. This place relates to anything AI and robotics. But what is the difference between those two?

Bram: [00:03:38] Yeah, for the public opinion, often robotics is equal to artificial intelligence, but we see more difference that a robotic is a physical entity working in the environment. So it needs sensors to sense the environment, it has a brain to think and actuators to manipulate that environment. And [00:04:00] in order to think, it often uses AI techniques. In AI techniques are algorithms that are able to learn from new experiences and from data it collects in the environment. But not all robots use AI techniques. There are also other techniques, and not all AI techniques do need a robot body in order to operate. Many AI techniques are also in your computer, in your smartphone, in the cloud. So they don't have a robot body in the environment. But of course, the two fields are growing more and more to each other and need each other. And that's also why we decided on the VUB to closely work together. So I'm a robotics scientist, but the AI researchers are in I-lab, but there is a very strong connection and so they are also unified in the AI Experience Center.

An: [00:04:49] And it's also helping because yeah, it's very challenging. We call it wicked problems. So they are very challenging to tackle and so you really need a holistic view. And in science [00:05:00] we are for many years, siloed in our own disciplines and working, really collaborating really together is difficult, but that's also with robotics, what we try to achieve. Also, they're having a place where you can also meet each other or experiment with each other. It's really interesting. Like we will experiment with a workshop that we created about letting people understand fairness notions. You can in AI implement fairness notions about is this technology fair in deciding who can get the job, for example, those are classical examples and we know there is a lot of bias creeping in actually those AI systems and mathematically there are ways to solve that. But of course the reasoning behind it, what is fair and what is not is a social aspect. So it's normative, it's culturally based, and to let people reflect on that, that's also what we are going to experiment with here in the AI Center.

Bert: [00:05:57] So we can see a lot of different things here in the Experience [00:06:00] Center, of course, a lot of robotics. But there's also a very specific book, Homo Roboticus 30 Questions and Answers on Man Technology, Science and Arts.

An: [00:06:10] Yes, it was the final point, the end point, actually, of a project that we did. And the goal was to have different departments from all over the university, all faculties. We invited colleagues to think about robotics and what the impact from their discipline actually was, but in a very accessible way. We first listed like what are different questions that people ask about robotics and in that way help them also to reflect from on that domain of robotics from their perspective.

Bram: [00:06:49] So the central line in the book is how we can keep the humanistic values of solidarity, equality and freedom, send ground in robotics world. And so the book is organized [00:07:00] around 30 chapters, which every time kind of provoking a question. So like, do robots need to pay taxes, for example.

Bert: [00:07:08] It's a good question.

Bram: [00:07:09] And then 56 academics work together to write those chapters. And at the end, we have a proposal for an inclusive robot agenda. So what are the measurements we need to take in order to start the debate how we can realize that synergetic approach between humans and robots?

Bert: [00:07:35] So now we're caught up with what AI and what robotics are and most importantly, the difference between the two and how they interact with each other. As An and Bram said, they try to bring a lot of different faculties and people together because that multidisciplinary collaboration can really make a difference.

An: [00:07:52] I meet a lot of people who also find it enriching because you never could find that kind of solution just on yourself. [00:08:00] But it's also sometimes difficult because you talk different languages, you have different goals also, things that you want to achieve. So you also really need to do a lot of things together to understand also each other better.

Bram: [00:08:14] Yeah. And it's also a challenge, especially for starting researchers because there are different expectations. But I think as soon as the collaboration starts and you give time to let it grow, then I think the interface between the different disciplines can create really the novelties and very important insights.

Bert: [00:08:32] Just to come back to AI for a second, because we started talking about AI. What are some misconceptions surrounding AI? Because, as you said, public can sometimes think robotics is AI and vice versa. Are there other misconceptions about AI?

Bram: [00:08:46] I think around AI, but also robotics. There is a lot of dystopia which a lot of things that can go wrong, AI systems that can learn everything, that even [00:09:00] maybe it's something can have a consciousness and so on. On the other hand, sometimes I think the possibilities of AI and robotics are under estimated, and I think we need to give a good expectation management. What are the possibilities of robots and AI? And a lot which we see in movies are not very far.

An: [00:09:25] The idea that we will be overloaded by all this technology. That's the dystopian and also that we will lose our autonomy as human beings. And especially in the northern west, people are very afraid of losing their individuality and being controlled. But that's not what robotics and AI, they are very good in specific tasks in a rather controlled environment. But there they can perform really great and can also give new insights that we can't yet imagine ourselves. So they really can be tools. But then we also [00:10:00] need a framework of regulation. And also at the European level, there is a lot going on in creating this kind of regulatory framework to also set the boundaries like what is okay to do with it and what not. And I think in that space a lot of innovation is still needed to gain. Like Bram just said, like this, this power that it's in the type of tools that we are making.

Bram: [00:10:24] Yeah. And in contrast to, for example, a hurricane or Corona, we create technology. So we need to do it in a very thoughtful interdisciplinary because sometimes we think we create new advantages, but we forget what are the disadvantages that it can create. And sometimes we tend to overestimate in the short future, but underestimate it on the long future.

Bert: [00:10:50] It's a weird contradiction to hear the difference between short and long future. How advanced in robotics are we as a society right now?

Bram: [00:10:58] Already rather well advanced. [00:11:00] For example, Belgium is in the top ten of most robotised countries in the world. I think they sneak in our society, now a lot of people have grass mowing robots or vacuum cleaning robots and they're not regarded as robots anymore. The same with AI. We rely so much on AI technologies, but because they are so part of our everyday life, we don't realize anymore what are behind computer programs that we use.

An: [00:11:26] Yes indeed and research that we do on our computer is AI. The hits that we get is AI, the news that we consume on social media and what we read. And so actually it's in every aspect already of our lives, but indeed we also do not recognize it anymore. And I think sometimes it's also really important that we also create that awareness, like where is AI in and what are the consequences? Every technology always has positive [00:12:00] and negative contributions, and if you don't look at both at the same time and the frictions between what you find important, like ecology and climate change versus equality on a gender level doesn't need to be in contradiction, but sometimes there is some prioritization to be made and and how to decide that, that's still a huge question.

Bert: [00:12:38] AI is everywhere around us, but not in the fatalistic or dystopian way that some might fear. At the same time, it's more than just some technological way to do cool stuff or to optimize certain work. It extends into the culture of our society, into questions of morality, and inevitably it touches on the subject of our data.

An: [00:12:59] If you talk [00:13:00] AI and robotics, talking about the data and how to collect and govern that data, that sounds very boring at first, but actually it's really crucial to think that through like, should I be the owner of my own data? How can I be involved in that? How much do I want to be involved with that? But we're still in the process of doing that. And the danger is, of course, if we are going to regulate it too strictly immediately, then we also can't experiment and we can't test out. So we need to also have a balance between that and that's a difficult balance.

Bert: [00:13:36] I was just going to say a very difficult balance I think.

Bram: [00:13:38] And also don't forget that robotics in Europe is very strong and one third of the industrial robots are produced in Europe. And if you look to the much smaller, of course, service robotics market, it's even two thirds. We also said that Belgium has a high robot density, so the amount of robots per 10,000 workers, but [00:14:00] if you include there, that of course our wages are much higher. So our reasons to go to robotic nations are also higher. If we take that into account, then the Asian markets are much stronger. And then of course, we are in an economic disadvantage. So we also need robotisation in order to, we think, bring back production to Europe. We need local production, which is also much more sustainable, but that we can only do with a high level of robotisation to go to mass customization. But at the cost of mass production is that you have individualized shoe or a car but at reasonable cost. Then you need the high level of automatisation to reach that.

An: [00:14:45] But then you also think normally immediately like what with the people that are working now there, what kind of jobs, what's the impact on jobs? And I think everybody should be aware that every [00:15:00] job, but really every job, we will have AI automatisation of certain tasks with AI. And sometimes people are like, oh, not my job, but really the job is different tasks. So everybody will have this transformation and adaptation that they will need to do in their jobs. But that's not, per se a bad thing.

Bert: [00:15:29] Robotics and AI will not ruin our jobs and rule our planet, according to An Jacobs and Bram Vanderborght. It is quite the contrary. They will help us as a society, as a company, as a worker to facilitate and possibly transform what we do for the better. That's what they are aiming for. And in that goal, sustainability is essential.

Bram: [00:15:51] It's very important that we continue to investigate that because robots and AI can be a huge enabler to reach that, [00:16:00] but can also be a huge inhibitor. Something that can be an advantage can also be used as a disadvantage. Also on a more policy level. Robots can help a lot to do jobs better, but who takes advantage of the profits that are generated by those robots?

An: [00:16:17] Well, I also think, like, for example, if you narrow down sustainability to climate and sustainable use of resources, there's still a huge way to go because until now, that was not a requirement on the list, like it's your prototype helping to reduce energy consumption. But you see this shift also by, for example, European Commission also asking that there is care for it, like there has been attention, more attention to ethics because the Commission is asking all our projects to take that into account. It's a process to be looking at those factors on the long term from the very early stage, I think.

Bram: [00:16:59] Yeah. And that's [00:17:00] what we investigate a lot because we develop wearable robots like prostheses for amputees or exoskeletons. So energy efficiency is crucial because you don't want to wear a huge pack of batteries. And so we've put a lot of focus on energy efficient activation, but now we also do sustainable robots because those robots are very complex, but you can break them quite easily when they fall and you need to over dimension them. And that's why we are investigating self-healing materials. Imagine that the system is able to repair itself. The materials are also now fully recyclable. We are investigating bio-based resources and biodegradability. So even if it's lost in the environment that the system can biodegrade, for example. So also the design in the construction of the robots, how can we make it sustainable? It's simple development.

An: [00:17:59] And you can also imagine [00:18:00] like and there are also projects on that to use AI, for example, to do smarter maintenance. Because if you do like prevention of things going down and you repair them earlier, they can be also more sustainable. So it's also that thinking that you can actually also use AI to actually create this benefit. Of course, not only and I think that's always an and and story. What can we do as humans do to reduce the resources and the energy that we're using? And in addition, how can we actually change the way that we work, change the way that we process things and make use of the tools that we are creating.

Bert: [00:18:42] An and and story, and thus a difficult story with a lot of different sides to it, but also a lot of exciting things. For example, the exploration of the senses in robotics.

Bram: [00:18:53] It's one thing to add a camera and it sees a lot of pixels, but it's another thing to make sense of [00:19:00] what the robot sees. So we have a department ETRO that works on cameras and to make awareness of what the system can see. So we're incorporating that also in our robot systems. Of course, there is also audio speech recognition, but that is things that we can borrow from the computer. What is often underestimated is our skin, the ability to feel touch and interactions, and that's for us very important in order to manipulate that environment. If I want to manufacture a part, we as humans can almost do it blindly because we feel the objects that we touch and we manipulate. And that is still largely underrepresented in robotics. And that's why we also collaborate with Imec to go to a new generation of sensors that will be able to feel and to better manipulate objects because our skin is the largest organ but almost non-existent in robotics. [00:20:00]

An: [00:20:00] And also the other way around because as humans we want to interact with those robots and they are not caged anymore. We can touch them. And also how do we want them to feel like? If you think, for example, about loneliness, companionship, there are robots currently also commercially available that try to fulfill that need. How do they need to feel? How they need to smell? How do need to sound? Actually, to give us also that feeling of comfort. So also understanding like what our expectations of this robotics, the form and the way that we can interact with them. Like for example, we did a project on collaborative robots in industry with Audi and there it was of course, it's a very loud environment. You can't do anything with voice there. So how can we imagine, like in the future, using gestures, for example, instead of buttons to interact with those machines and what's feasible, what's understandable, and what's also making work more [00:21:00] agreeable, but keeping in mind that it needs to be performative and efficient.

Bert: [00:21:04] That interaction between humans and robotics, what you are working on, it's so interesting because it's very unique and because in a way it seems maybe even limitless in what we can try to achieve and create.

Bram: [00:21:18] Yeah. So that's why first collaboration is so important because the challenges are so broad and there is still so much to develop that we need each other in order to do so. So that's why there is, of course, a bit of competition. But I think the challenges are not within Belgium between different research groups or even universities or even  in Europe. So that's what I also like on the funding scheme of Europe, that it's a very collaborative front because I think the challenges are to improve work and health and we need everyone because the possibilities and the challenges are very big. So we need each other to work on [00:22:00] those aspects because the number of applications and technologies are so big. Our human body is so complex, it's so amazingly engineered, maybe over millions of years of evolution and robotics is not yet that old. It's a bit I mean, after the Second World War, so it's still very young. And to make that same evolution, because we need capabilities of robots in order to assist humans which are at least similar physical but also cognitive performances as a human in order to assist. And so that's still an enormous challenge. And that also makes that I don't think that humans are so easily to be replaced.

An: [00:22:45] In our projects that we scope, the focus is on how can we enhance humans without being blind for the potential negative consequences that are maybe by-effects that we also take into account. But how can we enhance [00:23:00] them, keep them actually at the center, making tools with the technology in that collaboration but not diminishing them, disempower them. That's actually always our robotics point of view.

Bram: [00:23:13] And that's why we want to have excellence in our research field. But we want to be broad enough in order to connect to other disciplines. And so we try in our research groups to be excellent in what we do, but also be able to reach out to the other disciplines.

Bert: [00:23:39] Bram Vanderborght en An Jacobs, what are your expectations or what is your hope in your field of work for the next couple of years?

Bram: [00:23:44] Our hope is that robots will have indeed a role in our society and the number of robot companies we still need to increase. And that's why we also from robotics, we try to develop new robot companies. And the first one is [00:24:00] Exelis Bionics, which will bring on the market a bionic prosthesis to help amputees to have a better mobility. But also, we are trying to go for an exoskeleton company to improve the working conditions of workers on the long term. For example, we want to go for a company that goes to commercialization of those self-healing, soft robots, which can go also to applications beyond robots. So we try to go from very fundamental research on all the disciplines to more application driven and also to valorization how we can work with companies in Flanders, Belgium and Europe. And if they are not there, how we can found their own companies. And we hope that there is a vision where robots and AI and people work in synergy.

An: [00:24:54] Yeah. That we do not increase the digital divide. It's not one divide but exclusion. [00:25:00] And in general, like people that can have better health care, better kind of services because they have access to AI and robotics and people that are left out. That's really a huge challenge and we need to be aware about that and also think about that. It will be not one size fits all that we really need to investigate, like how do we keep everybody also in this trajectory so that they also can have benefits of this, this new evolution and this awareness, this awareness about the limitations, the choices that we have, the agency that we still have as humans, how that we can empower people on that. But of course, I also imagine, like within five years, that we have exoskeletons that are sleek, that are intelligent. So that can help me with heavy lifting, for example, in hospitals and in home care, because we have this aging population and the huge challenges ahead to [00:26:00] work with less people and do and give us the same kind of societal revenue.

Bert: [00:26:23] This was Scientists With A Cause, a podcast series of VUB, created by The Podcast Planet.