Will humanoid robots care for elderly and lonely seniors? Are we currently experiencing a robotic revolution? And isn't humanizing robots and artificial intelligence “playing God”? We discussed these questions in the Laboratory of Bioinspired Robotics with its head, Stanislav Věchet, from the Institute of Solid Mechanics, Mechatronics and Biomechanics.
We are sitting in the relatively newly established Laboratory of Bioinspired Robotics, and logically, there are robots and little robots all around. Can you introduce them?
This little one is called P.A.W.O., like the constellation Pavo. I decided to name robots after constellations. Pavo used to be seen as a mythical creature. P.A.W.O. can walk, wiggle, respond to touch – it's a fun companion for social interaction. For example, my kids love visiting it here.
Next to it is the under-construction humanoid Kentaur. It’s supposed to be about 1.3 meters tall and able to carry a crate. Originally, I wanted it to serve drinks and greet guests at the Night of Scientists, but we’ll see if I manage that.(laughs)
The lab and your department are relatively new at the institute, right?
Yes, about a year ago, a new fifth department was established at our institute, called the Department of Autonomous Systems, along with this lab. For now, the department is quite “intimate,” but we’re gradually getting it going. I teach a course called Mobile Robots, which had sixty students this year. And in the lab, I try to work on demonstrations of what today’s robots can do. We’re planning to buy a four-legged “robotic dog” from Unitree. I’d like the lab to serve as a kind of showroom for mobile robotics.
Marketing cookies must be enabled to play the video. You may change the setting here. You can open the video in a new panel. Let’s go back to P.A.W.O., which strongly resembles robots from animated films. Was that an inspiration?
Exactly, it’s my version of Disney’s BDX robot. It’s the same size and weight, I just don’t have Disney’s budget, so my robot has fewer degrees of freedom. It’s an example of so-calledoff-the-shelfrobotics, meaning it’s made from parts you can buy at a hardware store, the servos are for model-making, and the rest is 3D printed. My vision is for it to be an open-source project, so once it’s finished, other schools and students can build it and experiment with it. Because most robots we see online are either extremely expensive or not available for purchase.
Your next – still under construction – robot is supposed to be humanoid. Lately, it seems like humanoid robots are everywhere online. Where does this fascination come from?
I’m fascinated because I want to understand how we work. I realize how incredibly complex motor skills are – all the movements we make when holding a phone, writing… Or walking – the BDX robot walks so beautifully that I can’t stop watching it, and I just want to recreate that.
The unfinished Kentaur in the lab (left) and a visualization of the completed robot (right). | Author: S. Věchet archive Isn’t creating human-like robots a kind of “playing God”? Proving we can create something like a human?
In my opinion, it’s a fascination with our own likeness, just like with AI, which we try to personify and talk to like a friend. And yet it doesn’t make sense.
I don’t do it because I want to play God. I’m fascinated by solving the problem, and I think others who build robots feel the same. But to sell it, they sensationalize it by adding a human form factor. Often, it doesn’t make sense – I don’t need a humanoid to drive a car for me; it’s much more efficient to make an autonomous vehicle. I don’t need a robot that’s 180 cm tall and walks on two legs – 120 cm is enough if it serves its purpose. There’s no need to copy humans at all costs.
So the feeling that we’re living in a revolutionary time is just an illusion?
There’s no revolution happening – it’s a gradual evolution. And if we’re talking about humanoids, they’re cool, but I think it’s still too early. Many people don’t realize we still can’t properly assign them tasks. It’s nice that a robot can walk or run, but you can’t tell it to wash the dishes. For example, I think developing fine motor skills will take a long time. From a technical standpoint, human-like form is often a complication: imagine an industrial space, a hall with a flat floor, and a humanoid robot walking on two legs carrying boxes – but why? It’s much more efficient if it has a humanoid upper body and a Segway-like lower part. That’s why companies often show only the upper half of the robot – because running and doing tasks at the same time would consume too much energy.
There’s not only fascination with robots in society but also a certain fear of technological development. How do you perceive that?
I don’t share that fear. As I said, there’s no revolution. It’s a temporary hype that runs in cycles and has happened many times before. “Revolutions” are mostly created by sensationalists. When you see a robotics revolution in the media, chances are it’s just marketing – at least that’s how I see it.
And it’s not just robotics – even with AI, I feel like most people only know ChatGPT. But it’s basically Google on steroids. When Google came out, people were amazed at how it worked. Today, it’s normal. The hype always cools down, and the field continues with those who were already working in it… until the next supposed sensation comes along, and so on.
How do you build a robot “from scratch”?
You need motivation, otherwise you’ll get bored quickly. But I’ve been hooked since I saw the movie *Short Circuit*, so I’ve built a lot of mobile robots.
The beginning is always the same: pencil, paper, sit down, draw, and calculate. Then I made the first designs, tried printing parts on a 3D printer – of course, they didn’t work and fell apart. So second try, third… I’m on about the fifth version of P.A.W.O. now, and it finally works. You can never create a perfectly accurate model in simulations – 3D printed parts are very flexible and behave differently in reality. So it’s constant iteration, and you have to be passionate.
Robot P.A.W.O. | Author: S. Věchet archive What will Kentaur – the mentioned humanoid you’re building – look like when finished?
It kind of resembles a centaur, hence the name – it will be a combination of a bipedal and quadrupedal robot. Four legs can carry more, and you can load extra cargo on its back. I’d like it to carry at least 100 kg. It will be smaller than a human and its legs are divided into three parts instead of two, which works better for the drives. I think the first applications of these robots will be for carrying loads or monitoring environments that are too dangerous for humans.
From time to time, the media report on humanoid robots being developed as companions for elderly or lonely people. Do you think that will happen?
On one hand, there will probably be such a tendency, because humanoid robots will be more pleasant than a mobile monitor. On the other hand, I wouldn’t let a robot take care of my grandmother instead of me. A humanoid should be a service tool – we should use robotics to help us, not to replace us. I understand that some people might think it’s a good idea, but I can’t imagine fully replacing human social interaction.
Last question – a bit personal: how many robots do you have at home?
Here the saying applies: the shoemaker’s children go barefoot(laughs). I don’t want any unnecessary electronics at home – no smart house, not even a robotic vacuum. It’s about balance – I need to rest from it at home.
