
A household robot that vacuums, unloads the dishwasher, mows the lawn, and even drills a hole in your wall? It may sound like science fiction, but according to Professor Bram Vanderborght (VUB research centre BruBotics), that future is fast approaching. âWith mass production, robots could become as affordable as an average car.â
The research of Vanderborght and his team is already leading to concrete applications. VUB spin-off Axiles Bionics, for example, develops intelligent ankle-foot prostheses that combine robotics, artificial intelligence, and human biomechanics. The company has been selected for the federal campaign Belgium. Open for Business and will showcase its technology at the World Expo in Osaka in 2025. At the VUB AI Experience Centre in Etterbeek, research also meets entrepreneurship and society: a high-tech test and demo space where companies, researchers, and the wider public collaborate on AI and robotics.

Bram Vanderborght
Is Vanderborght right, or simply too optimistic? Big names such as Mark Zuckerberg, Elon Musk and James Dyson are already investing in smart humanoid robots. Yet many tasks remain a dream for the future. When Vanderborght recently hired a digging machine, it refused to work. âI felt it in my back. (laughs) Robots wonât be taking over those kinds of jobs anytime soon. A lot of technical work requires far more dexterity than youâd think. Skilled workers need not worry: much work will always remain human work.â
What can humans do so well?
Bram Vanderborght: âHumans have senses, brains, and muscles to observe, explore, and manipulate our surroundings. Robots need to do the same, but with sensors, processors, and motors. Intelligence is not just in our brains, but also in our bodies â what we call embodied intelligence. Our bodies evolved to do things âwithout thinkingâ, like walking through a forest or picking up a stick. Millions of years of evolution fine-tuned that ability. The challenge is designing robots that can imitate even these seemingly trivial actions as well as possible.â
For that processing, are AI algorithms used more and more?
âNot exclusively. The famous dancing robots from Boston Dynamics (Do You Love Me? on YouTube) donât actually use artificial intelligence. They rely on models based on Newtonâs classical laws of motion â describing the relationship between force and movement. From this, you can precisely calculate the forces needed to spin, jump, or grab an object. Thatâs called model-based control.
Outside the lab, this often isnât enough, because the real world is complex and unpredictable. Since we donât have models for everything, AI steps in. The best robots will combine both: a model-based foundation to prevent unsafe behaviour, such as colliding with humans, and an AI layer to learn new tasks.â
âIf robots are to walk among humans, safety will be the number one priorityâ
Is this comparable to ChatGPT-style AI?
âChatGPT is a large language model trained on vast amounts of text. That works well because the internet is overflowing with text. For human-robot communication, ChatGPT is relevant. Work is now expanding to images and videos, which require enormous amounts of data and computing power. Robot control, however, needs even more sensor data â and we rarely have massive datasets.
Humans donât need huge data volumes to learn. Children learn astonishingly quickly. Yesterday, my son saw a hot-air balloon shaped like Spongebob on Karrewiet. Thanks to the context, he immediately recognised it as a balloon. AI might have just seen the cartoon character.â

Massive data and computing power â that doesnât sound very sustainable?
âThat will indeed be a challenge. Computers controlling self-driving cars already process huge amounts of data and therefore consume enormous energy. Humanoid robots will have hundreds of sensors, so computing power will explode. Weâll need to develop new chips, sensors, hardware and software to cut down energy consumption.â
What other hurdles remain?
âSafety is the number one challenge. Teaching a computer to play chess or running ChatGPT happens in a virtual world. Mistakes donât hurt anyone. With robots itâs different. They share our physical space. That makes safety an absolute priority. Robots must not harm people, their surroundings, or themselves.â
Like self-driving cars, which still sometimes cause accidents. Are they robots too?
âA modern car isnât fully autonomous â the driver still needs to stay alert. But with its sensors it can detect the environment and potential risks. In that sense, yes, itâs a robot. The same goes for drones. If you command one to fly forward but it sees a wall, it wonât execute the command. A model aeroplane, by contrast, will crash if you steer it into something â thatâs not a robot.â
âIn care homes, robots could deliver meals and medicationâ

What do we still need to invent for robots to perform real-world tasks?
âTouch sensors. Seeing, hearing, speaking⊠those are less of a problem. We can borrow technologies from other sectors like gaming or telecom. But touch â true physical interaction â is another matter. Thatâs something robotics itself must invent.
Skin is our largest organ, but for robots itâs currently little more than a plastic shell. We can hold a pen, rotate it, let it slide between our fingers⊠Robots need to master that too. Theyâll need sensors that measure not just pressure but also friction and shear. Belgian company Melexis is developing such 3D force sensors for robots â weâre collaborating to bring them into robotics. With Imec weâre developing a new sensor skin to measure distances, so robots can work faster yet safely with humans.â
So it sounds like a long road ahead?
âIn 2023 many new humanoid companies were founded. Weâre now at the start of the Gartner Hype Cycle, the curve every new technology follows. First comes buzz. Then hype with inflated expectations. After the peak comes the âtrough of disillusionmentâ: the tech isnât as revolutionary as hoped, and interest dips. Next, realistic applications emerge, leading to practical implementation. In the end, the technology often outperforms expectations. We tend to overestimate the near future and underestimate the distant future. That was true for the mobile phone and AI. It will be true for self-driving cars and robots too.â
Where do you see robots working in that distant future?
âIn healthcare, hospitality, agriculture, horticulture⊠anywhere thereâs a shortage of staff. The key is deciding what to automate and what not. In a care home, for instance, a robot could move patients, deliver meals, and bring medication. That frees up caregivers for human contact.â
âRobots wonât steal jobs, theyâll work with usâ
So they wonât take away jobs?
âWeâll collaborate. Robots are precise and tireless. Humans are creative and versatile. We complement each other.â
Will everyone be able to afford one?
âToday a humanoid robot is like a Ferrari â hand-built and very expensive. But with mass production, theyâll become as affordable as a regular car. And if you think about all the technology that goes into a car, thatâs cheap.â
Are Flanders and Europe ready for the robot revolution?
âResearch receives strong support, but the leap to startups and businesses is huge, because developing hardware costs so much. Europe still has big robotics companies like ABB, Siemens, and Kuka â though the latter is now largely in Chinese hands. We must make the right choices now. Economically and geopolitically itâs vital to keep both design and production here. Elon Musk says: design is overrated, production underrated. His real innovation wasnât the Tesla, but how it was built at scale for an affordable price. Letâs not repeat past mistakes, where knowledge is built up here, but profits and monopolies end up in the US or China.â

What role does VUB play in all this?
âOur research group BruBotics coordinates several European research projects and launches spin-offs, hoping to spark entrepreneurship. At the same time, we aim to contribute to the many dimensions of the UN Sustainable Development Goals. It remains a huge challenge.â
VUB AugmentX lab for human-robot collaboration
In early 2025, the Vrije Universiteit Brussel, in collaboration with Flanders Make, opened the AugmentX lab: an advanced research centre dedicated to optimising human-robot collaboration in the workplace. Led by the BruBotics research group, the lab offers innovative tools and methods for analysing workplaces and refining the division of tasks between operators and cobots. AugmentX also investigates how new technologies can alleviate physical workload and makes that impact measurable.
Interview with AugmentX coordinator Tom Turcksin and Professor Tom Verstraten (BruBotics)
A peek inside the labs of VUB scientists?
Join us on 23 September 2025 for the Academic Opening and explore the renewed VUB laboratories
On 23 September, VUB will not only open the new academic year but also the doors of its laboratories. Discover live the technological innovations and scientific breakthroughs our researchers are working on â the very ones that could soon make the headlines.