Women in Engineering Day: UCLB’s Business Manager Physical Sciences & Engineering Beata Grzyb talks AI, robots and Argentine tango dancing

UCLB’s Business Manager Physical Sciences & Engineering Beata Grzyb discusses her international background working on neural networks and robotics. 

 

What inspired you to pursue a career in AI and robotics? 

My journey into AI and robotics began during my computer science studies, where I became captivated by how the brain processes information. I was especially drawn to artificial neural networks, but rather than the typical models used in mainstream machine learning, I focused on spiking neural networks—more biologically plausible systems that emulate the way real neurons communicate through electrical spikes. This led me to take part in several research projects, including modelling the behaviour of single neurons and later developing systems that used networks of biological neurons to interpret emotional expressions from facial cues. That early immersion in brain-inspired AI shaped my long-standing interest in building more human-like intelligence in machines. 

 

You’ve had a global research career in Japan, Spain, US and Netherlands – how has that international experience helped with your work in robotics? 

My international research journey has shaped my thinking in robotics in profound ways. I started in Spain, where my early work focused on robot manipulation and grasping—very much application-driven. But I gradually became more curious about the cognitive and developmental dimensions of robotics, which led me to spend several months in the US, training in developmental psychology to explore how those principles could be applied to robotics. 

After that, I spent a year in Japan, where I was deeply influenced by their pioneering work on social robotics and the creation of human-like machines. The emphasis there was less on function and more on how robots can help us understand human behaviour, emotion, and interaction. That experience reinforced my interest in using robots not just as tools, but as models for studying human development. 

I then continued this developmental robotics approach in the UK, where I began to explore how robots can serve as testbeds for understanding early human learning, especially language learning. Later, in the Netherlands, I further developed this line of work by investigating what drives human trust in robots—how aspects of their appearance, behaviour, or even subtle social cues influence the way we relate to them. 

All these experiences have given me a broad and interdisciplinary perspective on robotics—combining technical skills with a deep curiosity about human behaviour. It continues to inform how I approach intelligent systems today, especially in thinking about how they integrate into human contexts in a meaningful, trustworthy way. 

 

You got the nickname ‘Robot Mama’ for working with ‘child’ robots. Can you elaborate? 

Sure! I was exploring a relatively new approach at the time called developmental robotics. The idea was simple but powerful: if we want to build truly intelligent machines, we can’t start at the level of adult cognition—we need to understand how babies develop and learn. At that time, even a one-year-old child could move, grasp, and communicate far better than the most advanced robots built through multi-year, multi-million-dollar projects. That contrast really drove home the importance of studying development. 

During my time in Japan, I worked in a lab that was entirely focused on building child-like robots—some resembling toddlers, others even designed as robotic infants. The goal was to explore how physical embodiment and social interaction contribute to learning. It was both technically challenging and deeply interdisciplinary. 

Later, when I continued this line of research at the University of Nijmegen in the Netherlands, I became known for championing developmental robotics within the academic community there. I even secured funding to grow that research. Somewhere along the way, with all the time I was spending with these ‘child’ robots, colleagues jokingly started calling me the Robot Mama—and the nickname stuck! 

 

How have you seen AI and robotics progress over the last few years and what areas of life you think AI and robotics will change for us all in the next 5 years? 

The past few years have been transformative, particularly with the rise of large language models (LLMs). We’ve already seen how they’ve reshaped the way we interact with technology. One of the biggest changes has been in how software is written—AI is now generating significant portions of code, and this is already affecting the job market. Many computer science graduates are finding it harder to land entry-level programming roles because AI can now handle many of those tasks. Similar shifts are happening in other sectors too—for instance, in law, where AI is now summarising and even drafting legal documents. 

Looking ahead, I believe robotics will be the next major frontier. We’re already seeing robots preparing groceries in warehouses, autonomous vehicles transporting passengers, and drones delivering food or medicine. While industrial robots have long been used in car manufacturing and other structured environments, the focus is now shifting toward more dexterous, general-purpose robots—including humanoids that can work alongside or even support other robots. 

One area with huge potential is social robotics—particularly in healthcare and elder care. We’re starting to see early prototypes of humanoid robots that can assist not just with physical tasks, but also provide basic social interaction and emotional support. While widespread adoption may take longer, I believe the next five years will bring significant progress in this space—especially in pilot deployments and real-world trials. This will lay the foundation for a future where social robots become a meaningful part of daily life, particularly for aging populations and overburdened care systems. 

 

What are the areas of research are you working on with UCL academics? 

I work closely with researchers in AI and robotics, helping to translate breakthrough technologies into real-world applications. One exciting area is robot dexterity—we’re supporting projects developing advanced tactile sensors that allow robots to manipulate even fragile or irregular objects with remarkable precision. There’s also work on whole-body robotic “skin,” which could make robots much more aware of their physical environment and significantly safer when interacting with humans. 

Beyond robotics, there are also fascinating projects in the VR space. Some are exploring how virtual environments can be used to enhance social interaction or immersive training, opening up new possibilities in areas like education, therapy, and collaborative work. 

What makes these projects particularly exciting is that they combine deep technical innovation with clear societal relevance—and that’s where I see the real potential for impact. 

 

What were some of the challenges and successes you encountered founding your AI mental health startup in the UK? 

After leaving academia, I joined Zinc VC’s venture builder programme, which focused on tackling one of the UK’s biggest social challenges: improving children’s and young people’s mental health. It was within this programme that I developed my solution. One of the most valuable aspects was having access to a wide network of experts—from business, healthcare, and academia. It really made me realise that it takes a village to support a startup, especially in a space as complex and sensitive as mental health. 

One of the biggest challenges was securing funding at the early stages—just to build a working prototype and test it with real users. While Zinc provided some initial support, getting follow-on investment was highly competitive. Breaking into the healthcare and mental health ecosystem is tough. There are long validation cycles, strict regulations, and the stakes are high, because you’re working with vulnerable populations. There’s also constant pressure to raise more funding, often before you’ve had a chance to properly test and iterate. 

Despite the challenges, it was an incredibly rewarding experience. It taught me a lot about navigating uncertainty, building in complex systems, and staying focused on user impact—lessons I still draw on today. 

 

We hear that you are into dancing Argentine tango! Tell us more and what do you get up to outside of UCLB? 

Yes, I’ve been learning Argentine tango for almost three years now. It’s such a beautiful dance—really about creating a deep connection with your partner and with the music. It’s wonderfully complex, with many layers to explore, and something you can spend a lifetime perfecting. The social aspect of tango—the community and the shared experience—is equally rewarding. 

Outside of dance, I stay active with strength training to keep fit and balanced. I also enjoy jazz music and theatre—living in London is fantastic for that, with so many incredible performances and venues to explore. 

 

Learn more about Beata.