Inside Dyson’s robotics research

When I first heard about a job opportunity at Dyson in robotics, I thought of the robot vacuum cleaner. I had no idea about the scope of Dyson’s research. Dyson’s investment in researching and developing robotic technology goes far beyond their existing product range. There are lots of universities doing research in robotics, but very few companies trying to apply that to get an actual product going. In that regard, the job here is incredibly exciting.

When I was young, I loved the robot novels of Isaac Asimov. And ironically, the novels seem to have had quite an impact on the path of my career.

I left school with a penchant for maths and physics and studied Telecommunications Engineering in in my home country of Spain before I went on to complete a masters in Computational Neuroscience in Berlin. This combined the mathematical side of electrical signal processing from my previous degree with the study of the human brain.

I then completed a PhD in Robotics and AI at the Italian Institute of Technology and worked with the iCub humanoid robot. My work there focused on machine learning: programming the robot to learn by interacting with the environment, in order to predict the consequences of its own actions.

My current role at Dyson bridges the gap between academia and invention in machine learning and computer vision. I think about ways to develop state-of-the-art methods into a useful application, then create prototypes to discover if it’s feasible in practice. From control and mechanics to computer vision, AI and machine learning, the real innovation lies in bringing all the specific fields of robotics together to realise a fully formed idea.

With this said, the collaborative nature of work here is something I find really exciting. You can spend a good amount of time working in your lane, trying to get your very specific area of development to a good level. But it is encouraged to constantly talk to each other across the teams, allowing each area of development to inform the others. It means that we generate results much quicker. Our team is a cohesive network of experts, from different backgrounds and industries, all working together to fulfil a common goal. We are prototyping constantly and there’s a strong sense of the direction we’re going in.

When I worked in research, I often felt like my specific work was like a little grain of sand in a desert. But here I have a more tangible impact day-to-day. We are using the grains of sand to construct something useful and the atmosphere is amazing because of that.