The importance of research to PAL Robotics: a continuous loop between research and development
Research & development (R&D) and innovation in robotics are elements that depend on one another for success. At PAL Robotics we have always been fully immersed in the R&D community and have many team members join us from academia, including PhD researchers, active in a range of European research projects. The Scientific Advisory Board at PAL Robotics was formed in 2021 to further strengthen our aims in research, and here we take the opportunity to interview the members of the group to hear more about why robotics research is important to PAL Robotics, what they are currently working on, and future plans. We spoke to Séverin Lemaignan, Narcis Miguel, and Enrico Mingo Hoffman.
Research is vitally important at PAL Robotics, as Enrico Mingo Hoffman told us, “research is the way for a company to grow, in terms of capabilities and skills. The ideal way to work is to do a continuous loop between research and development.”
Enrico continued, “PAL Robotics has always been very active in R&D – and it’s nice to have people coming from academia or research centres as they can take on that part which can be a long path from development until production of a robot.”
Enrico added, “The main goal of PAL Robotics is to provide solutions to challenges that the customer wants to solve, and to improve the quality of people’s daily life through robotics, and research is a key part of this. This could be for example to help workers in stores via inventory tracking, or in industry via humanoid robots that lift heavy items, so that humans only have to check that everything is okay.”
Narcis Miguel commented, “our robots are where theory is put to the test. We are giving factual proof that research works and is useful, as we are implementing outcomes of the research into our platforms, and showing that this works out-of-the-box.”
Starting the Scientific Advisory Board: setting scientific direction and connecting trends in research to our robots
Narcis commented on the goals of the Scientific Advisory Board, telling us, “the aim is to connect the most recent trends in research to each of our robots, bringing this together with the interests of our team in research. We want to also demonstrate that our robots are participating in the latest research.”
Enrico added, “having an advisory board helps to give a direction to all of our research activities as a company, and also to be able to then plan all the activities that are related to this, for example being present in research publications and active in scientific conferences.”
Séverin Lemaignan continued, “we wanted to increase the visibility of research at PAL Robotics, as a way for the company internally and externally to signal that science is important, and that to enable academics to support scientific activities in the company including workshops, and outreach activities within the scientific community.”
Séverin concluded, “it’s also important for us to set the scientific strategy in terms of our priorities. Although our work is driven by the needs of our customers, we also need to look a little bit further: can more basic research contribute to our longer-term vision? The job of our Scientific Advisory Board is to make sure we always look one step further in terms of research and innovation.”
R&D challenges in areas including locomotion, whole-body control, HRI and social robotics
Regarding his work in research, Narcis told us “I’m a mathematician, specialised in dynamical systems – this links me to challenges like control in humanoid robots, ranging from manipulation to locomotion. I’m especially interested in learning and adaptive control. I’ve been involved in locomotion in general and specifically in the development of the Kangaroo robot – for both modelling and its control. Now my role is shifting to manipulation-related challenges, in particular working with the TIAGo robot.”
Enrico told us, “I’m responsible for research in humanoids and legged systems – including whole-body control, locomotion and loco-manipulation, planning and control, and I work with Kangaroo, TALOS and REEM-C. Recently I’ve worked on research on modelling of Kangaroo robot, I also edited a Research Topic in the journal ‘Frontiers In Robotics and AI’ related to trajectory optimization and model predictive control – this considerers research on many different types of robots, including PAL Robotics’ Kangaroo and REEM-C.”
Enrico continued, “At PAL Robotics I’m approaching something new – generation and control of dynamic motions which is very challenging, and I enjoy seeing the robots performing – running, jumping etc. These developments also help to build the know-how that allows the company to grow.”
Séverin explained, “my work is in social robotics, including working with ARI, and, really, our core question is: how to design and program a robot so that it is genuinely useful to people, without taking away roles that humans could do better. For instance, we are inventing ways of getting our end-users, sometimes even children, to directly teach the robot what they want the robot to do (or not to do!) for them. Our primary challenge is to get the robot to understand what is going on, we call that ‘social situation assessment’, and this is far from being a solved issue. Especially when more people are present at the same time: who wants to join in? Who wants to talk? Are people ‘just looking’ or do they want to interact? Those basic social skills are at the core of the research I conduct at PAL Robotics.”
Aligning research with trends including developing humanoid robots that work beyond human capabilities
Regarding aligning research at PAL Robotics with research trends, Enrico said, “a number of high profile companies are developing humanoid robots. With some of the videos we have seen on YouTube involving humanoid robots there is a renewed interest in pushing the limits further and further. Every time there is a new activity the peak of capabilities achieved is even higher than before. This takes us in the direction of developing humanoid robots that can work beyond human capabilities. PAL Robotics is pursuing this direction with the Kangaroo robot that has incredible potential.”
Séverin said, “it’s interesting to see trends in the community – there has been a huge push to get out of the lab and into ‘the field’, which means, getting robots into the hands of real people, in their homes or workplaces for longer period of time – weeks or even months. When you develop robots in a research lab, a lot of the real-world complexity is hidden. But working in a company that tries hard to put robots in everyone’s hands, we are forced to confront all that complexity. A big challenge, but an exciting one! And for that reason, I think PAL Robotics is a great place to make sure research will have a real-world impact.”
The future of robotics R&D: tackling real-world challenges
Regarding the future of robotics research, Narcis told us, “ the idea is to adapt our robots as much as possible to research, and this is important for all of our robots. For example, to adapt to the current trend in machine learning-based applications we provide already integrated add-ons for the TIAGo robot that allow using these methods to their full potential. In addition, we need to investigate and understand better how robots interact with humans, as more and more people are interacting with robots, in particular people who are not experts.”
Narcis continued, “PAL Robotics is a research company by nature, as our platforms are where research has to be implemented, and there has always been a following of research trends that concern types of robots we have, and known developments and ideas that the community is focusing on.”
Enrico explained, “with research, you never know if it will work or not – and you need time to make it work etc, therefore having a good blend between research and engineers is a loop and collaboration always that makes things go faster.”
Séverin commented “my ambition for research at PAL Robotics is to show that, as a company, not only we have to keep up with all the amazing research taking place in academia to stay on top of the game, but also that we are ourselves able to do great research and push the boundaries of human knowledge. And we are indeed very well placed to tackle complex, real-world challenges that need to be solved.”
Being active and encouraging the team to be active in R&D
Regarding being active in R&D Narcis told us, “this happens naturally for us, one example is within EU research projects we partner in, which have a huge research-related component. This keeps us in contact with communities that are directly related to robotics research”
On the theme of encouraging the team at PAL Robotics to take on new challenges in the world of research, Séverin told us, “we hope to hold a mini-conference where everyone can tell their colleagues more about research they’re working on, big or small, to solve challenges we have. Starting with small problems, that are exciting and need to be solved – that is having a scientific mindset.”
Read our previous interview with Enrico Mingo Hoffman. Find out more about Séverin Lemaignan’s previous work in our blog interview.
We would like to thank PAL Robotics’ Scientific Advisory Board for taking the time to talk with us. You can learn more about our recent research and our robotic research platforms on our website. If you have any questions for the team at PAL Robotics don’t hesitate to visit our contact page.