Welcome.I am an Associate Professor in Robotics with the Oxford Robotics Institute at the University of Oxford, and a lead at the Dynamic Robot Systems group. My research targets the combination of machine learning with dynamic whole-body motion planning and control, focusing on robots with arms and legs. Building on my previous work, I am interested both in the locomotion and manipulation aspects of autonomous robots. Before, I was a postdoctoral researcher at the Robot Learning & Interaction Group of the Idiap Research Institute in Switzerland. I worked on the Horizon 2020 DexROV project, where my research focused on machine learning approaches that enable robots to learn complex skills from demonstration, in an online, interactive and intuitive manner. Previously, I was a senior postdoc and the head of the Locomotion Group of the Dynamic Legged Systems Lab at the Advanced Robotics Department of the Italian Institute of Technology. I have a passion for robotics and love working on robots with arms and legs. I also have a strong background in machine learning and believe that learning is key for autonomous mobile robots. I hope that my work will enable robots to become more autonomous and widely used in a variety of applications, from inspection and monitoring, to maintenance and construction work, in industrial, business and domestic environments, performing a range of specialised and every-day tasks. |
- Robotics: whole-body approaches, dynamic motion planning and control, locomotion, optimal control, dynamical systems, optimization.
- Machine learning: data-driven and on-line learning, manifold learning, probabilistic approaches, geometric representations, compact and reusable skill/task representations.
I worked on robot learning with probabilistic models, mainly for skill learning, recognition and motion generation.
I worked on problems on and around legged locomotion with the HyQ. This included motion planning, foothold selection, control, body trajectory optimization, among others.