Dealing with challenging robots, environments and missions: operational space control in the context of motor learning

Robotics has evolved from an industrial, repetitive framework to application domains with much more variability of tasks and increasing complexity in uncertain environment. This is clearly the case for Service Robotics where the safety of the interactions is also a major issue. Based on these observations, part of the research activities of the Motion & Perception group at the Intelligent Systems and Robotics Institute (ISIR) from Université Pierre et Marie Curie (UPMC) is dedicated to human motion analysis and to motion generation and control for humanoid robots. This abstract gives a quick overview of the approach we are working on to solve the problem of motion generation and control for humanoid robots. It also tries to show how our approach relates to the existing literature in the domain and why it seems to be a promising alternative within the framework of Service Robotics and more generally in the context of humanoid robots evolving in uncertain environments.