Adaptive Whole-Body Manipulation in Human-to-Humanoid Multi-Contact Motion Retargeting

Kazuya Otani (1, 2) , Karim Bouyarmane (3, 2)

1 Inria - Institut National de Recherche en Informatique et en Automatique
2 LARSEN - Lifelong Autonomy and interaction skills for Robots in a Sensing ENvironment
3 UL - Université de Lorraine

We propose a multi-robot quadratic program (QP) controller for retargeting of a human's multi-contact loco-manipulation motions to a humanoid robot. Using this framework , the robot can track complex motions and automatically adapt to objects in the environment that have different physical properties from those that were used to provide the human's reference motion. The whole-body multi-contact manipulation problem is formulated as a multi-robot QP, which optimizes over the combined dynamics of the robot and any manipulated objects. The multi-robot QP maintains a dynamic partition of the robot's tracking links into fixed support contact, manipulation contact, and contact-free tracking links, which are re-partitioned and re-instantiated as constraints in the multi-robot QP every time a contact event occurs in the human motion. We present various experiments (bag retrieval, door opening, box lifting) using human motion data from an Xsens inertial motion capture system. We show in full-body dynamics simulation that the robot is able to perform difficult single-stance motions as well as multi-contact-stance motions (including hand supports), while adapting to objects of varying inertial properties.

Domaines: Robotique [cs.RO]

Kazuya Otani, Karim Bouyarmane. Adaptive Whole-Body Manipulation in Human-to-Humanoid Multi-Contact Motion Retargeting. IEEE-RAS International Conference on Humanoid Robots, Nov 2017, Birmingham, United Kingdom. ⟨hal-01569390v1⟩