Human-Humanoid Co-working in a Joint Table Transportation
Résumé
We address the problem of programming a human-humanoid haptic joint action consisting in transporting an beam. For such tasks humanoid robots plan footsteps to walk while keeping balance and coordinating whole-body manipulation-locomotion. In order to achieve such tasks, we devised a method consisting in assembling in a task-based control framework, a reactive walking pattern generator with a footprint planning strategy mapped from robot force cues interpretation and a task kinematic constraint. This approach basically decouples locomotion from manipulation but uses several heuristics to well tune each independent task to work in conjunction with the other. Experiments conducted on a real HRP-2 humanoid robot assessed the method.