ANR Cobot (2019 - )


Context of the project

Today, robots appear as a solution to help humans in the factory, at the hospital or at home. However, the direct and safe interactions with humans are still a great challenge. Humans are soft, flexible and their behaviour have been optimised through learning mechanisms and through evolution. Despite great advances in robotics, no robots can safely be involved in collaborative tasks with humans. At best, the robot is able to follow or to drive the manoeuvre, but in the first case, the robot is no more than a wheelbarrow (but is much more expensive) and in the second case, either the human does not accept to be driven or he has to compensate for the lack of reactivity and manoeuvrability of the robot. To avoid these problems, we must find a way for robots to collaborate safely and in real time with humans.
It is classical that human behaviours inspire robotics algorithms. However, the precise nature of the coordination between two humans allowing such collaboration is unknown while humans are able to develop behaviors where “the whole is better than the sum of the parts”. The challenge facing movement sciences is that facing all the research on complex systems. In spite of the apparent simplicity of a skilled movement, the organization of the underlying neuro-musculo-skeletal system remains unknown. A reason is the redundancy of the motor system: a given movement can be realized by different muscle and joint activity patterns, and the same underlying activity may give rise to several movements. Some theories, such as optimal command or motor primitives, provide tentative solutions to this conundrum, but none satisfies all empirical findings, which largely pertain to very basic movements, such as grasping or pointing. Studying an all-body movement like locomotion, mobilizing most joints and muscles and involving body balance, may be most revealing when submitted to parametrically varying external constraints, here an interaction with another subject.

Aim of the project
We designed this project to explore the collaboration strategies used in a load-carriage task performed by two human subjects and to develop robotic simulations to test our assumptions regarding the optimization of their collaborative behaviours.

Coordination: Bruno Watier (Team Gepetto, LAAS/CNRS, Toulouse)

Participants responsible of Workpackages: Pierre Moretto (Team CAB, CRCA, Toulouse), Nicolas Turpin (Team  IRISSE, EA4075, la Réunion)

P. Moretto of team CAB is responsible of Workpackage 1: Biomechanics of collective load carrying in humans. 1 PhD hired for the project: Nour Sghaier



Université Paul Sabatier
118 Route de Narbonne

31062 TOULOUSE Cedex

Annuaire général