Immersive teleoperation of a mobile manipulator for assistance at home
Auteur : Serena Ivaldi
General information
Encadrant | Serena Ivaldi | Luigi Penco | |
Adresse | |||
serena.ivaldi@inria.fr | luigi.penco@inria.fr | ||
Bureau | C014 |
Motivation
Current telepresence robots are only “screen-on-wheels” that cannot interact with the world. For instance, they cannot be operated to open a door or grasp an object on the floor. The team recently acquired a TIAGO robot (http://tiago.pal-robotics.com) that combines a mobile base, an arm, and a vision system. We aim at teleoperating it with a motion capture system (Xsens Suit) and/or an haptic control device, combined with an Oculus Rift virtual reality helmet with joysticks.
Sujet
The objective of the internship is to develop an immersive teleoperation software that can be used to tele-operate the Tiago so that it can effectively interact with the world (moving around, open a door, grasp an object on the ground, take the elevator, etc.).
Cadre du travail
The robot is controlled with the ROS framework. It will be deployed in the “smart apartment” (a replica of a small appartment equipped with many motion capture sensors, installed at Inria/Loria). The starting point will be the teleoperation software and the oculus driver developed in a previous internship.
References
Penco L, Clément B, Modugno V, Hoffman M, Nava G, Pucci D, Tsagarakis N, Mouret JB, Ivaldi S. (2018) Robust Real-time Whole-Body Motion Retargeting from Human to Humanoid. IEEE Humanoids 2018.