Sylvain (Ph.D. degree in Mechanical Engineering from University of Versailles, France, M. Eng. degree in Mechanical Engineering from University of Poitiers, France) joined IHMC, in April 2013, as a Post Doctoral Researcher and is working with Drs. J. Pratt & P. Neuhaus in the Robotics Lab.
His main interest is humanoid/bipedal locomotion, especially the running motion. Sylvain’s hobbies are juggling, fire breathing and having fun.
DARPA Robotics Challenge (DRC):
- DRC Trials was the second phase, held December 20-21 2013, of the DRC. Team IHMC scored 20 out of a 32 possible points related to task completion in real world with the physical humanoid robot Atlas, placing second in a field of 16 teams.Sylvain worked on implementing the walking controller and state estimator on the physical robot Atlas and was the operator controlling the robot for the terrain task. The migration from simulation environment to real world caused several issues that had to be addressed. The objective was to increase the controller robustness to real world imperfections and the creation of a new state estimator. Both the “upgraded” walking controller and state estimator were used for the competition and allowed the full completion the terrain task.
- Virtual Robotics Challenge (VRC) was the first phase, held June 18-20 2013, of the DRC. Team IHMC scored 52 out of a 60 possible points related to task completion in a simulation environment, placing first in a field of 22 teams.Among the three tasks (driving a vehicle, walking over varied terrain, and manipulating a fire hose) Sylvain was involved in software development for the driving task and was one of the operators controlling the virtual humanoid robot Atlas during the competition. More especially, he focused on the vehicle ingress and egress part of the task: he worked on an suitable high-level humanoid controller and scripted the sequence of robot actions, both used in the competition.
Sylvain worked on the FastRunner project. He focused the passive stability (no balance controller using global state feedback) that lies behind the running ostrich robot by studying simple bipedal systems. He also developed a running controller using a with balance control despite for the underactuated robot. The controller has the interesting property of using an inverse dynamics algorithm to allow a control of the hip and knee angles with a non-actuated knee.