[robotics-worldwide] [jobs] PhD position in Humanoid Robotics at LIRMM, France
We are looking for applicants to a PhD position in Humanoid Robotics at LIRMM, France.
Duration: 3 years
Starting date: October 1st, 2018
In the current state of the art, motions realized on humanoid robots usually alternate
between fixed contacts with the environment. Fixed contacts have been sufficient to
describe a number of scenarios, including walking. Yet, they also impose strong
restrictions on the range of motions that humanoid robots can realize.
Imagine humans climbing a steep staircase: when a ramp is available, they are likely to
keep a hand sliding on it, increasing their overall stability with the ability to
squeeze the ramp whenever necessary. Far from this, in current stair climbing
experiments, robots are grasping the ramp at predefined locations (see e.g. references
 and  below), resulting in slow and overly conservative (hence power-consuming)
motions. The reason for this conservative behavior is that non-fixed contact "modes"
(for instance, sliding contacts) are not handled by state-of-the-art humanoid
The goal of the successful PhD candidate will be to develop this new capability. In this
endeavor, s/he will design new contact models and control software that will be tested
on the HRP-4 humanoid robot used in our lab (see e.g. our team's YouTube channel ).
About the team
The IDH team  is part of the CNRS–UM Montpellier Laboratory of Computer Science,
Robotics, and Microelectronics (LIRMM), Montpellier, France. It works in close
collaboration with the CNRS–AIST Joint Robotics Laboratory (JRL) located in Tsukuba,
Japan, with which frequent exchanges of people, know-how and software take place. As
part of the team, the candidate will be surrounded by a group with strong expertise in
humanoid motion planning and control, where we value both theoretical developments and
their application to the real robot. Our team collaborates closely with Airbus Group as
part of the COMANOID project . In the course of the PhD, the candidate will therefore
interact with practical scenarios expected to have a strong return in the years to come.
Master in robotics, computer science or automation with strong programming skills (C++
or Python preferred). Linux fluency is a plus.