Positioning and Objectives
The Inria Center at the University of Bordeaux hosts various research teams that incorporate robots into their projects (e.g. Auctus team or Manao team). The Experimental and Development Service (SED) is responsible for establishing and maintaining the experimental platforms of the center, including the maintenance of its robots. These robots operate within ecosystems that encompass various sensors such as cameras and lasers, and they are controlled using the ROS middleware. Currently, ROS2, the second version of ROS, is at a relatively stable development stage, and we are seeking to migrate all of our platforms to ROS2. Furthermore, Inria has recently acquired a new robot, a KUKA IIWA 14. The objective of this internship is to commission this robot and then develop a ROS2-based solution that can seamlessly integrate into the existing experimental environment.
In addition to the purely computational aspect related to the installation of ROS2, the internship will also require gaining a deep understanding of the softwares developed by the Auctus team for the use of its robots. These softwares encompass various aspects, ranging from sensor data integration for high-level task generation to trajectory planning and robot motor control. The goal of the internship is to be able to use all of these software components with the KUKA robot. Special attention will also be given to the ability to simulate the robot’s behavior using Gazebo Ignition software. Finally, the Qontrol software developed by the Auctus team should be operational on this robot. This software allows for defining control problems for manipulator robots in the form of constrained optimization problems. At the end of this internship, various examples of Qontrol usage should be implemented on the robot, including:
- Using Qontrol for simple trajectory tracking tasks
- Exploiting the robot’s redundancy for secondary tasks (gravity compensation, breathing motion, etc.)
- Modulating constraints based on sensor data (e.g., speed modulation relative to the distance to an operator) Advanced work by the team’s doctoral students on Model Predictive Control may even be tested on this robot.
Research Environment
This internship will take place both within the SED and in collaboration with the Auctus team. The project will rely on the hardware and software resources of Inria’s experimental platform. The selected student will be guided throughout the internship with regular meetings.
Expected Skills
- Background in robotics (modeling, control, etc.)
- Proficiency in Linux environments.
- Knowledge of Python and C++
- Familiarity with ROS (1 and/or 2) is a plus
How to Apply
You can submit your application by email with your CV, a cover letter, and your Master’s degree transcripts at - Lucas Joseph - lucas.joseph@inria.fr