Previous | 
Home
Section: Highlights of the Year
Highlights of the Year
Fondamental results
Three PhD students have defended excellent thesis in 2019:
-
Eulalie Coevoet: Optimization-Based Inverse Model of Soft Robots With Contact Handling
-
Zhongkai Zhang: Vision-based Calibration, Position Control and Force Sensing for Soft Robots
In each of these thesis we presented fundamental results on the team's roadmap. Eulalie Coevoet presented the first algorithms that allow inversing the robot model in contact situations. It can be used for planning, manipulation and locomotion. Zhongkai Zhang's results allow the use of the robot as a generalized force sensor thanks to vision. We use that for feedback control for both position and force. Maxime Thieffry developped the first method for dynamic control based on model order reduction. The method is very generic and significantly improves the precision of soft robots.
Awards for software development
DEFROST actively contributed to the open source community by developing plugins for the SOFA framework. The team participated in the SofaWeek2019, during which the SOFA consortium organized the “Open-Source SOFA awards”. One prize was offered to the candidate who developed the best open source plugin for SOFA. Another prize was offered to the best open source plugin according to the public (conference attendants). Both prizes were won by the DEFROST team, for the Model Order Reduction plugin and the SofaPython3 plugin respectively: link.
Organization of workshops and tutorials
This year, special effort was expended on the promotion of our tools through the organization of workshops and tutorials. A full tutorial day about our tools was organized at the IEEE International Conference on Soft Robotics (RobotSoft019). We then organized the first Journée de Robotique Souple in Lille with 70 participants from 9 countries. The team also participated in the organization of the (2nd Workshop on Proximity Perception in Robotics) at the IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2019).
ANR project ROBOCOP
The project ROBOCOP (ROBOtization of COchlear ImPlant) was funded by the ANR (Agence nationale de la recherche) for the development of cochlear implants for the future that are activated by electroactive polymers. The project is in collaboration with the IEMN, the LPPI, Inserm and Oticon Medical. This project will allow us to fund a PhD Student and a postdoctoral fellow for 2 years.