The French ANR Project SESAME:

Singularity and Stability of Sensor-based Controllers

---

---

Objectives of the SESAME project

The request for manufacturing robots which are flexible, adaptive, and able to interact with their environment is increasing (e.g. in navigation, manipulation, cooperation). Therefore, since several years, there is a large scientific development of sensor-based controllers which allow including in the robot controllers external sensorial information.

SESAME is a project on fundamental research in which we propose to overcome scientific issues that prevent for the moment the acceleration of the industrial transfer of sensor-based controllers for robots, issues which are the study of singularities and of stability for sensor-based controllers.

 

---

Project methodology

The objective of the project SESAME is to synthesize globally asymptotically stable sensor-based controllers, whose performances (i.e. convergence properties towards the desired configuration, absence of singularities and local minima) are guaranteed in all object/sensor relative configurations.

In view of the state of the art on sensor-based controllers, this objective is ambitious. To reach it, the theory and algorithmic allowing to analyze the singularities and the basin of attraction of sensor-based controllers will be developed initially, then methods of optimal synthesis of sensor-based controllers based on these analyses will be proposed.

The objective will be reached thanks to a pluridisciplinary approach including skills in sensor-based controllers (teams ARMEN of LS2N, Lagadic of IRISA), mechanics (team ARMEN of LS2N) and computational mathematics (teams OGRE of LS2N and PolSys of LIP6).