Algorithmic motion planning

A strong axis of the research conducted in RIS concerns the algorithmic foundations of motion planning for complex articulated systems. We have contributed to the development of the sampling-based methods that have now emerged as an efficient framework for facing the high complexity of motion planning.  Our contributions concern the two main classes of methods:  the Probabilistic Roadmaps sampling algorithms (PRM) and the Random Trees diffusion variants (RRT) developed for solving multi-query and single-query problems, respectively. The obtained results allow to better face the complexity of the search spaces (constrained and highly dimensional) and of the mechanisms (closed-chain systems), as well as to incorporate additional motion constraints (dynamic obstacles, path quality and uncertainty).

Motion planning for a complex humanoid torso

We also devote substantial efforts to planning and control problems of manipulation tasks in human environment. We have built a planner capable of synthesizing a complete manipulation task defined by an object to grasp and to hand over to a human. 

Asymov, a hybrid task planner: We also addressed the challenging issue of bridging the gap between symbolic and geometric reasoning. We have extended our work on the manipulation planning problem and come up with a general framework to solve intricate motion, manipulation and task planning problems.  We introduced a new extension to classical action planning formalisms with a rigorously defined interaction between action and motion planning. On this formal basis, we have built a planner that is able to solve intricate geometric and symbolic constraints that are not solved by any other planner. It is worth noting that the concept of manipulation planning that we have developed has inspired a number of new algorithms and geometric reasoning approaches.


Relevant publications

M. GHARBI , J. CORTES , T. SIMEON, Roadmap Composition for Multi-Arm Systems path Planning. Rapport LAAS N°09541, International Confefrence on Intelligent Robot and Systems, IROS 09, Saint Louis (USA), 11-15 Octobre 2009

S. CAMBON,  R. ALAMI , F. GRAVOT A Hybrid Approach to Intricate Motion, Manipulation and Task Planning. Rapport LAAS N°09345, The International Journal of Robotics Research, Vol.28, N°1, pp.104-126, Janvier 2009

L. JAILLET , T. SIMEON, Path deformation Roadmaps. Rapport LAAS N°07448, International Journal of Robotics Research, Vol.27, N°11-12, pp.1175-1187, Novembre 2008

J. CORTES , L. JAILLET , T. SIMEON, Disassembly path planning for complex articulated objects. Rapport LAAS N°07484, IEEE Transactions on Robotics, Vol.24, N°2, pp.475-481, Avril 2008

E.LOPEZ DAMIAN , D.SIDOBRE , R.ALAMI, Grasp planning for non-convex objects. Rapport LAAS N°05012, 36th International Symposium on Robotics (ISR'2005), Tokyo (Japon), 29 Novembre - 1er Décembre 2005, 6p., Rapport LAAS N°05012, IFR Robotics Newsletter. Special Addendum, pp.1-7, 2nd- quarter 2006

K.MADHAVA KRISHNA , R.ALAMI , T.SIMEON, Safe proactive plans and their execution. Rapport LAAS N°05084, Robotics and Autonomous Systems, Vol.54, N°3, pp.244-255, Mars 2006

T.SIMEON , J.P. LAUMOND , J. CORTES , A.SAHBANI, Manipulation planning with probabilistic roadmaps. Rapport LAAS N°03177, International Journal of Robotics Research, Vol.23, N°7-8, pp.729-746, Juillet-Août 2004

J. CORTES , T. SIMEON, Sampling-based motion planning under kinematic loop-closure constraints. Rapport LAAS N°04110, 6th International Workshop on the Algorithmic Foundations of Robotics (WAFR'2004), Utrecht (Pays-Bas), 11-13 Juillet 2004, pp.59-74

 

Recent thesis

E. A. SISBOT Towards human-aware robot motions, Doctorat, Université Paul Sabatier, Toulouse, 1 octobre 2008

I. HERRERA AGUILAR Commande des bras manipulateurs et retour visuel pour des applications à la robotique de service, Doctorat, Université Paul Sabatier, Toulouse, 1 septembre 2007

E. LOPEZ DAMIAN            Planification de saisie pour la manipulation d'objets par un robot autonome, Doctorat, Institut National des Sciences Appliquées, Toulouse, 4 juillet 2006

L . JAILLET Méthodes probabilistes pour la planification réactive de mouvements Doctorat, Université Paul Sabatier, Toulouse, 19 décembre 2005           

S. CAMBON            Planifier avec les contraintes géométriques du mouvement et de la manipulation            Doctorat, Université Paul Sabatier, Toulouse, 24 juin 2005

F. GRAVOT Fondation d'un planificateur robotique intégrant le symbolique et le géométrique, Doctorat, Université Paul Sabatier, Toulouse, 24 Mars 2004, 156p.

J. CORTES Motion planning algorithms for general closed-chain mechanisms, Doctorat, Institut National Polytechnique, Toulouse, 16 Décembre 2003

A. SAHBANI Planification de tâches de manipulation en robotique par des approches probabilistes, Doctorat, Université Paul Sabatier, Toulouse, 17 Février 2003, 137p