AROBAS : Advanced Robotics and Autonomous Systems (project-team)

About the project : Activity report Research reports Theses Theme : Num Numerical systems Sophia Antipolis - Méditerranée research center

Project-Team Presentation

The AROBAS project-team works on fundamental problems associated with the autonomy of robotic systems. These problems are addressed with a "bottom-up" approach, rooted in the modeling and control of highly nonlinear mobile robot systems, extending to the control of the interactions of these systems with their environment, based on sensory information, and culminating with the execution of complex missions which relies on the construction of models of the environment. The ambition of AROBAS is twofold: to propose generic solutions to canonical problems independently of an applicative framework, and to apply these solutions to practical problems of industrial partners in the domains of ground, aerial, and underwater robotics.

Research themes

• Autonomous navigation and SLAM: many applications based on the use of robotic devices (ground vehicles, aerial robots, underwater autonomous vehicles, ...) involve navigation issues. A difficult problem is when neither the robot path nor the map are known, localisation and mapping must be considered concurrently. This problem is known as Simultaneous Localisation And Mapping and it is intensively addressed by a large and active scientific communauty. Whereas the problem of motion control is rarely considered as a constitutive part of the SLAM problem, we advocate that autonomous navigation and SLAM should not be treated separately, but rather addressed in an unified framework involving perception, modeling, and control.
• Perception for autonomous robotics: autonomous capabilities heavily relies upon the perception, the modelling of the surrounding environment in order to interact with it. Robustness and efficiency at the sensor processing level are the key points to ensure the robot performances. Robustness requires to explicitely handle photometric models, dynamic scenes and multisensor fusion. Thanks to efficient real-time algorithms, sensor data can be used directly in the design of sensor-based feedback control.
• Feedback control of underactuated and critical nonlinear systems: study of a general control approach for ground, marine, and aerial vehicles whose actuation and motion rely on a vectorized thrust force, and pursuit of the development of the Transverse Function approach for critical systems whose linear approximation about equilibria is not controllable.

International and industrial relations

• Academic collaborations with the Instituto Superior Technico of Lisbon (Portugal), and the CenPRA of Campinas (Bresil),
• Participation to the European Project FP6 Pegase with Dassault, EADS, ALENIA, EUROCOPTER, IJS, INRIA/ Lagadic, INRIA/ Vista, CNRS/I3S, IST/DEM (Portugal), Università di Parma (Italy), EPFL (Swiss), ETHZ (Swiss), Institut "Jozef Stefan" (Slovenie),
• Industrial contracts with Ifremer, ECA, Thalès, Bertin,
• Participation to the "Pôles de compétitivité" "Pégase" and "Mer" (PACA) and "System@tic" (Ile de France),
• ANR projects with academic partners: Lasmea, Heudiasyc, Amiens, Laas, LCPC, Université de Bourgogne.

Scientific leader

Patrick RIVES
+33 4 92 38 77 52
Patrick.Rives@sophia.inria.fr
Secretary : +33 4 92 38 77 48