PRASYS aims to develop the algorithms for the perception of mobile autonomous systems in the real world environment.
An autonomous system is supposed to simultaneously analyse the data collected by a variety of sensors, acquire a representation of the surrounding environment and its uncertainty, use it as an input for navigating and control, all of this in order to perform the mission as efficiently as possible. The autonomous system can be made of a single robot or a team of robots (possibly heterogeneous). A strong link then exists between navigation and sensor data processing.
To build such intelligent systems, the action-perception loop is studied within two themes : autonomous perception and autonomous action.