Perception for a transport robot in public environment

FIRST is a mobile robotic system suited to the transport within hospitals of heavy trolleys such as those containing laundry, meal trays, drugs, etc. The system comprises a fleet of mobile robots, a central control station and several distributed control stations. The robots may use any route following a network of paths taught during a teleoperation phase. They navigate autonomously in the environment of the hospital, in particular in the presence of humans, be they hospital staff, patients or visitors, all in complete safety. Autonomous navigation requires the ability to react to unexpected events, hence the importance of a perceptual system which has two functions to recognise obstacles in order to avoid them in a safe and autonomous fashion, and to detect all situations which could endanger people and goods during the transport and handling of the trolleys. This paper presents the design principles and the performance of the perception system observed during site trials