Exploration of an unknown environment by a mobile robot

Mobile robots can make important contributions to the integration of industrial operations. The improvement of mobile robot autonomy will lead to many benefits which include a decrease in programming effort, simplifications on shop floor design, and the lowering of costs and time to market. Additionally this will free human resources which may be applied to the areas of product development. All those factors will ultimately contribute for enabling more agile and sustainable industrial production. In this article we demonstrate the validity of the parti-game self-learning approach for navigating a mobile robot, by finding a path to a goal region of an unknown environment. Initially, the robot has no map of the world, and has only the abilities of sensor-based obstacle detection and straight-line movement. Simulation results concerning the application of the learning approach to a mobile robot are presented.