Building Maps Using Indoor Mobile Robots with Ultrasonic and Laser Range Sensors

A NEW METHOD FOR LAERNING PROBABILISTIC GRID-BASED MAPS OF INDOOR ENVIRONMENTS BY A MOBILE ROBOT IS DESCRIBED. WE CONTRIBUTE IN TWO SUBJECTS OF MAP LEARNING, NAMELY, SENSOR DATA FUSION AND EXPLORATION. IN PARTICULAR, NEW MODELS OF SENSORS AND A WAY OF SENSOR DATA FUSION THAT TAKES ADVANTEGE OF MULTIPLE VIEWPOINTS ARE PROPOSED. THE EXPLORATION APPROACH MEGES A LOCAL STATRGY, SIMILAR TO WALL FOLLOWING TO KEEP THE ROBOT CLSE TO OBSTACLES, WITHIN A GLOBAL SEARCH FRAME, BASED ON A DYNAMIC PROGRAMMING ALGORITHM. THIS METHOD IS TESTED USING A SIMULATED AND REAL MOBILE ROBOT WITH ODOMETER, ULTRASONIC AND LASER RANGE SENSORS (IMPLEMENTED WITH A LASER LINE GENERATOR AND CAMERA) WITH PROMISISNG RESULTS.