Sensor integation using concurrent computing on-board the ORNL mobile robot

The mobile robot prototypes developed at the Center for Engineering Systems Advanced Research (CESAR) at the Oak Ridge National Laboratory (ORNL) are equipped with sonar sensors, CCD cameras and a laser range camera that are used to support autonomous navigation and inspection tasks in an a priori unknown and unstructured dynamic environment. This paper summarizes work directed at extracting information from data collected with these sensors and integrating it, in order to produce reliable descriptions of the robot's environment. The approach consists in studying different world models and mappings among them, sensor models and parallel algorithms for sensor information processing, and appropriate integration strategies. Specifically, the paper describes the integration of two-dimensional vision and sonar range information, and the integration of laser range and luminance images. 16 refs., 3 figs.