WE PROPOSE DEVELOPMENT OF ALGORITHMS FOR SEMANTIC MAPPING OF COMPLEX INTRA-VEHICULAR ENVIRONMENTS AS WELL AS STRATEGIES FOR SAFE NAVIGATION WITHIN SUCH AREAS DESPITE CHANGES IN THE MAP. THAT IS OUR ALGORITHMS WILL ENABLE ROBOTS TO PERCEIVE THEIR ENVIRONMENT AS A COLLECTION OF HIGH LEVEL OBJECTS AS A HUMAN DOES NOT SIMPLY AS A COLLECTION OF POINTS OR OCCUPANCY GRID. WE EMPHASIZE RESEARCH ON SAFE AUTONOMOUS NAVIGATION IN AND CHARACTERIZATION OF DYNAMIC ENVIRONMENTS WHERE THE MAP MAY CHANGE. WE ALSO PROPOSE WORK ON CHARACTERIZING AND RECOGNIZING SOFT OR JOINTED OBJECTS WHICH ARE NOT RIGID AND MAY CHANGE IN CONFIGURATION. RESEARCH WILL FOCUS ON HIGHLY OPTIMIZED TECHNIQUES WHICH ARE CAPABLE OF RUNNING ON COMPUTATIONALLY CONSTRAINED PLATFORMS.

Trustees Of The University Of Pennsylvania, The

Investigators