Session 10: Sensing and Planning

摘要

This session was composed of four presentations spanning various aspects of sensor modeling, motion planning, and tracking. In each case the emphasis was on using insights from currently implemented systems to re-think the underlying models conventionally used in the literature today. The first paper in this session, "Adaptive Full Scan Model for Range Finders in Dynamic Environments" by De Laet, Smits, De Schutter, and Bruyninckx describes recent progress in probabilistic range finder models for dynamic environments. The authors presented an experimental validation of a prior approach (the igorously Bayesian Beam Model), and proposed a sample-based full scan model which is an improvement on the state of the art models due to the way it handles multi-modality of the range finder data.