A Unified Approach to Sensor-Based Terrain Characterization and UGV Mobility Prediction

Unmanned ground vehicles (UGVs) will play an important role in the nation's next-generation ground forces. One key limitation of autonomous UGVs arises from current approaches to sensor data analysis and interpretation. Most current approaches to sensor data analysis rely on classification: that is, analyzing remotely observable "features" to assign a terrain region as a member of a pre-specified semantic class. The purpose of this proposed Phase I STTR research program is to develop a method for terrain characterization that would generalize locally-sensed physical terrain features to remotely-sensed data to infer properties about UGV mobility through its surroundings. In the proposed concept, "local" (i.e. proprioceptive) sensors would measure signals related to physical UGV-terrain interaction. Local sensor feedback would be analyzed to identify terrain patches that possess unique mobility characteristics, and visual features associated with these terrain patches would be correlated with remote data, thereby creating a "mobility map" of the surrounding environment. Contrary to classical terrain classification methods, this map would not delineate semantically-labeled terrain boundaries, but rather would delineate mobility-labeled terrain boundaries. The proposed work would be performed as a collaboration between researchers at the Massachusetts Institute of Technology and Quantum Signal, LLC (QS).