Real time prediction and diagnostics for unmanned ground vehicle (UGV) mobility
Holger M. Jaenisch, James W. Handley, Michael L. Hicklen · Proceedings of SPIE, the International Society for Optical Engineering/Proceedings of SPIE · 2007
This paper describes a novel capability for predicting and diagnosing current unmanned ground vehicle (UGV) system health and status. Prognostication is the results of a multi-step process consisting of successful novelty detection, fault detection, fault diagnosis, and failure prognosis. UGV mobility prediction requires the fusion of both external and internal situational awareness, resulting in a course of action that can be executed by the UGV and confirmed by its own sensors. Our algorithms are analytical and enable both prediction and diagnostics to be performed in real time and within the limited processor speed and memory constraints of the UGV. This paper summarizes these algorithms.