Monitoring and Diagnostics for a Hydraulic Robot in Hazardous Environments

Abstract

Hazardous environments are an important application of modern robotic techniques. However, failures can be quite serious in such environments, especially if they trap the robot within or damage containment efforts. Thus robust early fault detection is especially important for such robots. In this paper we discuss the application of one such method, analytical redundancy, to a hydraulic system similar to the one found in hazardous environment robots such as the Rosie worksystem. Further, we extend the method beyond its linear formulation to better deal with the nonlinearities inherent to hydraulic systems.