The Use of Prognostic Health Management for Autonomous Unmanned Air Systems

Abstract

Unmanned Air Systems (UAS) show great promise for a range of civilian applications, especially „dull, dirty or dangerous‟ missions such as air-sea rescue, coastal and border surveillance, fisheries protection and disaster relief. As the demand for autonomy increases, the importance of correctly identifying and responding to faults becomes more apparent, as fully autonomous systems must base their decisions solely upon the sensors readings they receive – as there is no human on board. A UAS must be capable of performing all the functions that would be expected from a human pilot, including reasoning about faults and making decisions about how to best mitigate their consequences, given the larger context of the overall mission. As these autonomous techniques are developed their benefits can also be realised in non-autonomous systems, as realtime aids to human operators or crew. This paper proposes a novel approach to PHM that combines advanced Functional Failure Mode Analysis with a reasoning system, to provide effective PHM for autonomous systems and improved diagnosis capability for manned aircraft.