An Embedded Instrument for Online Condition Monitoring of Unmanned and Autonomous Systems

With the integration of Artificial Intelligence (AI) and robots, unmanned and autonomous systems (UAS) have been widely used both in civil and military fields. However, the complexity of UAS increases with the improvement of its autonomous ability, which puts forward higher requirements for online condition monitoring (OCM) to timely confirm the safety and availability of UAS. To achieve UAS OCM, an embedded OCM instrument combing perception, intelligent algorithms, and edge computing should be developed, where perception is responsible for collecting UAS operation data through sensors, the intelligent algorithm uses the operation data to assess UAS status, and the edge computing is utilized to support the online operation of an intelligent algorithm. Since intelligent algorithms such as deep learning usually have high time complexity, it brings great challenges to assuring the timeliness of the embedded OCM instrument, especially under the limitations of size, weight, and power (SWaP). Therefore, in this paper, first, the UAS OCM is introduced. Then, the design guidelines for the embedded OCM instrument are given. Finally, the embedded OCM instrument of the unmanned aerial vehicle (UAV) is presented in detail as a case study.