A simple HUMS approach to detect characteristic variation for mechanical systems

Abstract

In 2003 the Defence Science and Technology Organisation (DSTO) and a local Australian company, Industrea Ltd., initiated a research project aiming to create a simple low cost Health and Usage Monitoring System (HUMS), now called SmartHUMS. This paper discusses the development of SmartHUMS technology including hardware and software aspects and possible applications. The design and manufacture of the SmartHUMS hardware was primarily carried out by Industrea with requirements specified by DSTO. This paper particularly focuses on the algorithm development conducted by DSTO. In particular the autocorrelation and cross correlation algorithms. The former is a very efficient method to identify the unique characteristics of a raw data signal, whereas the later is a simple way of comparing the signal characteristics shared between two data sets. When using these two algorithms concurrently they allow the SmartHUMS unit to detect behavioural variations within a mechanical system.