Detection of mobile machine damage using accelerometer data and prognostic health monitoring techniques

Caterpillar, Inc. and Frontier Technology, Inc. (FTI) are investigating prognostic health monitoring technologies for application to Caterpillar equipment. In particular, robust detection of mechanical damage in a wheel loader has been demonstrated via processing of high-speed, three-axis accelerometer data. Data collected with and without the damaged parts show distinctive signatures that are quantitatively separable. FTI's Pattern Recognition of Health (PRoHtrade) technology drives the signature generation and abnormality detection process through the use of data-driven techniques that estimate deviation from normal behavior.