Miniature MEMS-based data recorder for prognostics and health management (PHM)

Novel prognostic sensors and reasoner algorithms are the core technology for detecting defects caused by accumulation of fatigue damage in electrical and mechanical systems over time. However, serious technical challenges to implementing a general health management strategy for helicopters and military aircraft still exist. For example, severe heat and vibration make it difficult to distinguish fault signatures from environmental noise. Moreover, bearing loads are very dynamic, making it difficult to distinguish subtle wear-out signatures from normal acoustic patterns. Detection can be improved by increasing the number of sensor locations, but this option is unattractive from the standpoint of added cost, weight, and data overhead of such a system. Our approach is to integrate MEMS sensors with a standard commercial microcontroller and measurement electronics. In this way, prognostic sensors can be positioned closer to the stressed components and provide higher fidelity data with lower cost. We present an innovative design for a prognostics and health management (PHM) data recorder that will facilitate sense-and-response logistics, and provide a small and inexpensive package. This low-cost, low-power, and lightweight solution is based largely on COTS components; it is implemented using a standard low-power lightweight microcontroller core and COTS MEMS sensors to record and process local temperature and vibration data, and status reporting is implemented using a short range wireless transceiver.