Neural Network-Based MEMS Failure Probability Prediction

This paper reports a neural network-based methodology for failure probability prediction and quality enhancement of microengine MEMS using attribute data derived from actual measurements on microengines. A backpropagation neural network was employed for failure probability prediction. Microengine attributes constituted the inputs while time-tofailure statistics (mean, median and shape parameters) constituted network outputs. Once neural network training was complete, independent data was used to validate results. Correct prediction of failure statistics was achieved with high confidence (~0.9). Low humidity (0-10% ) and high microengine resonant frequency coupled with microengine operation at 0.4 of resonant frequency was found to result in median times-to-failure of at least 200 million cycles – i.e. high reliability.