Data Strategies for Fleetwide Predictive Maintenance

Senior Technical Fellow PeopleTec, Inc. Huntsville, AL, USA ABSTRACT For predictive maintenance, we examine one of the largest public datasets for machine failures derived along with their corresponding precursors as error rates, historical part replacements and sensor inputs. To simplify the timeaccuracy comparisonbetween 27 different algorithms, we treat the imbalance between normal and failing states with nominal under-sampling. We identify 3 promising regression and discriminant algorithms with both higher accuracy (96%) and twenty-fold faster execution times than previous work. Because predictive maintenance success hinges on input features prior to prediction, we provide a methodology to rank-order feature importance and show that for this dataset, error counts prove more predictive than scheduled maintenance might imply solely based on more traditional factors such as machine age or last replacement times. INTRODUCTION Successful predictive maintenance is challenging not only because failures can prove multifactorial but also because maintenance forecasters often lack good training data.