Data Strategies for Fleetwide Predictive Maintenance
This work provides practical improvements for fleet maintenance operations by optimizing algorithm performance and feature selection.
The researchers tackled the problem of predictive maintenance by analyzing a large public dataset of machine failures, identifying three algorithms that achieved 96% accuracy with twenty-fold faster execution times than previous work. They also developed a feature importance ranking methodology showing error counts were more predictive than traditional factors like machine age.
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 time and accuracy comparison between 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.