Data Strategies for Fleetwide Predictive Maintenance

TL;DR

This paper evaluates 27 algorithms on a large public dataset for predictive maintenance, identifying top performers with high accuracy and speed, and introduces a feature importance methodology emphasizing error counts over traditional factors.

Contribution

It presents a comprehensive comparison of algorithms for predictive maintenance and a new feature ranking method highlighting the importance of error counts.

Findings

Three algorithms achieved 96% accuracy and were twenty times faster.

Error counts are more predictive than machine age or last replacement.

A methodology for ranking feature importance in predictive maintenance datasets.

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 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.