A MapReduce-Based Ensemble Learning Method with Multiple Classifier Types and Diversity for Condition-Based Maintenance with Concept Drifts

Condition-based maintenance in Industry 4.0 collects a huge amount of production datastreams continuously from the Internet of Things attached to machines to forecast the time when to maintain machines or replace components. However, as conditions of machines change dynamically with time owing to machine aging, malfunction, or replacement, the concept of capturing the forecasting pattern from the datastream could drift unpredictably, so it is hard to find a robust forecasting method with high precision. Therefore, this work proposes an ensemble learning method with multiple classifier types and diversity for condition-based maintenance in manufacturing industries, to address the bias problem when using only one base classifier type. Aside from manipulating data diversity, this method includes multiple classifier types, dynamic weight adjusting, and databased adaption to concept drifts for offline learning models, to promote precision of the forecasting model and precisely detect and adapt to concept drifts. With these features, the proposed method requires powerful computing resources to efficiently respond to practical condition-based maintenance applications. Therefore, the implementation of this method based on the MapReduce framework is proposed to increase computational efficiency. Simulation results show that this method can detect and adapt to all concept drifts with a high precision rate.