Source Component Shift Detection & Classification for Improved Remaining Useful Life Estimation in Alarm-Based Predictive Maintenance

Abstract

Predicting the remaining useful life (RUL) plays a significant role in reducing the downtime of operational assets and maintenance costs. While various predictive maintenance (PdM) studies have utilized sensor data to calculate RUL, valuable information from system alarms has often been overlooked or marginally used. This paper introduces the first-ever solely alarm-based PdM model for RUL estimation. To demonstrate the effectiveness of the proposed method in real-world production lines, the alarm data from two instances of milling machines, used for producing artificial bone joints, are utilized for PdM model development and testing. Additionally, given the varying operation hours of the machines and also the changes in the size of the produced bone joints, this study demonstrates the importance of pinpointing different subpopulations of the alarm data, called data modality (DM), for improved RUL estimation. As flexible and adaptable manufacturing required in Industry 4.0 results in constant changes in the distribution of the gathered data, this paper presents a systematic and generic approach for identifying different DMs, which results in a significant decrease of up to 25.20% in the mean absolute error (MAE) for RUL predictions on the whole test dataset, and up to 60.50% in the MAE of the RUL predictions for the minority DM in the test dataset.

Authors 4

1. Kiavash Fathi Aachen Chair of Information and Automation Systems for Process and Material Technology

   RWTH Aachen University · ZHAW Zurich University of Applied Sciences

   Affiliation as printed

   Institute of Mechatronic Systems, Zurich University of Applied Sciences,Winterthur,Switzerland,8400

   Chair of Information and Automation Systems for Process and Material Technology, RWTH Aachen University, Aachen, Germany

2. Marcin Sadurski

   ZHAW Zurich University of Applied Sciences

   Affiliation as printed

   Institute of Mechatronic Systems, Zurich University of Applied Sciences,Winterthur,Switzerland,8400

3. Tobias Kleinert Aachen Chair of Information and Automation Systems for Process and Material Technology

   RWTH Aachen University

   Affiliation as printed

   RWTH Aachen University,Chair of Information and Automation Systems for Process and Material Technology,Aachen,Germany,52064

4. Hans Wernher van de Venn

   ZHAW Zurich University of Applied Sciences

   Affiliation as printed

   Institute of Mechatronic Systems, Zurich University of Applied Sciences,Winterthur,Switzerland,8400