Journal of Industrial and Systems Engineering

Journal of Industrial and Systems Engineering

An IoT-Based Multi-Sensor Data Fusion Framework for Predictive Fleet Failure Using a Hybrid Survival Analysis and Machine Learning Approach

Document Type : conference paper

Authors
Aryan Akbarirad 1
Armin Akbarirad 2
1 Master of Industrial Engineering, Faculty of Management and Industrial Engineering, Malek-Ashtar University of Technology, Tehran, Iran
2 Master of Industrial Engineering, Faculty of Engineering, Islamic Azad University, Science and Research Branch, Tehran, Iran
Abstract
Cost optimization and increasing fleet reliability are among the main challenges in urban public transportation management. In this research, temporal failure patterns of critical bus components in various cities of Iran have been modeled and evaluated through analysis of real data collected from intelligent urban fleet systems. Comparative analysis of analytical methods showed that the Decision Forest model with an average accuracy (F1-score) of 89% (compared to Log-rank test 42.3%, stratified Cox 64%, and Decision Tree 85%) demonstrates superior performance in predicting component failures. Results indicate that environmental factors, operation, and utilization methods have significant effects on component lifespan. Accordingly, a predictive maintenance planning framework has been presented which, based on simulation results, leads to a 41% reduction in maintenance costs and a 65% decrease in fleet downtime.
Keywords
Component Survival Analysis
Fleet Management System
Internet of Things in Transportation
Preventive Maintenance
Machine Learning
[1] Elkateb, S., Métwalli, A., Shendy, A., & Abu-Elanien, A. E. B. (2024). Machine learning and IoT – Based predictive maintenance approach for industrial applications. Alexandria Engineering Journal, 88, 298–309.
[2] Zhu, T., Ran, Y., Zhou, X., & Wen, Y. (2024). A Survey of Predictive Maintenance: Systems, Purposes and Approaches. arXiv preprint arXiv:1912.07383.
[3] De Simone, M. C., Lorusso, A., & Santaniello, D. (2024). Predictive maintenance and Structural Health Monitoring via IoT system. IEEE Xplore.
[4] Schwendemann, S., Amjad, Z., & Sikora, A. (2021). A survey of machine-learning techniques for condition monitoring and predictive maintenance of bearings in grinding machines. Computers in Industry, 125, 103380.
[5] Wiegrebe, S., Kopper, P., Sonabend, R., Bischl, B., & Bender, A. (2024). Deep learning for survival analysis: a review. Artificial Intelligence Review, 57, 65.
[6] Thaduri, A., Galar, D., & Kumar, U. (2015). Railway assets: A potential domain for big data analytics. Procedia Computer Science, 53, 457–467.
[7] Ushakov, D., Dudukalov, E., Kozlova, E., & Shatila, K. (2022). The Internet of Things impact on smart public transportation. Transportation Research Procedia, 63, 2392–2400.
[8] Florian, E., & Soulat, D. (2021). Machine learning-based predictive maintenance: A cost-benefit analysis. International Journal of Production Economics, 236, 108114.
[9] Cerrada, M., Sánchez, R. V., Li, C., Pacheco, F., Cabrera, D., de Oliveira, J. V., & Vásquez, R. E. (2018). A review on data-driven fault severity assessment in rolling bearings. Mechanical Systems and Signal Processing, 99, 169–196.