Design of a smart real-time condition monitoring and predictive maintenance system for railway traction motors

Abstract

Railway traction motors are critical components in locomotive propulsion and are continuously exposed to harsh thermal, electrical and mechanical conditions that accelerate insulation degradation and increase the risk of unexpected failures. Conventional maintenance practices at the National Railways of Zimbabwe (NRZ) rely mainly on periodic inspections, which often fail to identify early-stage faults. This study presents the design and implementation of a low-cost IoT-based real-time condition monitoring system for railway traction motors. The developed prototype has a built-in ESP32 microcontroller combined with MPU6050 sensors for vibration and temperature measurements and ACS712 Hall effect sensors for measuring the current value. The system is capable of collecting data in real-time mode and processing the obtained parameters with embedded health assessment and threshold-based fault detection algorithms. The faults are indicated using LED lights, a buzzer and an emergency isolation relay. The experimental results showed that the system is able to detect abnormal operating conditions and give warning signals to prevent serious damage to the motor. However, despite the fact that it was impossible to develop a machine learning-based predictive model owing to the lack of fault datasets, the created system is capable of providing a scalable base for the further development of intelligent maintenance systems.