Condition Monitoring Maintenance in Train Bogie:: A Low-Cost Acceleration Sensor Proposal

In this work, a low-cost system to measure bogie vibrations using a set of ADXL355 accelerometers and a Raspberry Pi 4 for relatively fast data collection is proposed. Due to safety reasons, preventive maintenance is performed over bogies without knowing the exact condition, being based on operations time and/or distance. This procedure does not tackle the optimal moment or the effective bogie condition. With low-cost sensors, this process can be optimised. As a result, significant cost savings can be achieved (in our case, rolling stock and railway track conditions). Experimental results are presented, allowing for live monitoring and the acquisition of measurements in trains. Accelerometer sensors were installed at different suspension levels in the bogie: secondary suspension and passenger car suspension, making it possible to monitor events at both levels. This work shows that exact time to perform maintenance depends on the distance recorded by each bogie, however, mechanical endurance is not the same. It is also demonstrated that the ADXL355 sensor is suitable for use in condition control systems, or prediction systems, for monitoring bogies and railway track maintenance status.