Real-time implementation of optical sensor on lab rig model for speed estimation

The Indian railway traffic is increasing and is expected to increase more in the future, thus making safety a critical issue to focus on. However, there is very little work on smart monitoring with respect to railway infrastructure in India. This paper analyses this problem and proposes an optical sensor-based smart monitoring solution for railways. A fibre Bragg grating-based sensor is deployed on the sleeper of a lab train model, and an effective algorithm is proposed to successfully detect the train speed and the axle count using the sensor data. The train speed and axle count are fundamentals in the process of detecting railway faults such as derailments, wheel flats and corrugation or unbalanced loads. For a moving train, the instantaneous train speed is not readily available for monitoring purposes and thus the system proposed here focuses on analysing the safety of the moving train instantaneously by calculating the train speed and counting the axles of the train at a point. A lab train with one coach and two axles is moved to and from on the track over the sleeper which holds the sensor. Here, three iterations of to and from movement of the lab train are conducted to arrive at accurate results. For each of the iterations, four axles are detected-two when the train moves forwards and two when the train moves backwards. With the data collected and the algorithm proposed, the lab train speed is calculated to be 8.2739 km/hr which matches the known actual speed of 8.2 km/hr.