As the Chair of the Global Partnership on Artificial Intelligence (GPAI), India successfully hosted the GPAI Summit from December 12 - 14, 2023, at the Bharat Mandapam in New Delhi. 

The summit was attended by almost 22,000 individuals, with more than 15,000 AI enthusiasts participating electronically. Furthermore, other emerging companies and renowned technology enterprises showcase their artificial intelligence applications and solutions at the Global AI Expo. 

This week, we will delve into the Rail Track Monitoring and Alert System (RTMAS) from C-DAC (Thiruvananthapuram).

The collaboration between the Centre for Development of Advanced Computing (C-DAC) and the Konkan Rail Corporation Ltd. (KRCL) marks a significant milestone in railway safety technology. 

Through a Memorandum of Understanding (MoU), these two entities have embarked on a mission to develop a cutting-edge Rail Track Monitoring and Alert System (RTMAS) explicitly tailored for vulnerable regions. This strategic partnership underscores a shared commitment to innovation and safety within the railway industry. Combining C-DAC's expertise in advanced computing with KRCL's deep understanding of railway operations, the initiative aims to revolutionize track monitoring, enhancing the safety and reliability of rail transportation networks in susceptible areas.

Rail Track Monitoring and Alert System

Image processing is used to diagnose rail track deformations using Machine Learning (ML)/AI approaches. RTMAS is a cloud-based server solution designed to monitor the susceptible sections of railway tracks for occurrences such as tree falls, rock falls, soil slips, or any other obstructions. The system is founded on the active vision paradigm, which involves modifying the perceptual elements of biological vision to suit artificial systems. 

Alarm notifications

Advanced machine learning and AI algorithms are created and combined to handle image processing, image classification, and alarm notifications. The system employs cameras and ML / AI algorithms to surveil the susceptible cuts along the Konkan train lines, detecting any obstructions on the train track. It notifies the relevant engineers, resulting in enhanced safety and efficiency in transit along the Konkan sector. The project aims to create, build, deploy, and test the RailView module and RTMAS server-side software for trackside operations. 

RailView module

The RailView module, positioned adjacent to the railway track, can integrate several cameras to offer 180-degree monitoring. It comprises a high-speed system for processing and analyzing images. The RailView system is a dedicated hardware platform designed to execute image processing algorithms in real-time efficiently. The crucial design characteristics include parallel processing, multi-tasking, and easy configurability. The GPU-based platform integrates multi-core processors to deliver a fast image-capturing and processing solution. 

Specialized Digital Signal Processing

Specialized Digital Signal Processing (DSP) methods handle complicated and advanced picture applications, such as machine vision. After preprocessing, the recorded images are transmitted to the RailView High-speed AI/ML engine. AI will be utilized to analyze images and diagnose the status of rail rails. Once it detects any deformation, the identical photos will be transmitted to the central server via the 4G network to the RTMAS server.

RTMAS server-side software

A secure and fast communication interface allows the RTMAS server-side software to register the trackside RailView device and retrieve real-time track images. The RTMAS cloud server has a high-performance GPU that utilizes advanced machine learning and artificial intelligence algorithms to process images and provide alerts efficiently. The AI/ML-powered image processing algorithms analyze this data, identify the object/event on the track, and categorize the event. Alert messages are automatically created and distributed to relevant officials on any abnormal occurrence detected at the trackside. 

Regular and tailored reports

The RTMAS server software additionally offers regular and tailored reports to its users. The RTMAS server software provides a scalable solution for monitoring the rail lines for abnormal occurrences at the trackside. The equipment has been successfully installed and is operational at a susceptible area along the Konkan route. The ongoing operations include the final round of joint tests, evaluation, and enrichment of the learning algorithm. 

Features

The Konkan Railway traverses a challenging topography, with the Western coastline of India on one side of the track and the Western Ghats, a lengthy mountain range, on the other. The strata where the line is built consist of lateritic soil, a mixture of pebbles and soil, and jointed basalt rock. This system will enhance the safety of trains and passengers on this route. Timely alert messages sent to officials in near real-time facilitate prompt action and guarantee a secure and dependable train journey over the Konkan route.

Conclusion

Implementing the Rail Track Monitoring and Alert System (RTMAS) on the Konkan Railway represents a significant advancement in ensuring the safety and reliability of train operations in challenging terrains. With its intricate topography and varying soil compositions, the Konkan route poses unique challenges, which the RTMAS effectively addresses. This system enables swift responses to potential risks by providing timely alerts to officials, ultimately enhancing passenger safety and ensuring a dependable journey along the Konkan Railway.