Center-Director’s
Message

AI in Railways

Area Lead:
   Prof. Ashwini Nanda – Center-Director, AHRC & Visiting Professor, School of Electrical and Computer Sciences, IIT Bhubaneswar,
   Prof. Partha Pratim Chakrabarti - Ex-Director, IIT Kharagpur, and Professor of Computer Science, IIT Kharagpur
   Prof Manoj Choudhry, Vice Chancelor, Gati Shakti Vishwavidyalaya (GSV), Vadodara

Railway track inspection is vital for maintaining the safety and reliability of India’s 70,000+ km railway network. Traditionally, keymen walk long stretches of track, visually checking rails, fasteners, and joints, a process that is slow, labour-intensive, and prone to human error.

Modern practice requires systematic measurement and recording of key track geometry parameters such as gauge, alignment, unevenness, twist, rail profile, and wear. The health of critical components, including clips, bolts, and other fasteners, must also be monitored. At the same time, the surrounding environment needs continuous assessment to detect any infringements to the moving dimensions (MMD) or schedule of dimensions (SOD), such as encroaching structures or vegetation.

 Regular, high-quality inspection not only safeguards train operations and passenger safety but also underpins predictive maintenance strategies, allowing engineers to identify and address emerging problems before they escalate into failures, disruptions, or accidents.

At AHRC, we have developed an AI-powered Railway Track Inspection System, equipped with advanced sensors for autonomous monitoring of railway track components, the surrounding environment, and the track underlayment.

Our system integrates multiple cutting-edge technologies:

  • Phased Array Ultrasonic Testing (PAUT) equipment to detect internal defects in the rails.
  • 2D LiDAR sensors to measure track geometry, including gauge, alignment, unevenness, and twist.
  • Ground Penetrating Radar (GPR) to assess the condition of the track underlayment.
  • 3D LiDAR and RGB-D cameras to map and monitor the environment for any infringements to MMD or other potential hazards.
  • High-speed, high-resolution camera systems, combined with our custom-built computer vision and AI models, to automatically detect surface defects, missing fasteners, and other anomalies.

At AHRC, our solutions are designed in close consultation with railway professionals—people who have first-hand experience maintaining the backbone of India’s railway infrastructure.

Skip to content