Network risk monitoring through IoT-based system for improved capacity and safety in rail operation
Network risk monitoring through IoT-based system for improved capacity and safety in rail operation.
Gopi Chattopadhyay (Lead), Joarder Kamruzzaman (Assoc.), Harpreet S Kandra (Assoc.), Tanveer Choudhury(Assoc.),
Area of Research:
Risk Engineering under INTERDISCIPLINARY ENGINEERING: 091507
Covering Maintenance and Reliability Engineering, Asset Management, Civil Eng, Mechanical Eng, Mechatronics, IT including IoT, Machine Learning, Artificial Intelligence and data analytics.
Contact person and email address:
Gopi Chattopadhyay email@example.com
A brief description of the project:
Rail network is quite complex, especially for lager networks in Australia and India. For an example, Queensland Rail in Australia covers 2,670 kilometre of rail track with over 1000 curves, connecting major coal mines and their distributions to ports. The operational risk of the network is severely influenced by a number of factors including track stability (very hot or cold weather), weather condition, wear limits (lubrications issues and rail- wheel wear), rolling contact initiated cracks (initiation, timely and accurately detection and mitigation using rail grinding), traction and condition of the breaking system and human factors including driver behaviour. Currently safe speed of the rolling stock is determined based on analysis of risks based on limited data and therefore, same speed is enforced along the whole corridor (blanket speed restriction). Current practice causes reduction of operational efficiency and/ or missed detection of potential risks of derailments and accidents. With the advent of low cost sensors and easier deployment, the risks in various corridors of the track can be monitored at higher number of critical points of rail network using IoT based rail monitoring system, enabling operators and maintainers to gather more accurate and precise condition data in real time. In this project, using data through IoT-based system in the entire network as well as weather forecasting data from BOM, a predictive risk model using will be developed based on machine learning technique and artificial intelligence that will allow forecasting of the operating risks in a more realistic and accurate manner and allowing different operating safe speeds at different sections of the whole network resulting in increased operational efficiency and safety along with reduced operational, maintenance and capital replacement costs.