Newcastle-based engineering firms 4Tel and 4AI Systems (4AIS) have automated the rail corridor intrusion monitoring process by applying proven rail-specific artificial intelligence (AI) to create an advanced 4AIS Hazard Detection System (4HDS).
Safe and reliable operation of a driverless train system requires the rail corridor to be safe from unauthorised intrusions by people, vehicles and large objects that could impact the safety of train operations.
For example, the driverless Sydney Metro rail corridor is regarded as secure because it comprises platform screen doors at stations plus a rail corridor consisting of a tunnel and elevated track on a viaduct.
Together these components provide a significant physical barrier to deter unauthorised intrusion. However, ensuring security of the rail corridor when converting a traditional ‘at grade’ railway line to driverless operations presents a significant challenge.
The recent emergence of Artificial Intelligence (AI) classification technologies has resulted in rail authorities considering this new technology as a solution to the inefficiencies of current corridor intrusion detection (CID) systems.
Traditional CID systems require extensive field infrastructure and through life maintenance, resulting in high capital and operational expenditure (CAPEX/OPEX) and the possibility that field maintenance could impact train operations and present a workplace safety risk for the maintenance personnel.
A centralised AI solution can however detect and classify hazards with less field infrastructure resulting in CAPEX and OPEXS savings.
From a vantage point in the corridor, which may be fixed or an incident drone, information is collected for real-time remote processing to calculate the risk profile probability of an intrusion event based on the location, classification, direction of movement and proximity to the rail danger zone as pre-determined in scoping.
AI can assist the rail authority in minimising the impact of an unauthorised intrusion impacting train operations by enabling the rail operator to respond accordingly to the risk profile of the detected hazard.
The solution can automate the administrative alarm management process for network controllers to ultimately reduce the likelihood of human error, and provide an interface to the train control system to stop trains under agreed scenarios.
Since 2016, the 4Tel and 4AIS teams have been working with the Robotics Laboratory of the University of Newcastle to develop rail-specific AI algorithms for automating the task of detecting hazards in the rail corridor under 24/7 conditions.
Through dedicated research and development activities, it has taken a specialised team years to develop and apply advanced AI and machine learning algorithms specifically for automated rail hazard detection.
This rail-specific capability is now in use with both autonomous freight trains and in a large metro network, and is demonstrating how AI can assist in providing timely information for operational decision making.
This specific AI expertise in hazard detection is combined with 4Tel’s years of experience in tracking trains and assets in the corridor, and the ability to integrate control system information into a final solution.
The 4Tel and 4AIS hazard detection systems are proven and customisable to meet the needs of rail operators for a specific corridor hazard detection solution.