Railway Optical Detection and Obstructions-Tunnel & Station Monitoring

Vortex IoT is an award-winning Wales based SME heading up a consortium that also includes Network Rail and Transport for Wales, Kelos Amey, Balfour Beatty. The company has a proven record in product development and has assembled a highly qualified team of Engineers who are specialist Internet of Things (IoT), wireless mesh networks and Artificial Intelligence (AI). Whilst its partners are a prime rail contractor and integrator and Network Rail Infrastructure respectively (Railway asset owner/ customer). There is an increasing demand for high availability of rail infrastructure and rolling stock assets for rail passenger journeys and freight customers. Remote Condition Monitoring (RCM) of rail infrastructure is essential to maximise the reliability and maintainability of the rail network and is a key enabler for achieving the goal of 'Minimal Disruption and Delay' set by Network Rail’s Capability Delivery Plan. This funding bid is focussed on securing additional funds to accelerate a ‘first mover’ RCM solution to market. The proposed project - RODIO®-TSM (Railway Optical Detection of Obstructions and Intrusions-Tunnel and Station Monitoring) - will deploy and integrate 18 devices in two live rail locations: a live 1.2km rail tunnel and 200m of its either entrance at Melton offered by NRI and a live train station in South Wales (Bargoed) with 200m station area and 100m urban tunnel. The project will user test the intrusion/obstruction detection capability of the LiDAR sensor network, that are wirelessly connected to RODIO edge gateway where the sensor fusion and AI engine resides to process the data and then results (e.g. Notifications, Threat level, …) are pushed to Network Rail Telecom (NRT) cloud server and ubiquitous IP Network. The system uses data fusion and Deep Learning classifiers to identify intrusion and obstruction types and severities aiming for high recall (sensitivity) and high precision against false alarms. This system can accurately detect, differentiate and classify (a) Intrusions – Human and Animal movements (b) Obstructions – Rock fall, tree fall, brick fall, debris fall, (c) Geotechnical asset failures – localised rapid earthworks, flooding, landslides and then sends real time situational alerts to the rail control centre to prompt further investigation as an advisory system. The NRl Product Acceptance Framework defined by Rail Industry Readiness Level (RIRL) is a vital indicator of our product maturity. The RODIO-TSM project will advance the current position of the RODIO® product to RIRL8.