Public description The GPAI project uses geospatial data and an innovative positioning system to solve a Network Rail transport challenge Network Rail operates and maintains most of Great Britain's rail network. It is investing heavily in technology to increase capacity. Examples include digital signalling, traffic management and automatic train operation. These technologies require accurate train location to operate. Unfortunately, GPS does not operate well in the rail environment with deep cuttings, tunnels, covered stations and many structures blocking the signal or causing reflections. Even with good satellite visibility, accuracy is very often insufficient to determine which track a train is occupying when there are multiple parallel tracks. Consequently, Network Rail has to use expensive infrastructure-based systems to locate the train, including transponders, axle counters and track circuits. Challenge. The Network Rail Signalling Innovations Group (SIG), is involved in determining and validating the location of these infrastructure systems. The greater the confidence in their location, the smaller the error margin that needs to be built into the control system design and the greater the capacity that can be achieved. To make these measurements, Network Rail currently runs specialist engineering trains which are expensive to operate and, by their presence, reduce the track capacity available for passenger services. SIG wishes to move to low cost survey equipment mounted on the front of passenger trains. However, because of GPS limitations, it is not able to position the train to sufficient accuracy. It is therefore seeking solutions to improve the accuracy of low cost GPS-based location systems to within 1m. GPAI Project: Geospatial Position Accuracy Improvement is a proposed service that improves estimates of train location through repeat runs over the same route. It is particularly suited to passenger trains that make similar journeys on a daily basis. The improvement is achieved through innovative use of the Video Train Positioning System developed by RDS International. VTPS is a technology that locates a train using real-time image processing from a forward facing camera. The project will investigate GPAI feasibility through development of: requirements and business case proof of concept demonstrator technical and commercial routes to market. The project will be led by RDS, supported by Network Rail, One Big Circle and Porterbrook.

Precise train positioning is key to the vision for an innovative railway presented by the Rail Capability Delivery Plan 2017\. GPS is problematic for rail and currently the industry relies heavily on signs, transponders and balises in the track to locate a train at a specific point on its path. These are expensive to install and maintain. Each new application that needs them increases the complexity and management burden of the infrastructure. This project takes a completely new, lower cost, approach. Adapting recent developments in technology for autonomous cars, a train will be able to locate itself without needing any infrastructure equipment. Instead of a physical transponder or balise in the track, the system creates a virtual 'Video Balise' - Valise, which is stored in a forward facing camera mounted in the windscreen. The Valise is 'read' as the camera recognises that the train is passing the stored location. The Valise will unlock the potential for many high value applications which cannot economically be delivered with today's expensive physical positioning infrastructure. The project is being delivered by RDS in collaboration with Nottingham Scientific Ltd, Omnicom Balfour Beatty, First Group and Network Rail. RDS will develop the Valise technology and integrate it with its Video Train Positioning System (VTPS) to provide an autonomous low cost system suitable for deployment in service trains. The technology will be trialled in a real-time demonstrator on a First Group train, integrated with the RDS Driver Support System (DSS). The demonstator will show how the technology would be deployed for platform stopping and 'virtual' temporary speed restrictions. Balfour Beatty and Network Rail will validate the performance of the Valise against existing positioning systems and demonstate its use for remote monitoring and survey applications. NSL will bring its R&D experience in 'virtual balise' technology using satellite postioning (GNSS). With this, the project team will develop a safety approach for integerating GNSS and video virtual balises in order to realise a dependable virtual balise capable of use in safety related applications.