Public Funding for Zircon Software Limited

This project will demonstrate a solution which takes time based data in close to real time and broadcasts off measurment trains ito a cloud and (in parallel) the facility for a measurement train to be managed remotely (whilst being hauled by a manned traction unit), i.e. the operator can be in any remote centre. One operator team can then manage many measurement vehicles around the world. Data quality from unmanned vehicles will improve, crew costs reduced and data access improved. Analysis will be carried out in the cloud; more than one analysis can be used, and the results distributioned. Fault reports, statistics and alerts are taken from the cloud by users with mobile Apps or office systems. More than one display tool can use the data and the users can be anywhere in the world; in an office or on the line side. This combination of features, as well as improving the efficiency of the national railways has the potential to change the way in which this technology is delivered and will allow access to thie type of technology for smaller organisations such as tram and light rail infrastructure operators. It is envisaged that the creation of this rich dataset, and controlled access to it, will act as a stimulus for future system and product development from both within and outside the consortium.

Zircon software has a wealth of experience in developing high-integrity safety related applications within the rail industry. Our understanding of assisting clients throughout each individual RIRL's up to product deployment for well-known organisations such as Knorr-Bremse, Siemens and Petards will be key to our success. In addition, Zircon has experience in working with complex hardware and a number of COTS (Commercial Off-The-Shelf) products specifically designed for rail while our knowledge also covers supporting existing operational systems on the network. Zircon has been involved in Machine Learning (ML) research for over 5 years. In particular much of the research has utilised Video Analytics. After a successful 6 month trial, a previous research project called Platform Train Interface (PTI) Cam, was installed at London Undergound Victoria station. PTI Cam detected passenger interaction and intrusion from platform to the track. PTI Cam forms the basis upon which the SEBS solution can be built. PTI Cam delivered a successful demonstration. Zircon are taking the same approach in this competition and applying it to the scenarios outlined, notably: trespass (short-cuts, fare evasion, grafitti) and theft. The system may inevitably detect potential suicide attempts which is not it's intended purpose, although could be seen as a seen as a secondary use case. TIDS is modular stand-alone off-grid CCTV safety system capable of being deployed in all enviroments and designed to deter trespassers on the railway network. The system is fully automated utilising statistical analysis and machine learning algorithms, parts of which, already proven in past projects. By deploying an automated powerful deterrent device such as TIDs, we will reduce the number of train delays as a result of trespass which will have a direct impact on improving customer satisfaction.

Title: Smart Engineered Big Sister (SEBS) Text: Zircon has been involved in Machine Learning (ML) research for over 5 years. In particular much of the research has utilised Video Analytics. After a successful 6 month trial a previous research project, called Platform Train Interface (PTI) Cam, was installed at London Undergound Victoria station. PTI Cam detected passenger interaction and intrusion from platform to the track. PTI Cam forms the basis upon which the SEBS solution can be built. PTI Cam delivered a successful demonstration. Zircon are taking the same approach in this competition and applying it to the scenarios outlined, notably: Loitering; unusual behaviours and incidents; crowd trend analysis; threat detection and abandoned objects that are not part of station upgrade work. Using PTI Cam as a foundation, Zircon intends to integrate existing individual off the shelf ML detection technologies with a bespoke analysis platform to deliver an automated monitoring and alerting system that meets station control centre needs. Zircon will be utilising existing CCTV Infrastructure and as a result there will be no significant roll out cost to deploy our solution.