Improving Rail Safety With Autonomous Trespass Detection

Network Rail and Innovate UK released the Innovation in Railway platform end and edge technology funding call through the Small Business Research Initiative (SBRi) in December 2019. It is known that trespass on the railway causes a significant disruption to the operation and thus, the general public. Trespass can happen for a variety of reasons, for example, short-cuts, fare evasion, vandalism, suicide and theft. Figures recently published by Network Rail reveal that within the first month of lockdown alone, passenger and freight trains were disrupted 1,024 times which resulted in over 380 hours of delays, most of these likely to be carrying essential medical supplies and supermarket produce. Notwithstanding the operational delays, the estimated cost of trespass across the UK Network is more than £500million, however, as not all incidents are detected, the actual cost is expected to be much greater. Therefore, this competition aims to build a system capable of not only detecting trespass but to act as a deterrent to reduce the impact of these incidents on the railway.

As Zircon Software has a wealth of experience in developing high-integrity, safety-related applications within the rail industry, we felt best placed to submit a proposal to meet the aims of the competition.

The awarded consortium consists of Zircon Software, Pragmatex, and The University of Southampton who in conjunction will develop the Trespass Identification and Deterrent System (TIDS).

TIDS is a fully autonomous, off-grid CCTV solution designed to run autonomously and detect and deter trespassers on the network. Its unique modular design allows it to be installed in a variety of environments, take advantage of the existing infrastructure where possible, and be deployed in close proximity to the incident ‘hotspots’ identified by Network Rail.

The system uses unique video analytics built upon from previous Zircon’s projects. The most relevant being the Platform Train Interface (PTI) camera which was installed at London Victoria Underground Station for 6 months in 2017. The PTI camera detected passenger interaction and intrusion from the platform to the track. Other notable uses of Zircon’s video analytics have been for asset detection, location accuracy and passenger drag detection, all of which have been tested on the railway.

The University of Southampton will research the novel deterrent aspect to establish the most effective to discourage trespass. However, previous studies conducted in Europe of basic sound deterrents reduced trespass by up to 44%. The system also takes into account UK GDPR, whereby incidents are detected, but people are not uniquely identified, and data is only stored for a necessary amount of time. Co-operation with the British Transport Police (BTP) could see the system used for evidence in prosecutions, although that is not the initial purpose.

Further to the initial development phase, TIDS will be installed at Reading West and West Drayton both of which support the new Crossrail route from East London to Reading.

At these stations, TIDS is intended to:

  • reduce the number of trespassers on the Network,
  • mitigate the operational impact of delays,
  • improve the customer experience and
  • empower the station staff to conduct their role to the best of their ability, without focusing on incidents outside of the station.

In addition, TIDS will provide useful statistics to the operators and in future, could be used to analyze the exact nature of trespass.

All being well, and notwithstanding the current COVID-19 situation, you should see TIDS on the railway by the end of October 2020. Following the success of the 12-month demonstration period, we foresee the potential for TIDS to benefit a variety of other areas on the UK Network.

It is also worthwhile noting, that whilst Zircon is building the TIDS system, they are developing a similar system, the Smart Engineered Big Sister (SEBS) which again utilises video analytics but specifically to tackle challenges around platform security. The competition, also funded by SBRi, aims to detect scenarios such as anti-social behaviour, loitering, overcrowding and miscellaneous abandoned objects. The long-term vision is that both SEBS and TIDS may be integrated to provide an all-encompassing intelligent security system covering the platform, the perimeter and individual track-side locations.