National Rail looking for security, maintenance automation solutions
Between £350,000 and £500,000 is up for grabs for a company of any size that can help Network Rail improve security and maintenance with data analytics.
Network Rail is a non-departmental public body of the Department for Transport charged with maintaining and operating the UK’s rail infrastructure. The firm oversees 335 kilometres of track and maintains over 600 tunnels.
The Small Business Research Initiative has allocated £3 million towards automation technologies that improve how the rail network operates.
Currently to conduct maintenance in tunnels, humans need to be sent in to investigate for wear and tear or malfunctioning parts. However, this is an expensive and time-consuming process that can also pose risk to the workers, and using new technologies to help automate that process is a priority.
Not only is health and safety a clear concern for Network Rail, but incidents inside stations cause significant delays for the whole network. To mitigate these risks Network Rail are seeking innovative ways it can use data analytics to improve safety and reduce delays.
- The key challenges tackled by submissions should be;
- Reduce the need for manual examinations
- Remove human-centric bias
- Inspect both surface and sub-surface defects
- Use machine learning to identify and track defects
- Remove the need for human review and analysis
- Align with Network Rail IT systems and condition-marking systems
- Allow engineers to view and analyse results
Network Rail state that a successful project should: “Reduce the time taken to complete examinations through improved automation and increased repeatability and reproducibility of data.”
Network Rail Automation
Network Rail are also seeking to create a security system that can identify security threats such as objects or bags that have been abandoned in the station. The desired security camera system should also be able to identify ‘undesirable behaviour,’ as well as station overcrowding.
Successful systems will be able to handle video feeds from 30 cameras placed in different station environments. It must be able to process video resolutions of 1080p and be able to operate 24/7. The system should be able to work without external connectivity to cloud infrastructure.
Applicants must demonstrate the analytic capabilities of their offering on pre-recorded data in the first phase of their submission, followed by a live trial in a station during the second phase.
The completion process is facilitated by the Small Business Research Initiative which is open to organisations of any size that are based in the UK, EU, USA and Canada. The competition officially opens on the 16 of September and the closing deadline for submissions is the 6 of November.