DIGITAL RAILWAY

Peter Prater of Hexagon’s Safety, Infrastructure & Geospatial division explains why UK Government innovation spending should be directed towards safety transformation

As the UK Government pledges £7.6 million towards rail innovation, now is the time to seize upon digital transformation across UK rail to improve railway safety and efficiency. This is an opportune time for the Government to invest in UK railway infrastructure with passenger and freight traffic expected to double by 2050, making safety and security assurances for rail organisations more important than ever. Keeping UK railways safe can be a complex ask, with rail operators routinely facing a wide variety of hazards that must be tackled, from security incidents and asset failures to weather-related events, including landslides, floods, lightning strikes, and even fallen leaves.

While rail still remains far safer than other forms of transport, there were 20 recorded fatalities on the UK’s rail networks and 342 near misses with pedestrians at level crossings between 2020 and 2021, according to a report by the Office of Rail and Road. This was the greatest number since the report was initiated in 2002 / 2003. As passenger and freight travel increases, so too will the number of accidents unless safety innovation measures are enacted.

Fortunately, new technologies are emerging that can bolster the safety of trains, tracks, and the entire operational environment. To this end, the UK’s railway operators should introduce emerging solutions that can improve safety procedures and mitigate risk across the country’s railway networks, namely digital twins of railways featuring artificial intelligence that can enable smarter monitoring and maintenance.

A salient aspect of protecting people and freight is visibility into the environments that make up the national railway network. This can be accomplished through monitoring and detecting anomalies in network conditions that may indicate upcoming or present risks. This is no small task with railway operators having to conduct continuous inspections and to monitor any faults in the track, the internal functioning of trains, crossroad maintenance and more. This results in deep pools of data, in which potential risk indicators might lie – and, traditionally, this data had to be manually sifted through and analysed. Not only is the manual process inefficient and time consuming, but it is prone to information being lost through human error or blocked between organisational siloes, leading to delays.

The analysis and communication of such data has been revolutionised by the emergence of integrated transport network information systems that collate all asset and spatial data, including live sensor feeds, into a single source instead of multiple separate databases. These systems can help guarantee that important data is always up to date and easily accessible for staff across the organisation who are tasked with monitoring issues on the track. When hazards are detected in the network, railway operators can proactively solve risks without needless delay.

Railway network monitoring can also be enhanced by visualising this information in a 3D model of the network and its adjacent infrastructure. This forms an identical digital replica of the whole network and its features, including tracks and bridges, but also specific details, such as benches, rubbish bins and trees. When connected to the integrated data, these models provide deeper insight into specific areas and pressing issues in the network.

And the combination of digital twin models with AI-powered automation means that the system can automatically flag any variables that could affect railway network safety, such as high congestion. This provides operators with the chance to mitigate any potential issues before they result in risk and disruption.

One example is the combination of AI with laser scanning technologies for railway infrastructure mapping. A partnership between Network Rail and Innovate UK is a perfect use case of how these technologies can be combined – and where Government investment can be spent. This technique uses laser scanning in parallel with AI to automate the collection and analysis of railway data.

The mapping of railway networks is key to ensuring there are no nearby objects that could endanger railway tracks or the trains themselves. Rail operators continuously collect data from railway tracks and the surrounding environment to gain visibility into the spaces between trains and their immediate infrastructure. This data can guide safe clearance between the trains and the edges of nearby objects. Traditionally, this data has been collected and analysed manually, which is a time-consuming task that could stretch out over months or even years.

By creating 3D models through laser scanning and then leveraging AI, the software can analyse the data, identify different structure types and perform measurements on nearby objects to the railway track. The AI models automatically process unstructured data and accurately output target objects, for example nearby trees or infrastructure, for further assessment. The AI can then detect and alert any potential issues to operator teams. This allows for a more rapid identification of risks to the network, and also helps human teams direct their attention where it is most needed, improving railway security across the entire network.

Over the past few years, ScotRail has been involved in a number of initiatives which aim to break down stereotypes and attract more disabled people, women, and members of the Black, Asian and Minority Ethnic community into the railway. As part of ScotRail’s ambitious early careers strategy, the train operator has set itself the target of increasing the number of apprentices tenfold by 2025. To help achieve this, ScotRail took on 16 young people in Modern Apprenticeship roles during 2021, the highest intake of apprentices in the past five years. Among the new recruits was Ross Henderson, ScotRail’s first deaf Modern Apprentice.

In addition to monitoring and mapping infrastructure to maintaining railway safety for both passengers and freight, physical security is also an important factor in guaranteeing network safety. Physical risks include that of trespass and intrusion of people to the railway network with tunnels, terminals and other features acting as potential targets to theft or terrorism.

To guard against physical threats, railway operators are required to detect, assess, and respond to issues on a swift and continuous basis. Once detected, these risks must be addressed in coordination with third-party organisations – such as the emergency services – within the lifecycle of incident management. An integrated approach to security and surveillance where 3D surveillance systems are combined with security, dispatch, and collaboration platforms is the best way of accomplishing this.

By combining fixed and mobile-sensor technology with AI-powered incident management platforms, next-generation monitoring can detect security breaches across the railway network. This enables rail operators to immediately flag any risks in real time to the appropriate emergency response organisation. A rapid response is key to stopping or mitigating any physical threat to the railway network.

As traffic on the UK’s railways looks set to increase exponentially – and especially as the nation strives to make greener choices – it is pivotal that investment is directed towards security and safety measures that will allow for greater efficiency across the nation’s railway networks. To achieve this monitoring, mapping and emergency response are integral, but all these processes rely on analysing and acting upon vast quantities of data. By evolving technology to generate actionable insights, railway safety can be improved across the board.

is Managing Director of Hexagon’s Safety, Infrastructure & Geospatial division, UK