Building resilience on the back of data
The end of an era for SMIS
The Safety Management Intelligence System (SMIS) has long been the backbone of safety data collection in rail. Through collaboration and standardisation, we built an enormous bank of information, in which we could really see the bigger picture. As technology moves on, it’s time to reimagine SMIS for the new era.
Building on recent improvements that simplified the data entry process, and collected better quality data, a new system has been designed. Using feedback from industry, Safe Insights will replace SMIS by the end of March 2025. It uses AI to enable easier user input. The streamlined data model will deliver better insights than ever before. And, crucially, it will provide more opportunity for integration with other tools.
Better data bolsters performance
The Red Aspect Approaches to Signals (RAATS) toolkit is another well-established data tool. It estimates the number of times a signal is approached at red, and counts different types of approaches and factors like train type, or time of day. It’s already a valuable tool. But what does an improved RAATS look like? What new insights would we get by combining RAATS and SMIS data?
A reboot of RAATS will extend the coverage of the tool to most of the network, rebuild it on a new user-friendly platform, and add new functionality for analysing red aspect approaches and delays by train service. This will not only help us target SPAD risk reduction but also find pinch points where delays are being accumulated, supporting safety improvement and performance optimisation.
Safety data hubs: the data you actually want
Our industry works so hard to generate and maintain safety data. It’s a precious resource that some other rail systems don’t have so we should be proud of everyone’s work to create and maintain it. But we could maximise its benefits if access were easier. Even an immaculate bank of information is less valuable if it’s too hard to navigate.
We recently launched a suite of safety data hubs—interactive Power BI dashboards that allow you to cut through the noise. You can access and filter real-time data on the safety issues that are key to your role. The data hubs were created in response to feedback from groups. They wanted more access to the data than quarterly PDF reports. Now with self-service data at their fingertips, users are free to find different approaches to using data to meet their needs. This can help to identify, address, and demonstrate safety problems and find better solutions.
A resilient railway, rain or shine
There is another data-driven tool on its way to support making our railways more resilient. PRIMA, or Proportionate Risk Response to Implementing Mitigating Speeds to Assets, is an operational planning tool for speed restrictions in adverse weather. It’s designed to make sure rail’s response to extreme rainfall events is proportionate. By weighing up safety, cost, and practicality, it helps inform difficult decisions with complicted trade-offs.
It’s forward thinking. PRIMA suggests speed restrictions for an operational route section in advance for different weather scenarios. By taking a data-led scientific approach, it predicts derailment risk and the knock-on risk from service disruption, and combines these with delay cost information to evaluate the benefits and costs of different responses.
This forward-looking use of data is what will truly bring our network into a new technology era. And it underlines the benefits of a data-led approach, informed by operational understanding of the issues involved.
Why is risk modelling the way forward?
Data alone can’t run the railway. It's about applying the numbers in combination with your expertise to make those decisions and draw those conclusions. But what happens when there is no data?
If you waited until the unthinkable happened, you'd have a far smaller data set than if you used modelling to estimate it. How worried should you be about, say, a multi-fatality accident caused by a freight train carrying toxic goods derailing in an urban area? That is an extreme example, we know it’s never happened on Britain’s railway, but we know it's a credible risk. But how big is that risk, and what should be done about it?
Even without the data, having the capability to develop models and use them effectively is the key. So our future activities will continue to include risk modelling to help rail make even better safety decisions.
