There is a LOT of data in rail. For data to be useful however, it needs to be trustworthy and accurate, it needs to be in a standardised format for meaningful metrics and reference points, but most of all it needs to be accessible.
What’s the current situation?
The Joint Rail Data Action Plan, published by the government in August 2018, emphasised the need to improve the quality and openness of rail data to enhance collaboration between the rail and tech sectors. Since then, the industry has made some positive strides in this regard, with initiatives including:
- The establishment of the Rail Data Council in 2019 (although after achieving its objectives, dissolved in 2023)
- The development of the Rail Data Marketplace, a platform to facilitate sharing of industry data with the aim to foster innovation and create customer-focused products
- Publication of the Rail Technical Strategy (2020) which outlines out how data can help realise industry vision around decarbonisation, optimisation of train operations and enhanced design to improve reliability in rolling stock and fixed assets
- Predictive maintenance such as AI-driven sensors on tracks and trains can detect wear and tear before failures occur. Although this is relatively new, this tech is already employed by Network Rail, but we can hope to see greater benefits to the industry from this as the technology evolves.
There is, however, still a lot more to do.
Unlocking the full potential of rail data also requires a cultural shift within the industry. Data should be viewed not just as a by-product of operations, but as a strategic asset that can drive better decision-making and innovation. Encouraging open collaboration between public and private stakeholders, including operators, suppliers, and tech companies, is key to breaking down silos. A shared vision of interoperability and openness will ensure that data delivers tangible outcomes for passengers, operators, and the wider economy.
Competition between rail operators, retailers and other rail suppliers means data, particularly passenger data, is often disparate, fragmented and exclusive, meaning individual actors see only part of the puzzle. Renationalisation of the train operating companies may help somewhat but would have no real bearing on data owned by private sector suppliers.
Integration of legacy systems is required to enable a seamless flow of information necessary for real-time data analysis and operational efficiency however, aging infrastructure relying on outdated systems can make it challenging to integrate modern digital solutions.
There are also new challenges brought about by the digitalisation of rail services. Managing vast amounts of data, from various sources, requires robust governance frameworks to standardise formats, ensure data privacy and establish clear protocols for data sharing. Recent incidents, such as the cyber-attack on Transport for London (TfL) that compromised customer data, also highlight the pressing need for robust cybersecurity measures in a more digitalised industry.
Why we’re excited about the advent of AI in Rail…
AI represents the next significant technological leap forward and with it, potential for huge gains in productivity, whether through enhanced services or available capacity. Here are just a few examples:
1. Smart Ticketing & Demand Forecasting
AI can analyse historical travel patterns to optimise pricing and ticket availability via dynamic pricing mechanisms. Why is this a good thing? Similar to airline pricing models, dynamic pricing will help balance demand with capacity, improving industry revenues and making fares fairer.
2. Passenger Flow and Station Management
AI-powered CCTV and sensor data can monitor passenger movements in stations. There are already trials underway in major UK stations such as London Waterloo and London Euston but widespread implementation could help manage crowding, improve safety and optimise train dispatching by using AI to predict crowd build ups and suggest appropriate interventions such as gate closures or staff re-deployment.
3. Real-time Service Operation
AI can be used to analyse huge quantities of live data from multiple sources for example weather, passenger numbers and train locations. This analysis enables dynamic rescheduling, reducing delays and improving efficiency. Effective, instant communication with passengers (for example through push notifications) is key to maximising the benefits of this.
4. Data Sharing & Open Innovation
Platforms such as Rail Data Marketplace (RDM) offer widespread accessibility to rail data. Users have the ability to publish, share or consume real-time and historical data. This opens up opportunities for third party start-ups and app developers to create new rail tech solutions, driving positive change in rail.
5. Environmental Sustainability
AI can optimise energy use in trains by analysing speed, braking, and power consumption. By analysing loading data, AI can calculate load dependent acceleration profiles and adjust coasting distances to reduce carbon emissions and operational costs. However, there is a trade off between energy efficiency and timetable adherence, therefore using this functionality in conjunction with 1.) Demand Forecasting and 3.) Real-time Service Operations could offer an optimum allocation considering key factors of demand, capacity, continuity of service and energy consumption.
Thoughtful and considered implementation of the 5 themes in the Joint Rail Data Action Plan, particularly data standards and quality and data transparency, would provide fertile ground upon which these exciting seeds of AI transformational developments can flourish.
