With the help of an implementation-oriented AI roadmap, Deutsche Bahn has ambitious goals for future reliability and efficiency says Dr Bernd Peper, head of public sector at Sopra Steria Next Germany.
Historically, railway companies have always used new technologies to improve operations and performance. With the advent of artificial intelligence (AI), this development is continuing. This is because AI, and generative artificial intelligence (GenAI) in particular, is helping companies to improve planning and performance along the entire value chain.
Figures from a recent study by UIC and McKinsey show just how much. According to the study, railway companies worldwide can save between US$13-22 billion per year by using AI. The enormous potential of AI is also demonstrated by a study entitled ‘Navigating the AI Era’ by Sopra Steria Next. The experts estimate that the global AI market will account for 10% of global IT spending by 2028 and will grow three times faster than the overall technology market.
The use of AI in European rail
A look outside the box already reveals value-adding examples of AI use in European rail transport. At SBB AG in Switzerland, for example, demand forecasts are created with the help of AI. This helps to provide the right number of wagons. An AI-supported journey profile also enables precise time and speed calculations for the railway.
At SCNF AG in France, an AI solution ensures that capacity utilisation and revenue for high-speed trains (TGV) are optimised. The cognitive performance of staff will also be analysed using AI in the future. By measuring eye movements, perspiration and heart rate, stress and fatigue among staff will be measured in order to minimise accidents.
A number of AI use cases are also in use at Deutsche Bahn. These range from the Railmate feedback platform and peak spotting, an intelligent capacity utilisation control system, to delay forecasts based on artificial intelligence.
With its turnaround plan S3, which was launched in September 2024, Deutsche Bahn has launched a restructuring programme to become more punctual, reliable and economical. Digitalisation and the use of AI in the right places are helping to achieve these goals. In addition, the spread of GenAI is opening up new areas of potential every day. In order to utilise these systematically, DB Fernverkehr AG – a division of Deutsche Bahn – has developed a structured approach to identify AI fields of action for the coming years.
Creating an understanding of the market and competition
The first step was to gain a better understanding of the market and the use of AI. To do this, the company took a close look at the AI use cases employed by other players, from Rolls-Royce and Deutsche Bank to Unilever and Coca-Cola.
At the same time, the potential benefits of these use cases for DB Fernverkehr were categorised, i.e. whether the use cases contribute to punctuality, efficiency and customer satisfaction, for example. The company also analysed twelve competitors in the mobility sector to find out which use cases are already creating value there.
This market and competition analysis resulted in eight overarching fields of action for the use of AI, each with the most promising application clusters. These eight areas of focus included operational efficiency, customer focus, next-generation software development, human resources, predictive analytics, knowledge management, data management and cyber-security.
Prioritisation of the most relevant fields of action for railway operations
Although the various fields of action from the market analysis promise great AI potential for a wide range of industries, it is relevant to determine the areas that are realistically feasible for railway companies and offer the greatest possible added value.
In addition to the general market perspective, the second step was therefore to look at the specific needs of railway companies. To this end, interviews were conducted with experts from research and industry, as well as discussions with internal and external partners, in order to identify specific application clusters that will have the greatest relevance in the next 24 to 36 months.
The fields of action identified in the interviews include, for example:
- Automated scheduling (operational efficiency). This application cluster is primarily concerned with increasing the punctuality and efficiency of decision-making processes based on AI.
- Predictive maintenance (predictive analytics). Predictive maintenance of trains, tracks, points and signalling systems enables maintenance work to be planned precisely and carried out when it is necessary. Analysing the data obtained in this way leads to more efficient use of resources, lower maintenance costs and less downtime.
- Automated creation of marketing content (customer focus). Marketing content is developed on the basis of target group data and market trends. Using AI models, customised texts, images and videos can be generated and optimally tailored to the preferences of the target groups. The automated process enables marketing departments to produce content faster and on a larger scale or to optimise existing content.
Based on the market analysis and expert interviews, a list of recommendations for the most relevant AI fields of action for the railway was drawn up. Although the market sees and realises AI potential in a wide variety of fields, clear recommendations can be made for railway companies for focus areas that are based on relevant key metrics such as punctuality, quality, EBIT and customer satisfaction.
From field of action to company-wide scaling
Based on these fields of action, the AI in long-distance transport programme is currently working with the specialist departments to identify specific and feasible use cases. In our opinion, a holistic approach is crucial for this, one that keeps feasibility and technical benefits as well as regulatory aspects in mind.
An overarching technical target image for each field of action, such as in the area of predictive maintenance, which affects many departments across the company, creates the basis for realising more complex applications, avoiding isolated solutions and breaking down silos. It offers a common orientation under which the specialist departments, IT and the AI department can unite and work towards in a targeted and coordinated manner.
In addition, the right framework conditions must be created to ensure sustainable and economical scaling. This includes:
- Clear responsibilities and processes
- Efficient management and implementation
- Compliance with data protection regulations
- Efficient handling of fragmented data sources
- Continuous improvement of data quality
- Implementation of uniform technical standards
- Training of employees
If implemented successfully, AI helps to optimise processes, develop new business models and strengthen competitiveness.
A new technological era
With the use of AI and GenAI, railway companies are facing a new technological era that can significantly improve efficiency, reliability and customer satisfaction. With clearly prioritised fields of action, such as predictive maintenance, automated scheduling and customer-oriented content creation, companies can create added value along the entire value chain. A holistic approach that guarantees feasibility, interdisciplinarity and scalability is crucial for success. If AI can be implemented sustainably, companies will not only strengthen their operational efficiency, but also their competitiveness in an increasingly dynamic mobility market.