GIVING their keynote addresses to the opening session of Railway Forum 2025 in Berlin on September 3, the CEOs of three major European suppliers all turned to one of the conference’s principal themes, the digital railway.
“We should not talk about the digital future, we should talk about the digital present,” said Alstom CEO, Henri Poupart-Lafarge, highlighting how the manufacturer is already using digital processes to optimise rolling stock production, and deploying AI to power its product lifecycle management system.
“We already have what is required to take full advantage of digital technology if we work to together,” he said.
A similarly optimistic note was struck by Hitachi Rail CEO, Giuseppe Marino, who sees digital transformation at scale as one of the key pillars of the railway in the 21st century, enabling it to maintain its position as the backbone of sustainable mobility.
If the sector is to meet its potential for handling twice the number of passenger-km by 2050, when compared with 2015, and cutting CO2 emissions by a factor of eight, “we need a zero-emissions railway that is fully digitalised,” he said.
Siemens Mobility CEO, Michael Peter (above), struck a more cautious note in his keynote address, considering how rail can benefit from what he described as the third digital revolution, the rise of AI following the development of the internet and before that the personal computer.
For rail to make the best use of AI, four key requirements must be met, according to Peter. The first is interoperability based on open application programming interfaces (API). “AI will not fly with 25 signalling systems in Europe,” he said, stressing the need to roll out ETCS more extensively and switch off legacy systems.
Standardised data and cybersecurity were other key requirements, as well as pragmatic regulation. “Huge simplification is needed here,” Peter said. “If we lose five years on AI regulation, I am afraid that this might be the next big thing Europe misses out on,” he said.
“Today rail is still a great steel colossus,” Peter said. “AI can make rail a living organism,” he believes, one where the system can effective heal itself through diagnostics and predictive maintenance, leaving operators free to focus on customer service in order attract passengers out of their cars.
AI development
Dr Daniela Gerd tom Markotten (above), board member for digitalisation and technology at German Rail (DB), saw the deployment of new technology to future-proof the railway as working hand-in-hand with DB’s major programme of infrastructure upgrades, which aims to tackle the root cause of the poor punctuality that has bedevilled DB’s long-distance passenger services in particular.
With long-distance punctuality at its lowest-ever level last year, falling to 52.9% in June 2024, “the citizens of Germany are not satisfied with what we are doing,” she said, noting how the performance of the national rail network acts as a barometer of how Germany is performing in general.
Three projects provided examples of this “ambidextrous” approach of combining the renovation of existing infrastructure with the deployment of new technology, including the construction of the new high-speed train maintenance depot in Cottbus “in record time.”
A second was installing FRMCS masts on the Hamburg - Berlin main line as part of the current upgrade, which will enable real-time data transmission between track and train as well as improving connectivity for passengers as they will also be used by public 5G networks. “That is what real innovation is,” tom Markotten says.
The third example, and one that is most likely to resolve the concerns expressed by Peter, is the DataHub Europe joint venture between DB and Schwarz Digits, which will enable DB to bring its own AI projects into operation more quickly and in a legally-compliant manner.
Approved by the European Union (EU) in July, the new hub is intended to enable AI innovation using data centres and cloud infrastructure based in Europe. Its core component is a highly secure data room that enables companies and organisations to exchange data in compliance with GDPR, the EU regulation on corporate use of personal data.
The new hub will also act as a digital toolbox, providing common standards, specialised toolsets, expert knowledge and scalable computing resources for the development of AI projects. This is expected to accelerate innovation, reduce development costs and promote data-based cooperation between industry, the research sector and public bodies.
The first use case developed by DataHub Europe and now deployed at DB and Schwarz Digits is AuditGPT, an AI-supported digital assistant for the preparation of audit reports, which will be further developed for the wider market.
