Railway Technology UIC publishes study on the potential of artificial intelligence in railway companies

Related Vendors

McKinsey & Company Inc.

FAULHABER Drive Systems

Taiwan Machine Tool & Accessory Builders' Association (TMBA)

Wuxi InfiMotion Technology Co., Ltd.

Greater use of artificial intelligence could bring significant benefits to railways worldwide, according to a recent analysis by the UIC in collaboration with McKinsey.

In a new study, the International Union of Railways (UIC) together with the management consultancy McKinsey examined the possibilities and impacts of increased use of artificial intelligence (AI) in railway companies. To this end, the initiators of the analysis surveyed eleven railway companies in Europe and Asia. In addition, they conducted detailed interviews with 15 selected companies in the railway sector and OEMs. The participants finally summarized the results in the study "The journey toward AI-enabled railway companies", presented in February 2024.

The report describes use cases that are already in use on a larger scale or that have the potential for introduction. Robotics was not considered, as it focuses mainly on applications of machine learning (ML) and deep learning (DL). In addition, the authors examine the influence of the ecosystem of various stakeholder groups and demonstrate how the use of artificial intelligence throughout the railway sector could create added value. They focus on passenger transport.

Artificial intelligence holds great potential for railways

First, the report identifies four core areas along the railway value chain where AI could support different business activities. These are areas that are directly related to the provision of the product "railway", i.e. the travel experience. These would include, for example, the deployment planning of train crew, the maintenance of rolling stock, or the response to irregularities in operation.

According to the study, artificial intelligence also has great potential in terms of infrastructure, such as planning, route management, or maintenance. Passengers could also benefit, for example through improved communication or simpler booking processes. A final area would be activities that do not directly involve operating the railway, such as human resources, finance, or IT.

One in every four railway companies is using AI on a large scale

According to the study, 25 percent of railway companies are in an early phase of implementing AI, but have not yet started pilot projects. Thirteen percent are experimenting with one or more projects and 38 percent have made initial deployments on a large scale and are building up internal competencies. Finally, 25 percent are already using artificial intelligence on a larger scale and have their own resources for this.

Overall, the study identifies about 100 potential use cases for AI; 20 of them are already in use at different stages in railway companies. AI is currently widespread in predictive maintenance, energy management, and real-time information in intermodal transport. Newer areas of application include assisted automatic train operation according to GoA (Grade of Automation) 3 or digital twins for network infrastructure. In each case, the study refers to corresponding reference projects, such as the digital maintenance of the suburban train in Munich-Steinhausen (Germany).

The use of AI pays off financially

According to the report, increased use of AI would also have financial impacts. Thus, artificial intelligence could potentially create a value of around 700 million euros (approx. 757 million USD) per year for a company with a turnover of 5 billion euros. Globally, AI has the potential to generate 13 to 22 billion US dollars in value creation.

However, railway companies wishing to use AI face challenges. This concerns the availability and quality of data, as well as regulatory considerations such as the General Data Protection Regulation, the lack of standards, or the fear of loss of expertise. Moreover, the railway industry is traditionally risk-averse when it comes to new technologies. To properly use AI, the authors of the paper propose six points on which the railways should concentrate when implementing: a strategic roadmap, AI-specific roles and capabilities, the operating model, technology, data management, and implementation and scaling.

Finally, the authors address the potential risks associated with artificial intelligence and shed light on the role of the ecosystem of OEMs, technology companies, research institutes, politics, and industry associations. They all play a role in the use of AI.

In summary, the initiators of the study find that Artificial Intelligence has the potential to significantly change the way railway companies plan and deliver services along the value chain. The full report is available here. (se)

Railway Technology.

FCH2Rail: Tests of Hydrogen Train in Spain Successful.

Rail technology

World record: Hydrogen train travels 2,803 kilometers on a single tank

*The author is a freelance journalist ands writer 

This article was first published on our sister brand 'NextMobility' (German Edition), Vogel Commucniations Group