Punctuality, customer satisfaction, and a shortage of skilled workers: These are the key challenges currently facing Deutsche Bahn. How can digitalization, automation, and targeted deployment of artificial intelligence (AI) help master them? As the following insight into Deutsche Bahn’s digital transformation shows, there are many applications where AI can play a role – ranging from routine processes to AI-driven process optimization right through to AI-driven analysis of social media channels.

These days, rail is more popular than ever. Passenger numbers are rising, and the share of rail in freight transport is set to increase to 25 percent of the modal split. But this rise in demand must contend with an infrastructure in need of renovation and with a growing shortage of skilled workers. That’s why German rail operator Deutsche Bahn (DB) is shaping up for the future by embracing digitalization and automating various processes in different business units and areas of work.

Machines and AI are helping DB offset the declining number of skilled workers – for example, by automating time-consuming routine checks, such as manual inspection of freight car tarpaulins. This lightens the load on employees, freeing them up to focus on more value-adding, varied, and interesting tasks. Data analyses help further enhance the quality of DB’s services and provide passengers and companies with even better information.

How AI Is Driving Customer Satisfaction

Data Power and AI Enable Highly Complex Real-Time Analyses

With some 4.65 million passengers daily, DB Regio is Germany’s market leader in regional and local public transport and plays a pivotal role in the German mobility landscape. The DB Regio network is an important backbone for green mobility in conurbations, large cities, and rural areas.

Every day, DB Regio’s operations generate millions of data points – from sources including train operations, maintenance work, and customer feedback. The company leverages this data to ensure that its business processes and customer solutions are sustainable and efficient. Information from all business units is provided to the cloud-based data, analytics, and visualization platform OneSource in accordance with defined standards. And the quality of the data from the source systems is subject to AI-based checks.

This ensures a transparent source of consistently high-quality data for analytics and AI applications throughout the company. To ensure the technical accuracy of this data, DB Regio has also implemented a data governance system, which encompasses all the people, processes, and technologies involved and ensures that data is used effectively for maximum long-term value.

At DB Regio, it’s particularly important that data is available rapidly and reliably, especially for highly complex real-time analyses, such as scheduling of trains with tight timing and complex structures. A data-driven application supports planning processes by accurately forecasting the impact of scheduling interventions and enabling AI-driven optimizations. This avoids waiting times and congestion on busy routes and has helped reduce total delay times by 58,000 minutes over the past year.

Innovative Analytics Tool for Optimized Passenger Information

Passengers on regional and local transport place great value on being kept up to date about train arrival times, delays, cancellations, and alternative travel arrangements. Timely and consistent passenger information is therefore key to achieving high levels of customer satisfaction.

Transparent processes and comprehensive data analyses are used to identify information gaps and quality issues in communicating information. The aim is to understand which sources of information (for example, the DB Navigator smartphone app) passengers use at which point in their journey.

A multidimensional analytics tool delivers insight into how customers rate criteria such as the timeliness, comprehensiveness, and reliability of forecast times. The tools also shed light on differences in perception between commuters and leisure travelers, and between passengers on suburban and regional transport. Continuous improvements in passenger information will make rail a more attractive travel option and increase customer satisfaction.

Targeted Customer Communication – Also in Social Media Channels

To ensure that data is used optimally and to support sound decision-making, the Regiolytics dashboard tool visualizes extensive analyses of social media channels, providing a digital picture of customer communication and interaction. Regiolytics has shown, for example, that AI-generated suggestions for day trips achieve significantly better results and considerably improve user engagement – reducing the bounce rate by 50 percent, increasing average visit time by 50 percent, and doubling user loyalty. In this way, Regiolytics supports data-driven decision-making and helps tailor content to customer needs more effectively, creating content that is more relevant for customers and optimizing the use of resources.

These examples illustrate how DB Regio is addressing the lack of skilled workers and making regional and local transport a more attractive, environmentally friendly, and sustainable option. DB Cargo is also successfully leveraging data and AI to play its part in Deutsche Bahn’s Strong Rail initiative.

Improving Efficiency and Transparency with AI

Leveraging AI to Drive the Digital Transformation of Rail Logistics

DB Cargo, Deutsche Bahn’s rail freight company, is one of Europe’s leading providers of rail freight transport, operating in 17 European countries. As such, it has a key role to play in the modal shift of road freight to rail. Eventually, the company aims to replace 30 million truck journeys every year with rail. With the help of AI, rail freight transport is becoming more efficient and competitive, which is crucial if more road freight is to be transferred to environmentally friendly rail.

Through the company-wide adoption of AI, DB Cargo is making rail transport more attractive for customers and improving transport routes, while also improving its profitability. Real-time analysis of large data volumes using machine learning methods and AI algorithms calls for considerable processing power, which is provided by scalable cloud infrastructures. This allows complex algorithms to be processed in parallel and distributed across several computers.

Freight Car and Order Tracking with AI-Assisted Forecasting

For its Wagon Intelligence project, DB Cargo has equipped its entire freight car fleet with GPS devices and sensors. The resulting position data is then linked with existing systems, enabling automatic mapping of departures and arrivals, border crossings, and the position of freight cars.

The data generated is also linked with context data, such as order and infrastructure data, to generate value-added business events (VABEs for short). This information can then be used to check whether the movement of freight cars is in line with the customer’s order; in other words, whether the freight car is in the right place at the right time.

The major benefits are rapid information provision and end-to-end transparency, enabling DB Cargo customers to track their freight cars throughout Europe in real time. In conjunction with DB Cargo’s recently rolled out customer service and sales application (casa), this approach makes customized, easily accessible logistics solutions a reality. The real-time data is used to provide customers with AI-driven forecasts about the arrival time of their freight cars, for example. The result: more attractive services, higher revenue, and greater profitability.

Smart Automation of Manual Maintenance Processes

Optimizing maintenance processes, a central area at DB Cargo, increases productive operation of the freight car fleet. This, in turn, improves transport quality for customers, accelerating the shift in intermodal freight transport from road to rail. The move from manual to automated visual inspections is helping tackle the shortage of skilled workers.

With the aid of AI, damage to the interior and exterior of freight cars is detected quickly and efficiently. In particular, this allows minor damage to be repaired swiftly, reducing the costs associated with damaged freight cars, and improving maintenance and servicing quality.

Smart predictive maintenance is one of the areas that DB Cargo is focusing on. The aim here is to return freight cars out of maintenance and into productive operation more quickly. To shorten the inspection process, a total of thirteen camera gates were positioned on the humps at eight classification yards. Up to 10,000 freight cars a day are scanned for damage, creating up to 300,000 high-resolution images. These extensive data sets were used to train an AI to detect specific damage. As a result, more than 70 percent of freight car damage can be detected at an earlier stage of the logistics process, and the findings of the analyses made available to maintenance plant employees. This makes for more precise maintenance and allows repairs to be coordinated more efficiently.

The AI now also detects cargo residues or defective roof tarpaulins automatically. This lightens the load on employees and reduces inspection times from several hours to just a few minutes.

DB Cargo has also joined forces with the steel industry to trial automated scrap detection. Detecting scrap at an early stage, before deliveries arrive, makes it easier for customers to schedule incoming freight cars for optimized production processes.

In addition, DB Cargo is using AI to analyze brake-disk thickness. Here, the technology helps identify maintenance requirements in good time and complete the work proactively, reducing the number of unscheduled freight car outages during operation. Through this and other initiatives, DB Cargo is highlighting the interplay of AI and digital innovations, making its rail offerings a more attractive option than road transport.

Conclusion

AI is a game changer for rail freight transportation, both in terms of improving customer satisfaction and maximizing process efficiency. The effectiveness of the technologies is already being felt throughout the Group. AI is now an integral part of the long-term corporate strategy and will be continually expanded. In addition to increasing capacity and optimizing train services, it also contributes to greater customer satisfaction and reduces delays. All of which strengthens rail as an attractive, environmentally friendly, and sustainable transportation system that has a key role to play in climate-neutral mobility and the successful decarbonization of transport in Germany.

The Authors:

Arlene Bühler is CIO and CDO of DB Cargo AG and, together with her team, drives the digitalization of Europe’s largest rail freight operator. Following various IT positions at Siemens and VW, she moved to Deutsche Bahn as Head of IT Operational Excellence in early 2020, assuming the newly created corporate function Group IT Portfolio & Performance Management.

 

 

Tanja Schlesinger is Head of OneSource and Data Officer at DB Regio AG. She and her team created the AI-assisted data platform OneSource and develop sustainable, customer-centric analytics applications to reduce environmental impact, devise strategies to combat the shortage of skilled workers, expand rail capacity, and increase the attractiveness of public transport.

 

 

Kristina Sahling is a Data Analytics Manager at Accenture, specializing in the implementation of data analytics products with cloud technologies. Her work focuses primarily on supporting data-driven decision-making processes and decision-making behavior. She is also completing a PhD at Humboldt University in Berlin.