Digitalization for a Sustainable Rail System (DeepTrain)
This BMWK funded research project tackles pressing challenges in the rail sector, as outlined in the Railmap 2030 – Policy Milestones for a Transportation Transition. A key aspect of this roadmap is the role of digitalization in making rail transportation more efficient, sustainable, and intelligent. In particular, the project aims to develop systems that can support automated driving assistance and reduce energy consumption in rail operations.
One major obstacle in both regional and freight transportation is the lack of current, detailed digital data about the rail infrastructure. Information such as track layout, gradients, and the positions of signals or stopping points is often missing or outdated. This makes it difficult to implement advanced systems that rely on accurate and up-to-date infrastructure knowledge.
In addition, temporary or dynamic changes—like speed limits or schedule changes—are not yet available in digital formats that can be processed in real time. Accurate train positioning is also still a challenge, especially in tunnels or urban areas where satellite signals are unreliable.
Our project aims to overcome these limitations by developing real-time digital tools that can automatically capture and interpret railway infrastructure and support modern driving assistance systems.
Our team contributes to two main **research focuses** within the project:
Research Focus 1: Self-Localization Onboard the Train
We are working on solutions for determining a train’s position using only sensors mounted on the vehicle itself, without relying on external infrastructure. This includes:
- Selecting and integrating suitable sensors
- Developing robust algorithms for location tracking
- Handling typical rail-specific challenges like changing lighting or high speeds
Our approach combines information from different sensors to ensure reliable positioning, even in difficult environments. The system is designed to be flexible and capable of working across a wide range of conditions and rail scenarios.
Research Focus 2: Detecting Infrastructure Changes
The second research focus involves monitoring and detecting changes in the railway environment over time. We aim to automatically identify when something in the infrastructure or surroundings has changed. This includes:
- Building a digital reference map of the track environment
- Monitoring differences between repeated observations
- Using this information to either improve positioning or signal a relevant change
This change detection capability plays a vital role in maintaining up-to-date digital representations of the railway and enabling safe and autonomous operation.
Vision
Our developments will be tested and validated using a real-world demonstrator, operating under normal service conditions. This will show the practical value and feasibility of our technologies.
In the long term, we aim to support the autonomous operation of trains in specific rail sections and contribute to a smarter, greener, and more efficient rail system.
