Automatisierter Zug
istock/ zhangxiaomin

Automated driving on rails

News from Sep. 20, 2021

A consortium of 12 partners will develop and test technologies for automated driving on the railroads in the project “Berliner Digitaler Bahnbetrieb” (BerDiBa) over the next 4 years. Fraunhofer FOKUS will use its expertise in environment perception for the automated vehicle and adapt its artificial intelligence tools for traffic on rails.

The future lies in automated rail transport. The BerDiBa project is developing three aspects of digital rail operations: driverless trains, automatically scheduled maintenance, and automated remote control for situations in which autonomous travel is not possible, for example, due to the weather. The goal is to make rail transport safer and processes more efficient.

The Fraunhofer FOKUS research team from the Smart Mobility business unit has many years of experience in teaching automated vehicles to reliably and robustly perceive the environment around them. Fraunhofer FOKUS is contributing this expertise in close cooperation with the Daimler Center for Automotive Information Technology Innovations (DCAITI), which is based at TU Berlin. For example, the team will adapt its labeling tool FLLT.AI for generating lidar image training data for use in the rail environment. This will allow train artificial intelligence (AI) to use high-quality training data to learn how to correctly sense the environment. Using data from sensors, such as radar and lidar, mounted in and on the train, the AI can then form an accurate picture of the current situation after training and recognize and distinguish between static objects, such as signaling systems, tracks, and stations, as well as dynamic objects, such as people, animals, vehicles and other trains.

Dr. Ilja Radusch, head of the Smart Mobility business unit at Fraunhofer FOKUS, summarizes the similarities and differences of automated driving on road and rail as follows: “It sounds like a similar task at first glance but the problems lie in the details. For example, continuous route monitoring is something we are not familiar with in conventional road traffic. On the one hand, we bring in our knowledge and experience from road traffic, but we also learn a lot of new things.”

BerDiBa is funded by the city of Berlin with 7.6 million euros from the ProFIT innovation funding program and has a total volume of 13.7 million euros.

Collaborative industry partners:

  • acs plus GmbH
  • AAI GmbH
  • GSP GmbH
  • ITQ GmbH
  • neurocat GmbH
  • Siemens Mobility GmbH (Konsortialführer)
  • Teraki GmbH
Partners from research and science:
  • DFKI
  • Fraunhofer-Institute FOKUS und HHI
  • TU Berlin
  • Zuse Institute