Privacy Preferences

When you visit any website, it may store or retrieve information through your browser, usually in the form of cookies. Since we respect your right to privacy, you can choose not to permit data collection from certain types of services. However, not allowing these services may impact your experience.

  • For the demand-oriented design of our website we create pseudonymous user profiles with the help of Google Analytics. In this way we are able to recognise returning visitors and count them as such.
  • These technologies enable us to integrate external services such as video sharing into the website.


TERAKI enters €14M Partnership to deliver “ADAS for Trains” in Berlin – The BerDiBa Project

BERLIN, January 14, 2022 - BerDiBa is a consortium collaboration between Berlin-based research institutions and companies, including Teraki. The project seeks to narrow the gaps in railway mobility involving high-speed, autonomous train operations. The results will serve as the cornerstone for highly automated rail systems and enhance trains’ safety and efficiency, carrying more passengers in each timeframe. For this purpose, Teraki is creating algorithms for environment and obstacle detection through leading AI-enabled sensor data processing.

BerDiBa: Advancing autonomous railway mobility

BerDiBa stands for Berliner Digitaler Bahnbetrieb (Berlin Digital Railway). The project is funded by the state of Berlin, with a total budget of €13.7 million. Over four years, a consortium of 12 industry and research partners will develop and test the foundational technologies for automated train operations.

The project covers various research fields, encompassing autonomous driving, optimized operations, and automated planned maintenance. Teraki is advancing intelligent sensor data processing to improve obstacles detection and classification for high-speed trains, which is crucial to enabling partial and full driverless operations.

Teraki’s sensor processing to improve Berlin’s railway network

Nowadays, trains operate at their maximum capacity, posing challenges to the number of passengers served and operating times. Similarly, the city and its operators devote substantial resources for tracks and wagons maintenance. Sometimes, harsh weather conditions and unforeseen obstacles, like wild animals, cause failures and interruptions if not detected on time.

Teraki uses sensor data collected on trains to train AI models. Specifically, Teraki leverages camera, 3D point cloud, and telematics data to achieve greater accuracy on static and dynamic obstacle detection and perform automated infrastructure scanning.

Building upon its years of experience in the automotive industry, Teraki is developing algorithms specialized in detecting regions of interest and times of interest (“ROI” and “TOI” respectively). This approach selects the relevant information without losing crucial data. Simultaneously, it reduces noise and improves signal quality.

Teraki assembled a state-of-the-art signal detection sensor box to scan real-life environment data on the train. The sensor box contains four RGB cameras, a focusable, long distance lidar and an event camera (DVS) to cover different scenarios (lighting conditions, weather conditions, speed) as comprehensively as possible.

Recordings from the event camera (left) compared to RGB camera (right). Data collected from sensors under diverse weather conditions will improve detection algorithms.

Recordings from the event camera (left) compared to RGB camera (right). Data collected from sensors under diverse weather conditions will improve detection algorithms.

The sensor scanner setup provides valuable insights into the train’s environment and leads to greater operational efficiency and safety:

Enhanced scheduling and reliability: Monitoring and preserving an optimal distance between trains during run-time would lead to reduced gap times in between routes, completing more journeys with shortened waiting times, therefore, increasing passengers transported.

Intelligent maintenance: Continuous and automated monitoring of the track and wagon conditions could rapidly identify troubles, minimize the interruption periods, and predict the best possible maintenance times. This would reduce the overall costs of maintenance and improve the upkeeping tasks.

Improved safety: The safety of the train cars increases by detecting obstacles from up to hundreds of meters away, prompting well-timed reactions to avoid costly collisions and break-down times.

The BerDiBa project will create more jobs in AI and autonomous driving-related fields in the future. The project also strengthens collaboration between research institutes, universities, and companies in Berlin.

About Teraki

Teraki is a Berlin-based company focused on delivering safer vehicle autonomy at lower costs. In a lightweight manner and on the edge, Teraki smartly and 10x more efficiently selects and processes in real-time the relevant information from large amounts of sensor data, leading to 20% safer and more robust autonomous applications. Use cases include autonomous driving vehicles, trains, and delivery robots. The company currently has 50+ employees, with offices in Berlin and Tokyo.

About BerDiBa

The BerDiBa Consortium is integrated by 7 industry partners leaded by Siemens Mobility and 4 science and research partners. Industry partners include ACS Plus, AAI GmbH, GSP GmbH, ITQ GmbH and Neurocat GmbH. Science and research partners are DFKI, Fraunhofer FOKUS und HHI, TU Berlin und DCAITI and Zuse-Institut Berlin.

Share this Post: