Project Time:: Projektzeitraum: 01.01.2022-31.12.2024
Funded By: Gefördert von: Bundesministerium für Wirtschaft und Klimaschutz
Machine Learning (ML) is currently viewed as the most promising approach for autonomous driving. In safety-critical contexts, like autonomous rail vehicles, ML models have to be provenly robust in order to, for example, detect vehicles on roads reliably. The project SafeTrAIn ( Website) aims to set the foundation for the safe deployment of ML models in autonomous rail vehicles, thereby taking steps towards utilizing ML models in autonomous vehicles in general.
Goals
- Development of solutions that allow to prove the safety of machine learning-based autonomous rail vehicles
- Development of architectures and methods for the safety concept
- Derivation of relevant criterions for the safety proof
- Development of concepts for the fast and efficient constuction of GoA4-Rail vehicles.
Partners:
- Siemens AG
- Siemens Mobility
- Fraunhofer
- Merantix Labs
- Edge Case Research
- Bit technology Solutions
- Setlabs Research
- Bridgefield
- University Applied Science Düsseldorf
- TÜV Rheinland InterTraffic
- TÜV Süd Rail
- TÜV Nord
- German Institute for Standardization (DIN)
- VDE
- BSI
- ITQ
If you are interested in this area of research and want to gather some first hand experience (in the form of a thesis, a team project, a scientific project or a hiwi) please contact Konstantin Kirchheim.
SafeTrAIn