Technical reports · 2026.06.18
Automotive Foundation Models: Fundamentals, Training Concepts and Applications
Automotive Foundation Models: Fundamentals, Training Concepts and Applications
Marcel Aach, Laurenz Adolph, Stefan Baumann, Sabrina Benassou, David Bernhard, Sebastian Bernhard, Mohamed-Khalil Bouzidi, Markus Braun, Thomas Brox, Sven Burdorf, Daniel Dauner, Thies de Graaff, Samed Doğan, Bojan Derajic, Fabian Flohr, Silvio Galesso, Malte Ganzer, Hanno Gottschalk, Dominik Heid, Falk Heuer, Tao Hu, Christian Hubschneider, Marius Kästingschäfer, Mert Keser, Peerayut Khongsab, Kamil Kowol, Felix Krause, Edward Kromm, Giovanni Lucente, Artem Lukin, Annajoyce Mariani, Sudhanshu Mittal, Arian Mousakhan, Firas Mualla, Adam Molin, Jonas Neuhöfer, Joshua Niemeijer, Christian Ojeda Bernal, Björn Ommer, Christina Ourania Tze, Sven Peyinghaus, Pascal Piecha, Ulrich Prestel, Andrea Ramazzina, Jörg Reichardt, Christoph Rist, Werner Ritter, Dennis Rochau, Fabrizio Savarino, Philipp Schenkel, Christian Schlauch, Julian Schmidt, Ömer. Şahin. Taş, Laurenz Thiel, Abhishek Vivekanandan, Stefanie Walz, Jiangtao Wang, Zhaoze Wang, Julian Wiederer, Katharina Winter, Carolin Wunderlich, Harsh Yadav, Yue Yao, Yutong Yang · nxtAIM consortium · English
Overview
White paper surveying foundation-model architectures, training and interpretability, followed by camera, LiDAR and radar data generation, video models and abstract scene representations. Discusses synthetic-data metrics, domain transfer, planning and validation strategies for automated-driving applications.
Publication and version information
The linked PDF cover is dated 18 June 2026; the project index lists the work under 2025. These dates refer to different source/version evidence.
Version: Current PDF filename labels v1.2; cover dated18June2026