Safety in cooperative, automated driving by tackling uncertainties (BEYOND VALIDATION AI)

• Subject Area: Systems Engineering ; Electrical Engineering and Information Technology ; Computer Science
• Term: since 2026
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 549102058

The consideration and reduction of uncertainties in the development and testing phase is a key aspect in dealing with the technological challenges of safe automated driving. With the increasing use of Artificial Intelligence (AI) methods, data and model-related uncertainties arise that must be addressed. The inclusion of these uncertainties in Cooperative Automated Driving (CAD) requires a multi-disciplinary approach. This is made possible by Technische Hochschule Ingolstadt (THI) in terms of staff and infrastructure through (1) the nationwide scientific flag-ship center (“Forschungsbau”) for vehicle safety “CARISSMA,” (2) the Bavarian AI Mobility Hub “AImotion Bavaria,” which was established as part of the “Hightech Agenda Bayern,” (3) suitable research infrastructure such as THI laboratories, (4) regional test beds for CAD, (5) research professorships with specialist expertise for this multi-disciplinary approach, (6) a mobility-oriented industrial research environment, and (7) a strong focus in teaching/study programs on the area of innovative mobility. The overall goal of the Research Impulse (RI) is a permanent interdisciplinary competence center for improving safety in AI-based connected mobility systems through a deep understanding of sources of uncertainty, their propagation, and impact on decision-making. The fundamental scientific consideration of uncertainties in CAD is ensured by the research professors submitting the application. The two doctoral centers for “AI/Computer Science” and “Engineering Sciences” in the THI Graduate School support the recruitment of research-intensive staff. With the RI an institute-wide research group is to be established and coordinated in the medium term by a junior research group leader. In addition to understanding the sources of uncertainty, methods for quantifying and reducing uncertainty will be researched. The focus is on perception and prediction uncertainties of the vehicle environment. It will be investigated how knowledge gaps leading to epistemic uncertainties can be reduced, for example, by large AI models, the identification of so-called “corner cases,” cognitive/physical behavior models or the inclusion of domain knowledge. The remaining uncertainties in the representation and prediction of the vehicle environment, but also in V2X communication, are described using appropriate quantities that are part of the research work. These quantities form the basis for the use of AI methods in safety-relevant components of automated vehicles and are used in the RI for planning safe trajectories, for test methods, and for the design of resilient functional architectures. The existing equipment at THI and the technological ecosystem of the regional environment are intended to help further build up the research-intensive staff at THI and make THI a research beacon with the goal formulated above during the funding period.

• DFG Programme: Research Impulses
• Applicant Institution: Technische Hochschule Ingolstadt
• Spokesperson: Professor Dr.-Ing. Michael Botsch
• Participating Researchers: Professor Dr. Christian Facchi; Professor Dr. Andreas Festag; Professor Dr.-Ing. Werner Huber; Professor Dr. Andreas Riener; Professorin Dr. Stefanie Schmidtner; Professor Dr. Torsten Schön