Fraunhofer-Allianz VerkehrDirector: Prof. Dr.-Ing. habil. Mario Trapp
Contact: Dr. Reinhard Stolle
Fraunhofer Institute for Cognitive Systems IKS
- Click here for IKS's internet presentation (iks.fraunhofer.de)
Transformation of Mobility through Safe AI Systems:
The future of mobility, healthcare, and industry will be fundamentally shaped by connected cognitive systems. At Fraunhofer IKS, we combine research in Artificial Intelligence with software engineering. Our goal is to unlock the full potential of emerging technologies such as quantum computing and AI—without compromising on safety or reliability.
Continuous Safety Assurance: Safe Operation for AI-Based Driving Functions
Our “Continuous Safety Assurance” approach enables efficient and safe deployment of AI-powered driving functions into operational use. We support you from defining safety goals, acceptable residual risk, and precise specification of the Operational Design Domain (ODD) to create the architecture, defining suitable metrics, and optimizing verification processes—with the support of safe LLMs and agents.
Our structured, evidence-based safety cases, which integrate confidence assessments and standards-compliant argumentation (including ISO PAS 8800, SOTIF, and the EU AI Act), form the basis for successful certification. For ongoing operations, we support integration of monitoring solutions, definition of Safety Performance Indicators (SPIs), as well as governance structures, triggers for safe updates, and reintegration of runtime data directly into the safety argumentation. Through rapid prototyping, we also accelerate integration and validation steps within standardized environments.
This creates a closed loop of operations, development, and continuous learning: compliance remains demonstrable, risks are consistently controllable, and your application achieves measurable safety quality through iterative improvement.
Our portfolio covers all areas—from embedded AI at the microcontroller level (e.g., as a virtual sensor in engine control) to highly complex AI functions for perception, trajectory planning, or end-to-end AI architectures within the vehicle and infrastructure.
Our Services at a Glance:
| Engineering of Safety-Critical Systems | Definition of safety goals, ODD, risk management, and design of safe ML architectures and efficient V&V processes with LLM/agent support |
| Continuous Assurance & Monitoring | Monitoring of assumptions and SPIs, root cause analysis for incidents, and safe execution of OTA updates |
| Continuous Safety Argumentation | Creation of structured, dynamic safety cases including confidence assessment, re-argumentation, and compliance documentation |
| Rapid Prototyping | Fast development and validation of your AI functionalities using our APIKS framework |