SAFE AI ENGINEERING
AI engineering ensures safety of AI components for autonomous, connected driving
SAFE AI ENGINEERING
AI engineering ensures safety of AI components for autonomous, connected driving
Start: 03/2025
End: 02/2028
The Safe AI Engineering project focuses on ensuring the safety of AI-controlled components for automated and connected driving.
The project aims to develop a continuous, lifecycle-level safety methodology to ensure safety in dynamic environments. By integrating its results into established standards such as ISO 26262 and SOTIF, the project bridges the gap between AI verification and established safety assurance for the automotive industry.
The project focuses on camera-based detection of vulnerable road users (VRUs), such as pedestrians, using deep neural networks.
Three increasingly complex scenarios are defined – from the detection of a static pedestrian to the handling of multiple dynamic VRUs in urban environments.
Role of the FZI
The FZI’s main focus in the project is on approaches to explainability and monitoring for AI-based perception functions.
In particular, the FZI will investigate curriculum learning – targeted learning with increasing complexity – as well as redundancies at the layer level in neural networks using Mixture of Experts models.
Additionally, the FZI plans to evaluate and demonstrate the concepts and methods using CoCarNextGen.
Contact
Philip Schörner
Department Manager
Division: Intelligent Systems and Production Engineering
- +49 721 9654-358
- schoerner@fzi.de
- Headquarters Karlsruhe
Research Focus
Applied Artificial Intelligence
Das Hauptaugenmerk des FZI in diesem Forschungsschwerpunkt liegt auf der praxisnahen Erforschung der Schlüsseltechnologie Künstliche Intelligenz (KI). Innovative KI-Lösungen werden entwickelt und für unserer Partner und Kunden in Anwendungsbereiche wie Mobilität, Robotik, Gesundheitstechnik, Logistik, Produktion sowie Ver- und Entsorgung transferiert.
Safety, Security and Law
Um die sichere Digitalisierung zu ermöglichen, erforscht und vermittelt das FZI in diesem Forschungsschwerpunkt anwendungsnah innovative Konzepte, Methoden zur Absicherung von IT-Systemen sowie rechtliche Rahmenbedingungen.
The SAFE AI Engineering joint project is funded by the Federal Ministry for Economic Affairs and Energy.
Project partner:
Over 20 project partners from science and industry
More projects
SouveMed
Data trustee model protects use of individual health data in research
RISE
AI assistant serves as first point of contact for legal questions
MAD Urban
AI engineering empowering security reasoning throughout the entire lifecycle of an AI function
SDI4ECom
Secure and self-determined digital identities in e-commerce
CaMinoBW
Co-design for application-specific microelectronics with innovative chiplets in Baden-Württemberg
regioKArgoTramTrain
Jointly moving people and goods
PathoBot
Stopping pathogen transmission in hospitals quickly, reliably, and comprehensively using AI.
ISOLDE
Generating resilient system-on-a-chip architectures for RISC-V
VINNIE
Body-worn sensors and wearables for recording vital data and identifying health-related issues in mental and psychosomatic disorders