ANYMOS

Brief description of the project

Research is being conducted into which measures ensure the protection of personal data while at the same time enabling efficient processing. The results support companies in the mobility sector in identifying anonymization requirements for specific use cases, selecting suitable methods based on the current state of research and technology, applying them correctly, and systematically identifying and evaluating re-identification risks.

Project objectives 

The goal of Fraunhofer IOSB is to research and apply methods for anonymizing camera-based data collected in vehicle interiors. Anonymization is to be established and used as a sustainable key technology for future mobility systems in order to be able to offer and implement innovative assistance functions that will require camera-based passenger and interior detection in the future in compliance with the GDPR. To this end, IOSB is developing and implementing a process model based on a specific use case, which is to be expanded into a standard procedure in the future. 

Project outcome 

Based on a previous analysis that took security requirements, data protection, and quality criteria into account, Fraunhofer IOSB developed a methodology toolkit that was implemented and validated on a trial basis. This toolkit enables the use of anonymized data when training ML models. Data protection in accordance with the GDPR is maintained, while enabling targeted use for internal and external AI applications. A concrete example of its application is research into the detection of sleeping test subjects in automated vehicles. Here, camera images and EEG recordings are used to prepare for the safe takeover of the driver by the automated system. Fraunhofer IOSB is integrating the project results into the driving simulator as well as the Advanced Occupant Monitoring System (AOMS).

Fraunhofer IOSB complements the ANYMOS consortium with its research groups Perceptual User Interfaces, Explainable AI and Distributed Systems, among others.