SECREDAS focused on protecting against malicious intrusions into autonomous driving that would put road users at risk.
The customer
The challenge
Project duration: 01.05.2018 - 31.10.2021 | Website: secredas-project.eu
The SECREDAS project, a collaboration of more than 70 organizations, completed its technical work in October 2021. During its three-year duration, the project involved security and privacy experts in developing integrated security and privacy solutions for automated and autonomous systems.
The aim of the research project was to protect future autonomous driving from external malicious interference or hacking attacks that could put vehicle occupants or other road users at risk. To achieve this, SECREDAS created a common reference framework for security and privacy that allowed technology solutions to be designed, developed, and tested that addressed both of these areas simultaneously. The results were successfully tested in a series of realistic driving scenarios on the road and in the presence of threats from hackers and vulnerabilities. The project included both new security features in rail applications and health monitoring applications.
Solution
Results & effects
Merantix Momentum developed algorithms that enable robust image segmentation and provided evidence that model ensembling is promising. Merantix Momentum also designed and implemented a test framework for model evaluation. To demonstrate the system in a vehicle, we developed an algorithm to detect attacks on the live image stream from on-board cameras.
.png)
.png)
.png)
