Working student (m/f/d) - Generative AI-supported Testing Methods for Automated Driving Development

Complexity and scale of automotive systems are increasing, particularly in the area of automated driving and advanced driving assistance systems. The increased complexity of such safety-critical systems demands for advanced testing and verification methods to ensure safety under an abundance of operating conditions and traffic scenarios that may be encountered.

Recently, generative AI has shown promising results for content generation, such as for text (ChatGPT), image (stable diffusion), or video generation (OpenAI SORA). In our R&D work, we would like to explore the possibilities of using generative AI to create critical traffic scenarios learned from traffic data.

We are seeking for a talented and driven Working Student / Master thesis student (m/f/d) to join the Systems Engineering R&D team and support the development of testing methods by using recent results from generative AI for scenario and traffic video generation. In your role, you gain hands-on experience on the training and validation of generative AI models, and contribute to our R&D testing platform for automated driving. The work is part of the public funded project nxtaim.de.

Responsibilities:

• Adaptation of a diffusion model for traffic scenario/video generation 
• Develop criticality metrics for traffic scenarios
• Develop optimization-based techniques to guide the generation towards critical scenarios
• Integration of the methods into our existing testing workflow
• Evaluate the testing campaign to extract critical test case