Tuesday, October 13, 2026

From Scenarios to Benchmarks: Comparative Evaluation of Automated Driving Functions

Topic and Goal of the Thesis

Automated driving functions are tested using selected test scenarios within their scope of application, the Operational Design Domain (ODD). To assess their performance, they can be compared with another driving function or an earlier software version. This comparison is meaningful only if the test scenarios occur in both ODDs and the selected metrics are suitable for evaluation. This is particularly relevant for AI-based or end-to-end driving functions, whose behavior is difficult to predict.

The goal of this work is to develop a methodology that enables the comparison of the performance of automated driving functions through the selection of suitable test scenarios and metrics. It will be implemented using two driving functions or software versions as examples.

Working Points

• Literature research on the comparison of driving functions, test case selection, and metrics
• Development of a method for comparing driving functions
• Evaluation and comparison of two driving functions based on selected test cases

Requirements

• Good written and spoken German and/or English skills
• Reliability and commitment
• Ability to work independently and in a structured manner
• Enthusiasm for the topic
• Basic knowledge of Python and ROS 2

Note: Please attach brief resume and grade summary.