Building a Scenario Library for Autonomous Driving Evaluation
- Institute
- Professur für autonome Fahrzeugsysteme (TUM-ED)
- Type
- Bachelor's Thesis Semester Thesis Master's Thesis
- Content
- Description
Motivation
The validation and safety assurance of modern autonomous vehicle systems is a cornerstone in the development of highly automated driving functions (Level 4+). In this context, there is a fundamental trade-off between simulation and real-world testing: while physical tests offer high physical validity, they are capital-intensive, time-consuming, and limited in scalability. In contrast, simulation platforms such as CARLA enable reproducible testing of complex traffic scenarios at a comparatively low cost and with high scalability.
As a result, the evaluation of autonomous driving functions takes place across a variety of different tools and environments. For instance, planning algorithms are validated in formal 2D frameworks like CommonRoad, while overall software behavior is tested in realistic 3D simulators such as CARLA.
This often leads to a comparability issue: A software stack might function reliably in one environment but exhibit unexpected behavior in another—often due to discrepancies in sensor modeling or physical representations. To enable reproducible, comparable, and scalable testing, a systematic and structured scenario library is required to bridge these different evaluation environments.
Objectives
The goal of this thesis is the conceptual and/or technical development of a modular scenario library for the evaluation of autonomous driving functions in CommonRoad and CARLA.
Planned work packages include:
- Definition and structuring of relevant test scenarios for urban and/or highway applications (e.g., lane keeping, lane changes, merging, intersection scenarios).
- Identification of "Minimum Viable Scenarios" (MVS) for fundamental functional safety assurance.
- Analysis of differences between Scenario Definition frameworks used in both tools.
- Implementation of these scenarios in CommonRoad and CARLA.
- Cross-Platform Behavioral Analysis: Identifying and analyzing divergent AV behaviors in identical scenarios to determine the impact of simulation fidelity on software stack performance.
Requirements
- Interest in the simulation or evaluation of autonomous systems.
- Strong programming skills.
- A structured and independent working style.
- Proficiency in German or English.
- Prior experience with CARLA, CommonRoad, or autonomous software stacks is helpful but not mandatory.
Start Date: Immediate / As soon as possible.
How to Apply: If interested, please submit your application via email, including your CV, current transcript of records, and a brief cover letter explaining why you would like to work on this subject.
- Tags
- AVS Bank
- Possible start
- sofort
- Contact
-
Christoph Bank
christoph.banktum.de