Master Thesis in Software Engineering: Adaptive Architecture for Human-AI Annotation Systems

This thesis project involves designing, implementing, and evaluating a confidence-aware adaptive verification architecture for an existing camera-based annotation tool used in validating autonomous hauling systems. The work focuses on structuring, integrating, and testing adaptive human-AI workflows to improve annotation correctness, performance, and maintainability while mitigating automation bias. Students will collaborate with the Verification & Validation team to develop modular components, implement decision policies, and conduct empirical evaluations in an industrial setting.