Time is running, but still worth to mention that we recently had a quite some successful PhD defenses. Congratulations to , , and !
In his thesis entitled “Cost-Effective Regression Testing for Self-Driving Cars through Simulation-Based Strategies”, advances software testing for autonomous driving systems by introducing scalable regression testing techniques and open-source tools that improve effectiveness while reducing testing costs in simulation-based environments. We are glad that Christian stays with our research group and now, amongst other research activities, is pushing his results towards an innvovation project.
In his thesis entitled “Towards Validating and Debugging Scientific Data Analysis Workflows: Irregularity Localization and Contract-Driven Design”, introduces automated debugging and validation techniques for scientific data analysis workflows. An interesting part of the Q&A was the discussion on generalizability of the results, which may be explited for increasing trustworthness for both AI-based systems and AI-based software evelopment. Reflecting even further, the techniques developed in his thesis are founded in the very ideas of Andreas Zeller’s delta debugging and Bertrand Meyer’s design-by-contract. The lesson learned: Many of the ideas we need for managing the AI revolution in software engineering are already there, we just have to pick the right ones.
In his thesis entitled “Bridging the Data Desert: Mitigating Challenges of Model Accessibility in Simulink Research”, tackles the shortage of shareable Simulink models by demonstrating the value of open-source datasets and introducing scalable model synthesis and anonymization tools that enable more transparent, reproducible, and industry-relevant research in Simulink-based MDE. We are glad that Alex stays with the group and now pushes forward our vision on software merging and integration: turning merge conflicts from a manual chore into a solved problem, bridging everyday version control with the theory of software variability.