What do you plan to achieve in Challenge 1?
In Challenge 1, our research team is working towards a workflow that allows for rapid iteration on design requirements through computational design methods and automated simulation-based verification and validation activities.
My contribution focuses specifically on the application of Large Language Models (LLMs) to generate simulation-based tests that achieve coverage over the requirements. This involves processing the natural language embedded in requirements and also ensuring that the test configurations generated can be executed by the simulator (generating the tests right), and that the executed simulations ultimately also correctly test the requirement statements (generating the right tests). This will ultimately support faster design iteration while ensuring early discovery of unsatisfied requirements, speeding up requirements based design workflows.