AI Agents for Test Generation of SystemC Models

AI Agents for Test Generation of SystemC Models
Hochschule München is among the largest German universities of applied sciences, located at the heart of Munich, Germany. The working group "AEMY" focuses on safe, secure and smart systems. Our main working areas include RISC-V processor design, WebAssembly runtime development and open source chip design tools.
SystemC is the standard C++ library for modeling hardware and embedded systems at the system level, for example for virtual prototypes and architecture exploration. Although SystemC models are software, they are rarely tested as systematically as production code.
Our tool AMSGen uses AI agents to automatically generate GoogleTest unit tests for C++ code, guided by code coverage measured with gcov. In this thesis, you will extend AMSGen to SystemC. The agents will generate tests for SystemC modules, processes, and TLM interfaces, taking simulation time, events, and concurrency into account. The goal is to bring the systematic, coverage-driven testing of production code to SystemC models. Potential areas of focus include:
- Analyzing which SystemC constructs (SC_METHOD, SC_THREAD, signals, TLM sockets) need special test support
- Extending the AMSGen pipeline to generate SystemC test harnesses with GoogleTest
- Using gcov coverage of the model code as feedback for the agents
- Evaluating coverage and fault detection on example models with injected faults
- Comparing local open-weight LLMs in terms of quality, cost and runtime
This project offers hands-on experience in system-level modeling, software testing and agentic AI. For further information or to express your interest, feel free to reach out.