Graph-Driven Intelligence

Methodology and Technical Innovation

The project develops a methodology that combines ontologies, knowledge graphs, and Large Language Models (LLMs) to unify code, models, documentation, and runtime data into a shared semantic space. This enables automated reasoning, interactive visualizations, and traceability from high-level scenarios to low-level implementation. The approach supports capability-driven design and functional decoupling, reducing reliance on manual analysis and scarce expert knowledge. Prototypes include an ontology builder, visualization tools, and interaction inference techniques, all integrated into the Renaissance platform.

Industrial Validation and Broader Impact

The methodology is validated using real-world use cases from Philips, ensuring practical relevance and scalability. The project aligns with the HTSM Systems Engineering Roadmap and contributes to national transitions, particularly the digital transition in high-tech industries. It addresses key challenges like human capital shortages, system lifecycle complexity, and the need for cost-effective engineering. By enabling AI-assisted system analysis, it strengthens the Dutch high-tech sector’s innovation capacity and competitiveness.