Scope of physics-based simulation artefacts

Documentation is critical for trustworthy AI in simulations: The paper identifies specific metadata and documentation requirements needed to make physics-based simulation models explainable to users, emphasizing that transparency about how simulations work is essential for responsible AI deployment.

Three key factors define simulation scope: Understanding a simulation's trustworthiness requires clarity on three elements: the subject matter being simulated, the specific objective or purpose of the simulation, and the physical system being modeled—each of which can significantly affect how reliable the results are.

Current standards need harmonization: The paper compares two existing documentation frameworks (MODA and ModGra) and finds gaps, suggesting that organizations need clearer, unified minimum requirements for documenting simulation models to ensure consistency in responsible AI practices across industries.

Visual tools can help stakeholders understand simulation limitations: The proposed "object-objective abstractness diagram" offers a practical way to map different simulation use cases and their complexity levels, helping non-experts understand when and how simulations can be reliably applied to real-world problems.

Explainability requires systematic artifact documentation: The research demonstrates that making AI systems based on physics simulations truly explainable demands comprehensive documentation of all modeling and simulation workflows, not just final results—a foundational step for accountability in AI systems.