The research demonstrates that simulation components can be successfully converted to FMI-compliant FMUs through a two-layer wrapper architecture, though with important technical considerations. Key findings include: (1) FMI 2.0's limited support for complex data structures necessitates flattening of Ada records, introducing processing overhead; (2) type system incompatibilities require workaround solutions; and (3) execution frequency coordination was handled internally within the components due to configuration complexity in the simulation engine.
Despite these challenges, workshops with Saab engineers indicate that FMI and SSP could satisfy a large extent of their simulation needs, particularly for non-real-time applications with moderate complexity. The standards' primary value lies in facilitating passing of models between departments, potentially reducing integration time with automated processes. Testing with the open-source OMSimulator revealed additional limitations, including difficulties with algebraic loops and sequential execution patterns, suggesting that tool maturity remains a consideration.
The findings suggest that while FMI 2.0 and SSP offer valuable standardisation benefits for aerospace simulations, their adoption requires careful consideration of performance trade-offs and implementation complexity. Future developments in FMI 3.0, particularly improved support for complex data structures and execution coordination, may address many current limitations. This research provides practical insights for aerospace organisations considering the transition from proprietary to standardised simulation interfaces, highlighting both opportunities and challenges in adopting industry standards for critical applications.