Compositional Formal Analysis Based on Conventional Engineering Models
This addresses the problem of applying formal methods more broadly in critical software systems, though it appears incremental as it builds on existing engineering models.
The paper tackles the difficulty of generating formal specifications for real systems by introducing SLICED, a methodology that uses conventional engineering models to derive compositional formal state models and assertions, demonstrated with NASA's Virtual ADAPT model and validated via Simulink replication.
Applications of formal methods for state space exploration have been successfully applied to evaluate robust critical software systems. Formal methods enable discovery of error conditions that conventional testing may miss, and can aid in planning complex system operations. However, broad application of formal methods has been hampered by the effort required to generate formal specifications for real systems. In this paper we present State Linked Interface Compliance Engine for Data (SLICED), a methodology that addresses the complexity of formal state machine specification generation by leveraging conventional engineering models to derive compositional formal state models and to generate formal assertions on the state machines. We demonstrate SLICED using the Virtual ADAPT model published by NASA and validate our results by replicating them using Simulink.