Implementation of airborne ML models with semantics preservation
This work addresses safety-critical compliance issues for ML in aviation, but it appears incremental as it builds on existing regulatory frameworks like EASA's concept paper and ED-324.
The paper tackles the challenge of ensuring safety and compliance for machine learning models in airborne systems by clarifying the distinction between an ML model and its unambiguous description (MLMD) and refining semantics preservation for accurate replication, applying these concepts to industrial use cases to build and compare target models.
Machine Learning (ML) may offer new capabilities in airborne systems. However, as any piece of airborne systems, ML-based systems will be required to guarantee their safe operation. Thus, their development will have to be demonstrated to be compliant with the adequate guidance. So far, the European Union Aviation Safety Agency (EASA) has published a concept paper and an EUROCAE/SAE group is preparing ED-324. Both approaches delineate high-level objectives to confirm the ML model achieves its intended function and maintains training performance in the target environment. The paper aims to clarify the difference between an ML model and its corresponding unambiguous description, referred to as the Machine Learning Model Description (MLMD). It then refines the essential notion of semantics preservation to ensure the accurate replication of the model. We apply our contributions to several industrial use cases to build and compare several target models.