Publikation
On the Origins of Self-Explainability in Cyber-Physical Systems: Model-Based and Data-Driven Approaches
Laleh Akbari; Rolf Drechsler
In: 29. Workshop zu Methoden und Beschreibungssprachen zur Modellierung und Verifikation von Schaltungen und Systemen (MBMV). ITG/GMM/GI-Workshop "Methoden und Beschreibungssprachen zur Modellierung und Verifikation von Schaltungen und Systemen" (MBMV-2026), March 17-18, 2026.
Zusammenfassung
Cyber-Physical Systems (CPS) operate in complex environments where system behavior results from the interaction of
software, physical processes, and changing external conditions. This inherent complexity makes it difficult to fully
understand, verify, or explain how such systems behave. Explainability has therefore become an important system
property, helping users make sense of system decisions, supporting developers in diagnosing unexpected behavior, and
enabling systems to justify actions of other cooperating systems. This review classifies work in explainability according
to a fundamental question: which system artifact is used to generate an explanation? Based on this criterion, we distinguish
between model-based and data-driven approaches. The discussion considers how these approaches generate explanations,
what assumptions they rely on, and the constraints that affect their practical use, while also pointing to issues that remain
insufficiently addressed.
