KIT | KIT-Bibliothek | Impressum | Datenschutz

Self-adaptation via joint global-local explanations for autonomous cyber-physical systems

Elahi, Ehsan ORCID iD icon 1; Camilli, Matteo ; Mirandola, Raffaela 1
1 Institut für Informationssicherheit und Verlässlichkeit (KASTEL), Karlsruher Institut für Technologie (KIT)

Abstract (englisch):

$\textbf{Context:}$
Autonomous and self-adaptive systems increasingly rely on black-box predictive models to support runtime decision making in critical domains. However, the lack of transparency of such models enlarges the adaptation search space and increases optimization costs.

$\textbf{Objective:}$
This paper addresses the problem of multi-objective optimization in opaque autonomous systems through cost-effective explanation-driven adaptation.

$\textbf{Method:}$
We propose GLASS, a self-adaptation approach that integrates model-agnostic interpretable machine learning into a feedback loop. Building on our previous work, GLASS combines global explanations (Partial Dependence Plots) with local explanations (counterfactuals) to guide runtime adaptation decisions and reduce the search space toward goal-relevant configurations.

$\textbf{Results:}$
An empirical evaluation on three cyber–physical systems benchmarks shows that GLASS achieves competitive or superior adaptation effectiveness while significantly reducing execution time compared to three state-of-the-art baselines, including up to two orders of magnitude lower overhead than black-box optimization.
... mehr


Verlagsausgabe §
DOI: 10.5445/IR/1000198447/pub
Veröffentlicht am 05.11.2026
Cover der Publikation
Zugehörige Institution(en) am KIT Institut für Informationssicherheit und Verlässlichkeit (KASTEL)
Publikationstyp Zeitschriftenaufsatz
Publikationsjahr 2026
Sprache Englisch
Identifikator ISSN: 0950-5849
KITopen-ID: 1000198447
HGF-Programm 46.23.01 (POF IV, LK 01) Methods for Engineering Secure Systems
Erschienen in Information and Software Technology
Verlag Elsevier
Seiten Article no: 108370
Vorab online veröffentlicht am 21.10.2026
Schlagwörter Explainable AI, Cyber-physical systems, Self-adaptive systems
Nachgewiesen in Web of Science
OpenAlex
Scopus
KIT – Die Universität in der Helmholtz-Gemeinschaft
KITopen Landing Page