Engineering Proactive Self-Adaptation in Cyber-Physical Systems under Uncertainty
2026 (English)Doctoral thesis, comprehensive summary (Other academic)
Sustainable development
SDG 9: Build resilient infrastructure, promote inclusive and sustainable industrialization, and foster innovation
Abstract [en]
Cyber-Physical Systems (CPS) are characterized by their complex interactions between computational and physical processes, uncertainty, and continuous change. Self-adaptation was introduced to address uncertainty and continuous change. Most existing self-adaptation approaches are reactive: the system monitors for goal violations and initiates corrective action upon detection. In CPS, however, some actions are irreversible, and delayed or incorrect responses can cause safety issues. This becomes even more challenging under runtime uncertainties, such as incomplete or inaccurate information during operation, in which the system may react only to visible symptoms, leading to repeated violations rather than stable operation.
To deal with these limitations, this thesis advances the engineering of proactive self-adaptation in CPS under uncertainty. It argues that, by shifting the decision point ahead of any violation, proactive self-adaptation enables CPS to anticipate future states and act while corrective action is still effective, and adaptation decisions can be refined as new runtime information becomes available. However, achieving this requires addressing fundamental challenges, including complex system dynamics under real-time constraints, the presence of uncertainties beyond pre-defined adaptation logic, and the need for richer representations of systems in the case of collaborative CPS.
To tackle these challenges, the thesis adopts a design science approach and develops a set of complementary architectural and methodological contributions. It introduces PANCS, a reference architecture that integrates adaptive model predictive control into the MAPE-K loop to enable anticipatory adaptation in nonlinear CPS. It further proposes a digital-twin-augmented architecture that continuously refines system knowledge and supports reasoning under evolving and partially unknown conditions. Finally, it presents the EA Blueprint Pattern for constructing and evolving Digital Twins of the Organization, extending proactive adaptation principles to higher levels of abstraction, i.e., collaborative CPS scale.
The contributions are demonstrated and evaluated through simulation across multiple CPS domains, including autonomous ground vehicles, robotic manipulators, and an automotive organization. Results show that proactive self-adaptation, when supported by predictive reasoning, continuous model maintenance, and principled architectural structure, continues to satisfy CPS's functional and non-functional adaptation goals during its operation as system and environmental conditions change.
Place, publisher, year, edition, pages
Växjö: Linnaeus University Press, 2026.
Series
Linnaeus University Dissertations
Keywords [en]
Cyber-Physical Systems, Self-adaptation, Proactive adaptation, Uncertainty, Runtime, Reactive adaptation, Nonlinearity, Digital Twin, MAPE-K, Software architecture, Robotics, Organization, Digital Twin of the Organization
National Category
Computer and Information Sciences
Identifiers
URN: urn:nbn:se:lnu:diva-150396DOI: 10.15626/LUD.633.2026ISBN: 978-91-8082-502-3 (print)ISBN: 978-91-8082-503-0 (electronic)OAI: oai:DiVA.org:lnu-150396DiVA, id: diva2:2114189
Public defence
2026-11-17, N1017, N-building, Linnaeus University, Växjö, 13:00 (English)
Opponent
Supervisors
2026-11-032026-10-302026-11-03Bibliographically approved
List of papers