Modeling Sustainability for Autonomous Cyber-Physical Systems

Autonomous cyber-physical systems (ACPS) are becoming central to modern society, powering applications such as autonomous mobility, smart industry, and intelligent infrastructure. As these systems increasingly rely on AI and machine learning, their energy consumption, resource use, and long-term maintenance raise important sustainability concerns. With this PhD project, we aim to investigate the development of a sustainability quality framework tailored to ACPS. The plan is to start by identifying sustainability-related quality attributes relevant to ACPS and then model these attributes using established software architecture techniques according to direct and indirect sustainability impacts. In parallel, the research explores how LLMs can help practitioners navigate and apply the framework by proposing refinements and generating guidance that improves the accessibility and practical adoption of the framework. Most importantly, this sustainability quality framework will be empirically validated through case studies involving real-world ACPS from several domains. Finally, the insights gained will be synthesized into best practices, and anti-patterns to support architects and engineers in designing sustainability-aware ACPS. The results of this PhD are expected to advance both academic understanding and industry practices by offering a structured, actionable framework to reason about and improve sustainability in ACPS.