Generative AI for Self-Adaptive Systems: State of the Art and Research Roadmap

TL;DR

This paper investigates how Generative AI, especially LLMs and diffusion models, can enhance self-adaptive systems (SASs) by mapping GenAI capabilities onto the MAPE-K feedback loop and HOTL interactions. It provides a two-fold analysis: (i) augmenting SAS autonomy through improved monitoring, analysis/planning, execution, and runtime knowledge, and (ii) improving human-in-the-loop collaboration via preference acquisition, transparency, and cooperative workflows. Through a systematic literature survey across AI, SE, HCI, and robotics, the authors propose a comprehensive roadmap outlining design-time-to-runtime transfer, LM-as-a-service, observation/representation, decentralized control, personalized interaction, ethics, evaluation artifacts, self-testing, and self-evolution, while highlighting inherent GenAI limitations and mitigation strategies. The work serves as a practical guide for researchers and practitioners to identify opportunities and challenges in deploying GenAI in SASs, emphasizing rigorous evaluation, safety, and governance. Overall, the paper consolidates current insights and charts an actionable, multi-year research trajectory toward robust, transparent, and user-aware GenAI-enabled SASs.

Abstract

Self-adaptive systems (SASs) are designed to handle changes and uncertainties through a feedback loop with four core functionalities: monitoring, analyzing, planning, and execution. Recently, generative artificial intelligence (GenAI), especially the area of large language models, has shown impressive performance in data comprehension and logical reasoning. These capabilities are highly aligned with the functionalities required in SASs, suggesting a strong potential to employ GenAI to enhance SASs. However, the specific benefits and challenges of employing GenAI in SASs remain unclear. Yet, providing a comprehensive understanding of these benefits and challenges is complex due to several reasons: limited publications in the SAS field, the technological and application diversity within SASs, and the rapid evolution of GenAI technologies. To that end, this paper aims to provide researchers and practitioners a comprehensive snapshot that outlines the potential benefits and challenges of employing GenAI's within SAS. Specifically, we gather, filter, and analyze literature from four distinct research fields and organize them into two main categories to potential benefits: (i) enhancements to the autonomy of SASs centered around the specific functions of the MAPE-K feedback loop, and (ii) improvements in the interaction between humans and SASs within human-on-the-loop settings. From our study, we outline a research roadmap that highlights the challenges of integrating GenAI into SASs. The roadmap starts with outlining key research challenges that need to be tackled to exploit the potential for applying GenAI in the field of SAS. The roadmap concludes with a practical reflection, elaborating on current shortcomings of GenAI and proposing possible mitigation strategies.

Figures (6)

• Figure 1: Trend in the number of papers with different title keywords in conferences across various fields. Here, the solid line represents papers with "Transformer" in the title, while the dashed line represents papers with "language model" or "LLM" in the title. We can observe that Transformer has consistently been a prominent research topic in AI and robotics, while the application of language models in each research field shows rapid growth from 2023 to 2024.
• Figure 2: Self-adaptive System with MAPE-K Feedback Loop Andersson2009.
• Figure 3: Literature Categorization Overview. One piece of literature may be involved in multiple categories.
• Figure 4: Overview of Empowerment of MAPE-K Modules via GenAI.
• Figure 5: Overview of Empowerment of Human-on-the-loop via GenAI.