What is it about?
This work introduces AWARE, a new framework for adapting complex software systems. Unlike the traditional MAPE-K loop—which is centralized, reactive, and sequential—AWARE relies on intelligent agents capable of learning, collaborating, and acting proactively. These agents can anticipate changes, generate adaptation strategies, carry out actions, assess results, and enrich their knowledge. Through this distributed, modular, and continuously learning design, AWARE enables systems to be more autonomous, scalable, and resilient. A real-world case study shows how AWARE outperforms MAPE-K in dynamic environments.
Featured Image
Photo by Mohamed Nohassi on Unsplash
Why is it important?
AWARE represents a significant advancement in the field of self-adaptive systems. By introducing distributed intelligence and continuous learning, it addresses the needs of modern systems that must quickly adapt to ever-changing environments. AWARE enables proactive problem-solving instead of reactive responses, supports better resource management, and improves system reliability. It also simplifies the integration of generative AI (like LLMs) without adding excessive complexity. This framework opens the door to a new generation of autonomous systems in domains such as cloud computing, IoT, critical infrastructure, and more.
Perspectives
Writing this article was a rewarding experience, as it marks a milestone in our research on adaptive systems. It challenges a 20-year-old paradigm (MAPE-K) and proposes a new approach centered on intelligent collaboration, learning, and proactivity. This work is already generating interest in Industry 4.0, cloud infrastructures, and cybersecurity, where adaptability and resilience are essential. We hope this framework will serve as an inspiring foundation for researchers and engineers working on distributed adaptive intelligence.
Brell SANWOUO
Read the Original
This page is a summary of: Breaking the Loop: AWARE is the New MAPE-K, June 2025, ACM (Association for Computing Machinery),
DOI: 10.1145/3696630.3728512.
You can read the full text:
Contributors
The following have contributed to this page
