Manisha Bhimrao Mane, N. Vijayakumar, P. S. Sruthi, R. Arulmozhi, D. Suresh
Summary: The study proposes an AI-CPS framework combining DRL, digital twins, and blockchain for autonomous smart infrastructure. Validated in smart grids and transportation, it reduces failures by 45%, boosts efficiency by 50%, and enhances cyber resilience by 35%. DRL enables adaptive control, digital twins predict maintenance needs, and blockchain secures data. Edge-cloud integration cuts latency by 45%. Future work explores federated learning and quantum optimizations. This framework establishes self-optimizing, secure CPS for critical infrastructure, bridging AI and physical systems for a resilient, interconnected world.
CompSci & AI Advances 2(1), 84-94 (2025)
https://doi.org/10.69626/cai.2025.0084