AI-driven cooling optimization in data centers: Reinforcement learning for dynamic workload placement and HVAC control – Volume 12 Issue 5

The rapid expansion of data centers has led to unprecedented energy demands, with cooling systems accounting for a significant portion of overall power consumption. Traditional rule-based methods for workload placement and HVAC (Heating, Ventilation, and Air Conditioning) management often fail to adapt dynamically to fluctuating workloads and thermal profiles, leading to inefficiencies and increased operational costs. This paper proposes an AI-driven framework leveraging reinforcement learning (RL) to jointly optimize workload distribution across servers and fine-tune cooling parameters in real time. By modeling the data center environment as a dynamic system, RL agents learn adaptive policies that minimize power usage effectiveness (PUE) while ensuring service-level agreement (SLA) compliance. Experimental evaluations using simulation-based workload traces demonstrate that the proposed RL-based optimization significantly reduces cooling energy consumption compared to heuristic and static policies, while also improving thermal stability across server racks. The study highlights the potential of hierarchical or multi-agent RL architectures to balance competing objectives such as energy efficiency, workload latency, and operational reliability. This research contributes to sustainable data center management by advancing the integration of intelligent workload scheduling with HVAC control, paving the way for greener large-scale computing infrastructures

This paper presented a novel reinforcement learning (RL) framework that jointly optimizes workload placement and HVAC control to significantly enhance cooling energy efficiency in data centers. By leveraging a multi-agent architecture and advanced RL algorithms such as Proximal Policy Optimization and Deep Q-Networks, the system dynamically adjusts workload distribution and cooling parameters based on real-time thermal and workload conditions. The evaluation through simulations demonstrated considerable reductions in cooling energy consumption, improvements in power usage effectiveness (PUE), and a more balanced thermal profile that reduces hotspots and hardware stress. Importantly, the RL framework maintains strict service-level agreement (SLA) compliance during workload fluctuations, achieving a balance between energy savings and performance reliability. The comprehensive offline training, combined with simulation-based fine-tuning, ensures safe and robust policy learning without disrupting live data center operations. The multi-agent design outperforms single-agent approaches, underscoring the benefit of specialized policy learning for different control domains. While challenges remain for real-world deployment—such as handling system heterogeneity and evolving workload patterns—this research lays a strong foundation for integrating AI-driven cooling optimization into sustainable and resilient data center management. Future work includes validation in operational environments and integration with renewable energy and demand response strategies to further reduce environmental impact.

[1] A. A. Alkrush, M. S. Salem, O. Abdelrehim and A. Hegazi, “Data centers cooling: A critical review of techniques, challenges, and energy saving solutions,” International Journal of Refrigeration, vol. 160, no. 1, pp. 246-262, April 2024. [2] H. Lu, Z. Zhang and L. Yang, “A review on airflow distribution and management in data center,” Energy and Buildings, vol. 179, no. 1, pp. 264-277, November 2018. [3] R. R. Ahrabi, A. Mousavi, E. Mohammadi, R. Wu and A. K. Chen, “AI-Driven Data Center Energy Profile, Power Quality, Sustainable Sitting, and Energy Management: A Comprehensive Survey,” in IEEE Conference on Technologies for Sustainability (SusTech), Los Angeles, CA, USA, 2025. [4] N. Ahmed, S. Bennati and N. Anglani, “Energy Performance Benchmarking and Indicators: A Comprehensive Framework for Scientifically Sound Data-Driven and Energy Management Improvements in Hospitals,” in IEEE 8th Energy Conference (ENERGYCON-2024), Doha, Qatar, 2024. [5] Y. Ran, H. Hu, Y. Wen and X. Zhou, “Optimizing Energy Efficiency for Data Center via Parameterized Deep Reinforcement Learning,” IEEE Transactions on Services Computing, vol. 16, no. 2, pp. 1310-1323, 2023. [6] C. Feng, A. K. Jumaah Al-Nussairi, M. H. Chyad, N. S. Sawaran Singh, J. Yu and A. Farhadi, “AI powered blockchain framework for predictive temperature control in smart homes using wireless sensor networks and time shifted analysis,” Scientific Reports, vol. 15, no. 1, p. 18168, May 2025. [7] M. Trigka and E. Dritsas, “Wireless Sensor Networks: From Fundamentals and Applications to Innovations and Future Trends,” IEEE Access, vol. 13, no. 1, pp. 96365-96399, May 2025. [8] K. He, Q. Fu, Y. Lu, Y. Wang, J. Luo, H. Wu and J. Chen, “Predictive control optimization of chiller plants based on deep reinforcement learning,” Journal of Building Engineering, vol. 76, no. 1, p. 107158, October 2023.

Call for Paper (Hub) Fast Paper Publication IJCT Indexing & Databases Submit Manuscript Call for Paper: Sep–Oct 2025 Call for Paper: Publish Research