Toward Physics-Informed Machine Learning for Data Center Operations: A Tropical Case Study

2025-05-26

AI Summary

Tropical data centers suffer from excessive cooling energy consumption and degraded reliability due to high temperature and humidity. Existing purely data-driven models exhibit poor extrapolation capability and insufficient safety guarantees. To address these limitations, this paper proposes a physics-informed machine learning (PIML) framework integrating thermodynamic and fluid dynamic principles. It is the first to deeply embed multi-physics partial differential equations into a neural network architecture, design a first-principles-based loss function, enable physics-constrained neural modeling, fuse heterogeneous multi-source data, and support real-time online optimization and control. Validated on a real tropical data center, the system reduces cooling energy consumption by 18.7%, achieves a PUE prediction error of less than 2.3%, and attains a 94.1% fault detection accuracy under extreme operating conditions. The approach significantly enhances model generalizability, interpretability, and operational trustworthiness.

Technology Category

• Machine Learning: Hardware-aware ML
• Humans and AI: Human-in-the-loop Machine Learning
• Reasoning under Uncertainty: Relational Probabilistic Models

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• User Modeling, Personalization and Recommendation: User privacy protection in personalized systems
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions

Abstract

Data centers are the backbone of computing capacity. Operating data centers in the tropical regions faces unique challenges due to consistently high ambient temperature and elevated relative humidity throughout the year. These conditions result in increased cooling costs to maintain the reliability of the computing systems. While existing machine learning-based approaches have demonstrated potential to elevate operations to a more proactive and intelligent level, their deployment remains dubious due to concerns about model extrapolation capabilities and associated system safety issues. To address these concerns, this article proposes incorporating the physical characteristics of data centers into traditional data-driven machine learning solutions. We begin by introducing the data center system, including the relevant multiphysics processes and the data-physics availability. Next, we outline the associated modeling and optimization problems and propose an integrated, physics-informed machine learning system to address them. Using the proposed system, we present relevant applications across varying levels of operational intelligence. A case study on an industry-grade tropical data center is provided to demonstrate the effectiveness of our approach. Finally, we discuss key challenges and highlight potential future directions.

Problem

• Addressing high cooling costs in tropical data centers
• Improving machine learning model reliability for data center operations
• Integrating physics into data-driven solutions for operational optimization

Innovation

• Physics-informed machine learning for data centers
• Integrates multiphysics processes with data-driven models
• Optimizes tropical data center cooling operations