Data Centers, Artificial Intelligence, and a New Control Approach: A Weidmüller Perspective

[Ahmet Ö.]

## Automation, Security, and Sustainability in Data Centers

Data centers form a fundamental infrastructure for electricity grids and industries, supporting network analytics, distributed energy resource management, and increasingly, artificial intelligence/machine learning applications. However, they also face challenges such as high energy consumption, water demand, and the threat of cyberattacks.

For control engineers, the real question is not whether these pressures are real, but how it is possible to measure, automate, and document performance and security, while also not compromising reliability.

### Weidmüller's Perspective and Industry Approaches

Pete Tecos, Director of Data Centers and Energy at Weidmüller, provides important insights into the challenges focused on sustainability and security. Although there is no federal legal requirement for data center efficiency in the US, DOE best practices and EPA ENERGY STAR certification are widely used. Companies are preparing for the SEC's climate-related disclosure rules and state-level water usage regulations.

In the EU, under the Energy Efficiency Directive, data centers with an IT load of 500 kW or more are required to report environmental performance indicators (such as PUE, WUE, and renewable energy share). Weidmüller is a leader in measuring KPIs compliant with EED and ENERGY STAR.

### Modern Approaches to Data Center Security

Security vulnerabilities in data centers mostly stem not from highly secure servers, but from automation systems. Cooling systems, pumps, valves, and the gateways and protocol converters that connect them pose risks. Weidmüller implements a Zero Trust architecture based on NIST standards (SP 800-53 Rev.5, SP 800-82 Rev.3, SP 800-207).

The risks associated with artificial intelligence are not that the system is not transparent enough, but that it creates a large amount of sensitive data and a wide connectivity area. As a solution, data classification, isolation, policy-based firewalls, and comprehensive logging are implemented.

### Balancing Continuous Updates and High Reliability

The IT world's need for updates must also be met in data centers, but these updates can conflict with long-term reliability. Isolation is possible but not practical. Weidmüller addresses this by designing redundant power and cooling systems for continuous operation, supporting updates at different speeds – fast-IT and planned-OT. Furthermore, every change is observable and reversible.

Weidmüller's superior sensor and telemetry systems are used not only for sustainability reporting but also to objectively prove the impact of software updates on performance and stability.

### Technical References and Standards

• DOE Energy Efficient Data Center Design Guide (2024)
• EPA ENERGY STAR Data Centers
• SEC Climate Disclosure Rules (2024)
• EU Energy Efficiency Directive Data Center Reporting
• ISO/IEC 30134-2:2026 (PUE)
• ASHRAE TC 9.9
• NIST SP 800-53 Rev.5, SP 800-82 Rev.3, SP 800-207

These approaches ensure the efficient management of sustainability, security, and reliability, strengthening the critical role of data centers today and in the future.