Harness the Power of AI to Optimize Newly Commissioned Data Centers for Reliability and Sustainability

As data centers become larger and more complex, operators face increasing pressure to meet sustainability goals, manage energy constraints, comply with new regulations, and maintain uptime during extreme and unpredictable weather events. Newly commissioned facilities are especially vulnerable, as teams must understand and tune systems that have never been tested under real-world conditions.

The challenge is clear:
How can operators meet PUE and WUE targets while maintaining reliability and uptime?

AI offers a solution. With the right intelligence platform, operators can simplify complex environments, reduce energy use, and improve reliability using the sensor data they already have. No new hardware and no complex IoT rollouts required. Time to value: measured in weeks.

Optimizing a Newly Commissioned Data Center

Lucend has optimized both new and legacy data centers globally, delivering more than 18,500 MWh of proven energy savings across customer deployments (including reaching the 14,000 MWh milestone highlighted in an earlier update).

One recent example involved a 20 MW, state-of-the-art data center equipped with modern hybrid cooling systems. Within a few months, results included:

• More than 4,000 MWh per year in energy savings
  
  
• Successful achievement of PUE targets
  
  
• Improved asset reliability
  
  
• An operations team empowered to focus on mission-critical work
  
  

After integrating facility sensor data into Gradient, Lucend’s intelligence platform, several major opportunities were quickly identified and quantified:

• Hybrid cooling units were running in mechanical mode even when outside air temperatures were as low as 8°C / 46°F
  
  
• Many cooling units were operating simultaneously, elevating operational risk and increasing energy consumption

Getting to the Root Cause

Data centers generate hundreds of billions of sensor readings. Gradient uses machine learning to analyze these patterns continuously and pinpoint the true causes behind inefficiencies or risks. This enables:

1. Automated analysis of complex behaviors occurring across the full environment
   
   
2. Clear diagnosis of root causes despite the thousands of potential interactions that can influence outcomes
   
   

In this data center, optimizations centered around three key issues:

1. Free cooling strategy hysteresis

Inefficient control logic caused cooling units to enter mechanical mode unnecessarily during winter months.

2. Pump and pressure misalignment

ML models identified large gaps in pump sequencing setpoints and high differential pressure in the chilled water ring. This caused too many cooling units to run simultaneously at high speeds.

3. Inconsistent operation of adiabatic cooling

Some chillers were operating with adiabatic cooling enabled, while others were not. When adiabatic mode was off, energy use increased significantly.

Gradient provided clear diagnostics along with quantification of wasted energy. Visualizations such as free cooling operation per unit showed thousands of hours of “unnecessary mechanical cooling,” enabling the team to take immediate corrective action.

Effective Measures and Assessing Impact

Once operators understood the underlying issues, Lucend’s specialists recommended targeted changes. These included:

• Adjusting control strategies
  
  
• Lowering differential pressure step by step
  
  
• Increasing sequencing setpoints to reduce the number of hybrid cooling units running simultaneously
  
  

Using Gradient’s advanced monitoring capabilities, operators tracked the impact of each change to ensure the entire cooling ecosystem returned to stable, efficient operation as quickly as possible.

Next Steps

The customer is now onboarding a second 20 MW data center on the same campus. With Gradient and the insights gained from the first facility, the team can achieve rapid onboarding, hit PUE and WUE targets sooner, and meet customer SLAs with greater predictability.

This example reflects a broader trend. As data centers scale in complexity, AI-enhanced decision-making is becoming essential. Gradient helps operators identify opportunities quickly, understand root causes clearly, and improve performance sustainably.

Lucend’s advisory services teams continue to work alongside local operations teams to share best practices, evaluate changes, and support long-term reliability and sustainability goals.

Are Your Data Centers Operating at Their Full Potential?

Inefficiencies often stay hidden until they affect performance or reliability. Lucend can help uncover them early.

To learn how Gradient and our expert teams can support improved reliability, reduced energy costs, and faster operational insight, contact us at hello@getlucend.com.

Stay optimized with Lucend.