Data centers are extremely energy-intensive facilities. However, efforts indicate automating their advanced control systems could curb unnecessary resource usage.
Improving User-Friendliness Through Automation
Many data center products offer partial automation, requiring people to confirm specific settings or otherwise exert some control over how the facility operates. These offerings typically include numerous user-friendly features. People can benefit from automated data center platforms without extensive knowledge of the underlying technology.
One example deployed in an Estonian data center will aid the facility in achieving a 1.2 power usage effectiveness metric when the industry average is approximately 1.6. This artificial intelligence-powered system optimizes the facility’s cooling controls for better efficiency.
It also allows users to access a single control panel to monitor, control, optimize and visualize the facility’s operations. An executive associated with the Estonian facility says this platform meets international security standards. He anticipates the technology will allow the data center to be 25% more energy efficient than the industry average.
It would be significant if that expectation plays out in real life since forecasts suggest data centers will use progressively more energy. A 2024 report indicates their electricity consumption could reach more than 1,000 terawatt-hours by 2026. It was only 460 terawatt hours in 2022. The report’s executive summary mentioned the importance of efficiency-related improvements to curb the expected power consumption increase.
Enabling Unstaffed Data Centers
Since data centers have so many pieces of critical, expensive equipment, they must operate within particular environmental parameters. For example, data center best practices suggest maintaining a relative humidity level between 20% and 80% to keep critical equipment running smoothly. Strategic decisions about layouts can also promote energy efficiency by improving airflow.
A 2023 survey showed nearly two-thirds of data centers experience staffing-related challenges. Some people advocate for tackling those difficulties by operating more unstaffed facilities. That option requires automating advanced control systems.
Sameh Yamany — chief technology officer for VIAVI Solutions — envisions a future full of lights-out data centers equipped with various advanced technologies. He explains that less than 1 degree of excessive heat can lead to equipment failures.
However, Yamany believes future data centers will have continually optimized temperature profiles that change according to the real-time data automatically fed into them during operations. Such an automated, optimized system should significantly boost energy efficiency because it means the data center will only consume precisely the resources required.
Most data centers still have staff, but as technologies become more advanced, some jobs should at least become less manually driven.
One commercially available data center automation product has improved environmental monitoring features. It allows people to track temperature, power usage and much more. The specialized information helps technicians identify and respond to hardware problems faster than they previously could. That advantage brings energy efficiency gains by targeting inefficiencies.
Removing Guesswork and Assumptions
What features does a next-generation data center have? After in-depth discussions, one vendor’s data center team unveiled their findings, explaining that these facilities must be sustainable, simplified, reliable and autonomous.
When clarifying the automation aspect, they mentioned operations and maintenance-related features. This tech rollout included participants talking about an energy efficiency optimization feature. It can deliver the best cooling strategy by choosing from 1.4 million possible configurations within a minute.
In another example, a data center company has deployed an artificial intelligence solution across 16 sites in the Asia-Pacific region. Those involved anticipate approximately 18 gigawatt-hours of energy savings stemming from this decision. That’s the equivalent of the electricity used by more than 1,600 homes in the U.S. each year.
The in-house platform can detect abnormal operations and make suggested optimizations without human intervention. Staff members can also refer to all tweaks on a dashboard, seeing them listed in order of estimated gigawatt-hours saved. That helps people make meaningful choices quickly and immediately know what should happen if they do. Such arrangements help them meet energy-efficiency goals in their data centers with fewer doubts.
Is It Time to Use Automation in Data Centers?
Saving energy in today’s data centers makes sense for the operator’s bottom line and helps the planet. Most facilities already have advanced control systems, so automating some features associated with them is a logical next step.
However, anyone considering this option should set goals related to how much energy they hope to save over a particular span and how much of a change such improvements would be over current average power usage metrics.
Doing that is an excellent way to see if a company’s milestones are within reason for the automation’s capabilities. For example, even if an automated system can optimize settings without human oversight, decision-makers may still need to upgrade the data center to match expectations.
Data center leaders will also get the best results by reserving enough time for employee training. Even if the chosen system requires minimal human interaction to work well, people will still need to understand its functionality and how to troubleshoot when necessary.
Planning for all these things will help people determine if now is the right time to implement automation in data to save energy. They can then feel confident about selecting, installing and testing an advanced system.
