Data centers and IT management are up against some significant industry changes at the moment. One of the most important is the major workforce shift involving baby-boomers retiring at a rate of approximately 10,000 a day — a trend that started in 2011 and is expected to continue until 2030. This is happening while the Internet of Things (IoT) continues to expand, with experts predicting a jump from 2017’s count of 6.4 billion connected devices to more than double that in the coming years.

These two topics may not seem connected, but they’re a bit more intertwined than they appear.

The shift in workforce demographics is challenging data center employers to replace industry veterans with green hats, which is problematic due to the loss of institutional knowledge that will occur as retirees depart. It’s an obstacle that will impact most industries, but even more so in highly-skilled professions, such as IT and critical infrastructure.

On the other hand, the growth of IoT is ushering in new data center and network architectures, with more edge locations required to keep up. Subsequently, as the need for edge locations continues to rise, so does the need for distributed data center management. Machine learning can provide data centers with the opportunity to take a more efficient approach to infrastructure management, providing automated operations, predictive alerts, and proactive servicing.

Through machine learning, data centers can identify operational trends (both normal and abnormal) and implement automated management of infrastructure systems, such as power and cooling. By proactively identifying opportunities for efficiencies, machine learning can help systems learn to automatically adapt when triggered, potentially eliminating the need for a technician be on site to make adjustments manually.

While technology may not always be able to fix the problem at hand automatically, machine learning also can assist technicians in their day-to-day work. By recognizing patterns and trends, machine learning also provides the opportunity to shift technicians’ servicing approach from reactive to proactive. Through the use of predictive alerts, technicians can handle maintenance before it creates an issue, minimizing last-minute emergency service calls that require an employee to be dispatched to any location at any hour of the day. What’s more, technicians that have access to trends and a comprehensive knowledge base in advance and onsite (via mobile apps), have a leg up on identifying the problem and solution at hand.

In the above examples, machine learning and artificial intelligence (AI) has the potential to minimize the need for on-site technicians as well as aid new hires with insights on operational trends, best practice procedures and solutions. But there are two key components required to reap these benefits — large amount of data and domain expertise. In order to take advantage of these possibilities, infrastructure management must have the right processes in place to capture and analyze infrastructure data. Even better, IT professionals can strengthen machine learning capabilities with stronger data. Data center companies that aggregate, anonymize and analyze data from various deployments around the world can provide learnings from the industry as a whole as opposed to any one specific location.

It’s clear that neither the retirement of IT, cooling, and power industry experts, nor the explosion of data and the edge network are slowing down. Fortunately, machine learning is just one of the ways data center management can keep pace with both trends.