AI is actively reshaping operational priorities across the data center, forcing operators to rethink how facilities are designed, managed, and optimized end to end. Simultaneously, the pressure to maintain always-on operations has never been greater.

Yet, delivering five nines of availability has become considerably more complex as AI workloads rewrite the rules across power, cooling, and long-standing infrastructure.

Against this backdrop, in a recent DCD>Talks episode, Marcelo Tarkieltaub, regional vice president for Southeast Asia at Rockwell Automation, explains why the next generation of data center automation is less about adding new layers of technology and more about simplifying operations through flexible, integrated controls.

Always on

For data center operators, reliability has always been the benchmark. But as infrastructure becomes more dynamic, achieving continuous uptime requires a new approach. Tarkieltaub draws parallels with other industries where downtime simply isn't an option:

"If you go to an offshore platform or an industrial facility, those operations cannot stop. We've spent more than a century designing systems that are built to operate continuously, even if something unexpected happens."

Today, resilience must be engineered into data center architecture from the outset rather than tacked on later. This means designing systems with redundancy across every critical layer, ensuring that if one component fails, another immediately takes over without interrupting operations.

"As we build these systems, we introduce redundancy so that even in the unlikely event something fails, the operation continues running," adds Tarkieltaub.

As AI infrastructure becomes increasingly mission-critical, this industrial approach to resilience is finding a natural home inside modern data centers.

Breaking down operational silos

While uptime remains the ultimate goal, achieving this level of reliability is becoming increasingly difficult as facilities grow larger and more complex.

Today's data centers often rely on multiple independent systems to manage everything from building management and electrical infrastructure to fuel systems and DCIM platforms. While each individual system performs its own function effectively, managing them collectively can introduce new challenges.

Tarkieltaub believes the industry must move towards a more unified operational model: "A lot of the time we see different silos. Different systems were designed by various vendors using multiple technologies, and you're trying to manage all that complexity within one operation."

Rather than operating separate control platforms, industrial automation allows many of these functions to be brought together under a common controller architecture to simplify and streamline management.

"The beauty of modern controllers is that they provide a single control platform across your entire facility," continues Tarkieltaub.

Beyond simplifying day-to-day operations, a unified platform also creates a consistent source of operational data.

"We deliver data that's already connected and ready to use," says Tarkieltaub. "Instead of trying to bring together information from multiple disconnected platforms, you create what operators often describe as a single pane of glass."

Beyond upfront cost

Despite the advantages of integrated automation, adopting new technologies or approaches is rarely straightforward. According to Tarkieltaub, conversations with operators often begin with two familiar concerns: if existing systems are working, why change? And can the investment be justified?

"The first thing we hear is, 'We've always done it this way.’ The second is usually about budget."

While modern automation platforms may require higher initial investment than traditional control systems, Tarkieltaub argues the discussion should focus on total cost of ownership rather than upfront capital expenditure.

Data centers are evolving so rapidly that operational requirements often change before construction is even complete.

"From the moment you start designing a data center to the day it goes live, the requirements have already changed," says Tarkieltaub.

As a result, flexibility is becoming just as important as reliability. Rather than deploying systems designed around today's workloads alone, operators must increasingly consider technologies capable of adapting alongside future operational demands.

"The problems we're solving today aren't necessarily the same problems we were solving five years ago,” he continues.

Turning operational data into efficiency

The benefits of modern automation extend well beyond uptime. As AI workloads place increasing pressure on power infrastructure, operators are looking for new ways to improve energy efficiency without compromising performance.

For Tarkieltaub, granular operational data is becoming one of the industry's most valuable assets:

"With modern controllers, you can manage operations in much greater detail. Combined with AI technologies and advanced control systems, that allows you to optimize energy consumption."

Rockwell Automation has already deployed these approaches across existing facilities. "We're seeing energy reductions of between three and eight percent simply by implementing advanced control systems and optimizing operating strategies," adds Tarkieltaub.

While energy efficiency has become a central priority across key markets, Tarkieltaub is confident interest is growing globally as operators seek to maximize existing infrastructure.

"Even in countries where energy hasn't traditionally been the biggest focus, companies are asking how they can optimize operations using these newer technologies."

Simplifying for people, too

One of the less obvious advantages that comes with standardizing automation platforms is its impact on the workforce. Finding experienced professionals remains one of the industry's greatest shared challenges, particularly as AI accelerates construction worldwide.

By adopting technologies already widely used across manufacturing, oil and gas, food production, and other industrial sectors, operators can significantly expand the available talent pool.

"When you have a simpler platform, it's easier to train people," explains Tarkieltaub. "More importantly, you're no longer looking for talent that's unique to the data center industry."

Engineers familiar with industrial automation can transition between sectors far more easily. "That talent can work across food manufacturing, oil and gas, or data centers. Using common technologies allows you to share that workforce rather than building an entirely separate ecosystem."

The same principle applies across the wider supply chain, making it easier to find systems integrators, implementation partners, and long-term maintenance expertise.

"It's a problem many customers weren't expecting to solve. Everything becomes simpler – not just operating the facility, but designing, building, and maintaining it over its lifetime."

The future is flexible

As AI continues to reshape digital infrastructure, Tarkieltaub believes the industry's greatest challenge lies in designing beyond today's requirements:

"Don't just look at the technologies that work for you today. Think about what you'll need five years from now. It's about choosing systems and partners that can grow with your business as your requirements evolve."

Flexibility must increasingly become a defining principle across automation, infrastructure, and technology selection.

Increasingly, future-ready infrastructure won’t be defined by uptime alone – the next generation of data centers will also need to be simpler to operate, easier to optimize, more adaptable to changing workloads, and better equipped to support the people who run them.

To hear more about optimization via automation, watch the full DCD>Talks episode with Marcelo Tarkieltaub, here.