As AI and high-density computing take over, data center cooling loads are increasing, making them more complex to manage. Data centers face increasing challenges that demand cooling solutions that are designed not just to perform, but to work harmoniously with the building and each other to maximize efficiency, resilience, and future readiness.

DCD spoke to Airedale by Modine’s leadership team to understand how they are responding to emerging industry trends by developing scalable cooling solutions that address the growing complexity and thermal demands of modern data centers, both now and in the future.

Can’t take the heat?

Medium-density data centers are increasingly shifting toward high-density models to support AI applications. These generate significantly more heat and demand cooling solutions that are not only more efficient but also enhance safety by minimizing overheating risks. This shift not only presents technical challenges but also introduces strategic considerations for future-proofing infrastructure.

Airedale’s approach is holistic, rooted in designing system components as parts of an integrated ecosystem – rather than as isolated units – to manage the complexity of data center cooling environments. While each product is engineered for individual efficiency, Airedale’s ultimate goal is to create an efficient overall system.

Designed with built-in redundancy and up to Tier 4 compliance, Airedale’s integrated solutions ensure that when units work together, the combined system outperforms the sum of its parts.

“Customers are concerned about building a traditional data center with medium density, only for it to become a dinosaur very quickly, requiring them to pivot swiftly to liquid cooling. Our focus is on partnering with them to design flexible cooling strategies that evolve alongside their needs, avoiding disruptive pivots and ensuring long-term resilience,” says Jonathan Jones, business development director at Airedale by Modine.

With over 30 years of industry experience, Jones helps ensure Airedale’s cooling technologies meet fast-evolving client needs – particularly the growing demand for hybrid systems that manage both air and liquid cooling.

Balancing efficiency with performance is a constant challenge, further complicated by the rapid, unpredictable evolution of AI workloads.

As the shift toward high-density infrastructure accelerates, it’s not just the cooling hardware that must evolve. Control systems – the intelligence behind the infrastructure – become just as critical in ensuring energy efficiency, scalability, and responsiveness.

Control freaks (in a good way)

When outfitting a data center, the initial focus typically falls on securing the housing structure and major equipment – elements with long lead times that define the facility’s capacity. However, what’s often overlooked at this stage is the integration of control systems that manage and optimize that equipment. Without early consideration of controls and the building management system (BMS), the entire cooling solution risks inefficiency and operational challenges.

Once the capacities of the chosen cooling equipment are determined, seamless integration of controls becomes vital to ensure the system operates at peak efficiency and reliability.

Reece Thomas, controls director at Airedale by Modine, has dedicated 18 years to developing Airedale’s controls division. His leadership focuses on integrating cutting-edge control systems that maximize energy efficiency while keeping pace with dynamic data center workloads – but even this comes with its own hurdles:

“This is especially difficult when you’ve got different manufacturers for indoor and outdoor equipment, and then a generic BMS provider in the middle trying to make it all work. The biggest challenge is that everyone is trying to build as fast and dynamically as possible, often testing and delivering the system on-site in real time.”

Airedale’s expertise lies in bridging this gap early in the design process, as Thomas explains:

“Having that design upfront and being able to optimize it with a control system and a product like ours – already deployed across many data centers totaling over 750MW and counting – gets you one step ahead. You can test everything off-site before delivery and before the ready for service (RFS) dates.”

One of the key enablers of this early optimization is Airedale’s Cooling System Optimizer – a platform that supports pre-deployment testing and adjustment in various data center environments.

“When you look at some of the Nvidia GPUs that are being introduced for liquid cooling and the high densities they introduce to the cooling system, they operate in parallel with very high computing demands. What you will see with that is a strong fluctuation in load, impacting the cooling system. Without understanding how valves or components respond to these load swings, you’re going to run into trouble pretty quickly.”

Flexing their mechanical muscles

All of this points to the importance of partnering with a company that understands every interconnection – from component to system – making Airedale an industry standout, with the business model, experience, and practical skills to provide a full solution alongside every key system component.

“We have a strong understanding of the intricacies of each component within the product, and how the product functions within the overall system – so we build many layers up,” explains Jones.

Drawing from his engineering background, Thomas emphasizes that unlike standard BMS engineers who mainly focus on monitoring and adjusting parameters, Airedale’s approach is more proactive:

“We are cooling engineers who design controls with a deep understanding of mechanical systems, allowing us to flex the mechanical components to their operational limits and maximize system efficiency.”

This hands-on approach enables Airedale’s engineers to adapt system design to real-world challenges,  since every site is unique, with differing layouts, piping, mechanical constraints, and local conditions.

Cooling in perfect sync

Initially, Airedale’s legacy sequencers (pre-programmed steps that a system follows to optimize its performance) were designed to manage outdoor equipment by sequencing chillers. But as systems evolved from primary-secondary setups to variable flow, and eventually to primary-only water delivery, it became clear that traditional sequencing could no longer ensure both efficiency and high redundancy across the full cooling chain.

This challenge was another catalyst for the development of Airedale’s Cooling System Optimizer, a robust control system that fine-tunes every variable, from coil surface areas and pump speeds to water pressures and temperatures.

Then came the next step: optimizing the Optimizer.

“I’ve worked at Airedale for 18 years. Back then, innovations like turbo compressors, EC fans, and electronic expansion valves brought major efficiency gains, cutting energy use for cooling by half,” says Thomas, adding:

“Recent years have seen fewer new technologies, and while refrigerants are now more eco-friendly, some have trade-offs in thermal efficiency, sometimes requiring larger units. I believe we can use data to find the next set of marginal gains in efficiency, so that's where we started our AI venture...”

Beyond the buzz

It’s true, plenty of systems and companies on the market claim to use AI, from apps on phones to building controls. However, many of these solutions aren’t truly leveraging AI. Airedale has embarked on its own journey, with a patent pending for its innovative approach to integrating AI into mechanical and cooling systems.

This method focuses on uncovering marginal gains and enabling systems to anticipate and adjust in advance. In this way, Thomas asserts that Airedale’s goal is to offer the most efficient cooling units on the market – units that will continue to improve over time, driven by AI and data.

Keeping IT cool and hyper-efficient

In the era of AI, it’s essential to view power systems as a crucial part of the cooling conversation – not just temperature control. Across the industry, and increasingly in the public eye, data centers are recognized as major energy consumers. With that in mind, large tech companies are working hard to flip the script, to garner a reputation as both green and sustainable, without impacting performance.

Operational teams tend to prioritize system redundancy to ensure reliability, while senior management focuses on sustainability goals. Airedale's experience and internal resources enable it to develop solutions aimed at balancing operational reliability and sustainability goals.

But true AI isn’t just about optimizing temperature. Airedale is actively using AI to optimize energy use across the entire system, improving power usage effectiveness (PUE) and feeding that intelligence back into the design of products like CDUs, chillers, and fan walls. According to Thomas:

“Our products can easily deliver a really good PUE at the system level, but AI lets us take it to the next step.”

The scale of today’s data centers makes efficiency gains even more critical. What used to be a 100kW hall is now a 10-20MW hall. Even small improvements can result in massive savings in terms of power, carbon, and cost.

“Scale is everything. A one percent efficiency gain today could mean ten times the impact just a few years down the line,” says Patrick Cotton, product management and marketing director at Airedale.

As the third expert in our Airedale leadership trio, Cotton combines a 15-year engineering background with strategic insight, helping shape Airedale’s product portfolio into what it is today, clearly communicating the value of sustainable, efficient technologies to a diverse range of customers.

With limited and sometimes variable power availability, it’s vital to squeeze every bit of efficiency from the system. Airedale’s optimizer achieves this by shifting workloads and fine-tuning component performance to maximize energy efficiency – dynamically balancing between green energy sources and traditional power.

A harmonious relationship

Baked into Airedale’s product design is the concept of “harmony” – a key term that captures the expertise and services aimed at bringing balance and coherence to what can otherwise be a complex and overwhelming data center environment.

Airedale gathers all of the essential ingredients of a cooling system with its Optimizer platform to manage overall system performance, rather than focusing solely on individual unit efficiency. As Cotton puts it:

“We don’t just design coolant distribution units (CDUs) or computer room air handlers (CRAHs) to be the most efficient units in isolation, because individual efficiency alone doesn’t guarantee the most efficient overall system. Instead, we design each component to perform optimally within the complete thermal management system,” adding:

“This holistic approach ensures all parts work seamlessly together, supporting both high- and lower-density loads, maintaining redundancy, and maximizing overall efficiency.”

This approach ensures that all units work in harmony – so when the load increases on one unit, the others respond accordingly, maintaining both redundancy and efficiency.

Reflecting on Jones’s earlier points, there remains uncertainty around how data center cooling will evolve – whether it will lean more heavily on CDU-to-air, CDU-to-liquid, rear-door air cooling, or other emerging technologies. With so many options available, ensuring harmony and flexibility within the cooling system is essential to accommodate future changes. Many data centers operating today could look very different in just a few years, which is why having a system designed to adapt, evolve, and scale over time is crucial.

Collaboration from the first valve to the control room

The collaborative process behind designing a suite of solutions for a BMS is just as important as the products themselves. After all, what successful product is truly designed in isolation? Jones reveals that the Airedale team began by focusing on both component-level and system-level control design, leveraging their multi-disciplinary expertise. By actively engaging with customers, they gained a thorough understanding of how system components would integrate, ensuring customer needs were fully captured before deployment.

Throughout the process, the team works closely with facilities management, operations, engineering, and senior management to ensure monitoring systems target critical areas and provide the necessary data:

“We’re working from the bottom up – starting at the very first valve in the system, all the way through to senior management extracting data from it,” explains Thomas.

The bottom line? Managing development and integration internally allows for closer alignment of solutions with customer requirements.

The power of precision

Airedale’s systems approach to cooling design focuses on optimizing heat exchange and energy use, helping data centers reduce costs and improve sustainability. By refining approach temperatures and adjusting waterside temperature differences, pump power consumption can be reduced, and free cooling potential increased.

For an example of how this works in practice, the Airedale by Modine Energy Efficiency Whitepaper demonstrates how savings of up to 44 percent can be achieved through an integrated, systems-based approach that applies advanced cooling technology from a single provider.

Wrapping up

In today’s fast-evolving data center landscape, adopting a harmonious system that can be continuously optimized and upgraded by a team familiar with the full thermal circuit is a sensible and efficient approach. With diverse stakeholders – from operations teams to senior management – each with their own priorities, customers need a provider that can deliver across the entire ecosystem: components, products, and integrated systems.

Airedale’s approach doesn’t just meet these needs – it evolves intelligently alongside the data center, giving customers confidence that their cooling solutions will continue to perform, no matter how their requirements grow and change.

For more information, please visit www.airedale.com.