Q&A: How AI is Affecting Commercial HVAC Systems

That being the case, experts do warn there are learning curves to consider, and whether equipment can connect with and optimize AI-enhanced controls and systems can be a factor. 

We spoke with Trevor Schick, CEO of AIIR Holdings, and Jamie Cameron, vice president and general manager of OpenBlue at Johnson Controls, about whether AI is making commercial operations more efficient and what HVAC professionals should know as technology advances. 

Editor’s Note: Responses have been lightly edited for clarity and brevity. 

ACHR NEWS: How is AI making HVAC building operations more efficient? 

Trevor Schick (TS): Most of what the industry calls AI in HVAC today is really connectivity. Units report up to a dashboard, which represents progress, but the loop for self-sustaining impact is still open. The unit is not being given authority to change how it operates. And, even when it is, most equipment was not built with real intelligence in mind. A single-stage compressor and a fixed-speed fan give the controls one decision to make: on or off. Without fully variable operations and data-rich sensing built into the controls, there is not much for AI to act on. Close that loop with the right equipment and the math changes.  

Jamie Cameron (JC): AI can evaluate large amounts of data, including equipment performance, occupancy, weather, energy use, and facility conditions, and adjust operations based on what is happening in the building. With OpenBlue, we’re seeing this shift from simply monitoring facilities to actively optimizing them.  

For example, Energy & Comfort Intelligence uses AI to tune air handling based on real conditions. At financial software company DATEV, the technology reduced airside energy use by nearly 10% on applicable equipment while the system autonomously executed thousands of data-driven decisions. Importantly, the facility team did not receive a single comfort complaint. 

ACHR NEWS: Is there a learning curve to implementing AI into commercial buildings and their controls? 

JC: There is a learning curve, but the first challenge usually isn’t AI itself. It’s making sure a facility has the right data, organized in the right structure, and a clearly defined operational problem to solve. AI readiness starts with data readiness. The organizations seeing the fastest results are the ones that have invested in connecting and contextualizing building data so it can be used to drive operational decisions. 

TS: Yes, and it is mostly a human learning curve, not a technical one. Building staff and technicians have spent their careers thinking in setpoints. Intelligent controls work toward a specific outcome instead: a comfort target that combines temperature and humidity. The system chooses the path to get there. That is a mindset shift that is easy to underestimate.  

The system is powered by AI at its core and learns what it needs to know about the space over its first weeks of operation; no one onsite needs to “learn the system.”  

ACHR NEWS: How do AI controls change what HVAC technicians actually do on service calls? 

JC: AI can give technicians something they have historically lacked — a much clearer picture of what is happening before they arrive on site. Instead of starting with a broad alarm or a report that something “isn’t working,” AI can analyze equipment behavior, identify abnormal patterns and help pinpoint where attention is needed. That means a technician can arrive with more context about the likely issue and the equipment involved. 

TS: Today, a technician walking into a guest room closet is diagnosing by feel, often on a unit that was installed by someone else years ago with no records. With connected equipment, the call starts with recorded readings, telemetry measurements that tell what the unit was doing when the alarm fired, how it compares to its own history, and what the most likely root causes are. That shrinks the diagnostic portion of the visit and cuts down on no-problem-found trips that frustrate everyone.  

ACHR NEWS: What safeguards are in place to prevent the AI from operating equipment outside specifications, or sacrificing occupant comfort for efficiency? 

TS: About equipment protection, the AI does not have authority over the protective layer. Compressor limits, refrigerant pressures, coil temperatures, and low-ambient behavior are enforced by the equipment controls regardless of what the optimization layer requests.   

Comfort is the objective, and efficiency can be the constraint on how we reach it. It is not a trade-off with the algorithm deciding a guest can be a little warm to save a few watts. The comfort target is the fixed point; the system is choosing between paths, such as dehumidifying versus cooling, all aimed at reaching that target. 

JC: Efficiency cannot come at the expense of safety, reliability or occupant comfort. AI-enabled controls are designed to operate within defined parameters rather than having unlimited authority over a facility. For example, OpenBlue’s Energy & Comfort Intelligence can adjust zone heating and cooling setpoints while maintaining defined minimum and maximum limits.  

ACHR NEWS: How is equipment keeping up with AI-driven HVAC controls? 

JC: AI doesn’t require every piece of HVAC equipment to be replaced with a new generation of “AI-ready equipment.” In many cases, the opportunity is to make better use of the equipment and data that buildings already have. 

There may be cases where a facility needs additional sensors, connectivity, or controls upgrades to take full advantage of AI. But AI should be viewed as a way to build on existing infrastructure, not something that is only available to brand-new facilities. 

TS: Broadly, it is not. Controls are only as smart as the equipment lets them be. To get value from AI, the unit needs to be updated with real degrees of freedom: a variable-speed compressor, modulating fans, electronic expansion, and enough onboard sensing to know what the system is doing. This is why we design our units and our controls together from the start, rather than layering software onto a legacy platform.  

ACHR NEWS: Where do you see the technology heading in five years? 

TS: First, room-level systems will operate cohesively as a fleet, not as isolated units. A property will manage load, sound, and comfort across the whole building, with the units coordinating among themselves.  Second, the comfort index will replace the setpoint as the target. Owners will stop asking, “What is the room set to?” and start asking, “Is the room comfortable and what did it cost?”  Third, intelligent packaged heat pumps will become a credible alternative to VRF and chilled water for mid-scale hospitality and multifamily. 

JC: I think we’ll see buildings become increasingly intelligent and adaptive. Today, we’re moving from dashboards and alerts to AI that can identify patterns, make recommendations, and automate specific decisions. Over the next five years, that evolution will continue as buildings become better equipped to coordinate systems, anticipate changing conditions, and continuously optimize performance.