AI and Manufacturing: How HVAC OEMs are Using AI

Daikin Applied Americas, Carrier, and Trane Technologies pulled back the curtain to show how AI has become an integral, yet monitored, part of making life easier for HVAC contractors. 

Where OEM AI Meets Customers 

Ashish Srivastava, chief digital officer at Daikin Applied Americas, said Daikin has implemented AI in a customer experience transformation agenda, including field service optimization, customer support, and remote assistance. 

“The best AI we run today is almost boring in the best way, meaning it’s quietly, consistently useful. We measure our AI agents the same way we’d measure a new team member, with objective performance checks, so the value is grounded in real work rather than a demo,” he said. 

Trane has rolled out its AI Control, which optimizes building systems in real time. It works in the background to predict needs and make adjustments through Trane’s web-based Tracer SC+. Alessandro Araldi, vice president of digital, CHVAC Americas, at Trane Technologies, said this can result in up to a 25% reduction in HVAC energy consumption. 

The company’s ARIA, a conversational AI building agent, helps facilitate teams, diagnose equipment issues, access real-time and historical data, and create charts for troubleshooting. 

“The clearest customer-facing applications of AI are the ones delivering measurable operational outcomes,” he said. “The point is faster decisions and a more intuitive way to manage building performance.” 

In February 2026, Carrier introduced “Tell Me More,” a generative AI-powered feature in its Abound Insights Assistant that gives context and guidance for building operations teams. The Abound platform is connected to more than 150,000 pieces of equipment across a wide range of building environments. 

“The goal is to make support faster, more proactive and easier for everyone involved,” said Nathan Yang, vice president of digital products and technology with Carrier. 

Internal Operations 

In addition to customer-facing programs, OEMs are applying AI within their own operations to improve workflows. These include areas like customer service, procurement, forecasting, supply chain, knowledge management, and product development. 

Yang said Carrier’s focus is to use AI to enhance productivity and empower employees, keeping people at the center of decision-making while offloading repetitive tasks. 

“In certain customer service applications, Carrier has observed significant reductions in call volume and call handling times through the use of AI tools. Within certain SmartHome support experiences, AI is able to resolve a substantial portion of customer inquiries without requiring human intervention,” he said. 

As part of Daikin’s five-year strategy, known as Fusion 30, Srivastava said the company built an enterprise framework known as DaikinIQ that lets it apply AI consistently and at scale. 

“Practically, we’re applying it across functions like engineering, sales and marketing, supply chain and our corporate operations, always starting from the problem we’re trying to solve rather than the technology itself,” he said. 

Contractors can expect product development and engineering to improve as well. All three manufacturers said they’ve incorporated AI into engineering processes. For industrial companies, the problem isn’t necessarily headcount; it’s engineering bandwidth. AI helps close that gap by training it on engineering data to perform repetitive scaffolding work that otherwise consumes an engineer’s time. 

“That frees up more of our engineers’ expertise to be applied where it matters most: evaluating tradeoffs, refining designs and advancing development with greater speed and confidence,” said Tom Gallant, vice president, engineering and technology, CHVAC Americas, Trane Technologies. 

Srivastava said offloading those tasks means engineers can concentrate on higher-value work like conceptual design and system trade-offs. Yang said AI helps engineers analyze large amounts of performance data far faster than before, leading to improved efficiency and smarter connected products. 

“It’s also helping us build capabilities like predictive diagnostics and system optimization within our connected products,” Yang said. 

Protecting Against Errors 

AI is capable of many things, but the technology is far from flawless. It’s prone to hallucinations — when it makes up information or sources out of whole cloth — and “drift,” which is the gradual decline of AI performance caused by its data sets becoming outdated, resulting in outputs being less accurate.  

These can result in mistakes, which is why safeguards need to be put in place when implementing AI-enhanced programs. 

Carrier, Daikin, and Trane experts all agreed that humans need to be a part of the process, as they can detect any errors that AI produces. This also requires training employees in using AI responsibly and appropriately. 

“Carrier approaches AI with a focus on responsible deployment, including governance, testing and oversight processes,” Yang said. “AI is used to support employees by providing information, insights and recommendations, while human expertise remains an important part of business operations and customer interactions.” 

Srivastava said cybersecurity needed to be transformed to ensure that the risk AI tools bring can be remediated and managed. At Daikin, measures are taken to invest in its security framework and ensure only relevant, vetted information is used to train Large Language Model AI. 

“Underneath all of it sits governance through DaikinIQ, an AI Council, a Center of Excellence and a review board that provide oversight,” he said. “Human judgment stays at the center. AI strengthens judgment; it doesn’t replace it.”  

Araldi said that, alongside keeping humans in the loop, Trane Technologies begins with strong data security and governed data pipelines. 

“It also means clear governance around navigating regulatory requirements and maintaining model performance and accuracy,” he said. “These safeguards are what turn AI from an experiment into an operational capability.” 

The Future of AI and HVAC 

It can be impossible to predict how AI will evolve over the next five years, let alone its impact on the industry. Even so, manufacturers peered into their crystal balls and see efficiency gains as the biggest development. 

Araldi said there are three areas he envisions AI affecting the most. The first is automating functions like helping thermal management systems adapt in real time to improve uptime, comfort, and energy performance. The second is helping technicians work more productively to deliver higher-quality service. 

“The third is facility management, where AI can make interacting with building systems much more intuitive, whether that means helping a facility manager quickly understand the source of a comfort issue or generate the right real-time dashboards and reports to manage building performance,” he said. 

Srivastava sees improvements to engineering and product development in the near future thanks to AI developments. 

“That’s where the bandwidth constraint is most acute and where the amplification effect is largest,” he said. “As AI agents mature from handling scaffolding work toward taking on more of the routine engineering load, the firms that learn to integrate them well will unlock more output from the talent they already have.” 

Yang predicts maintenance and service will experience the biggest transformations. 

“We expect that AI will help contractors prioritize service calls, diagnose issues remotely, optimize system performance, and provide more proactive customer support,” he said. “Ultimately, that means fewer surprises for homeowners and more efficient operations for contractors.”