Across industries, fast and complete adoption of AI is dominating conversation. The problem with speed is it often creates noise, duplicate tools and a workforce that feels like change is happening to them instead of with them.

Chasing AI just to say you’re using it misses the bigger return and keeps your business in the pack rather than breaking out ahead of it. You don’t have to look very far for data backing that statement up. A recent report from PwC states clearly that out of all the companies across global industries racing to adopt AI, only 20% of them are actually leaders, and for a very specific reason: a focus on growth and scaling human work, not just productivity (https://bit.ly/49SFdZ2).

Now, at first glance, it might seem surprising for a mechanical industry magazine to be discussing wholesale adoption of AI. Maybe you’re thinking: “Well, it is the hottest topic of today, so I guess it makes sense.” At the same time, the skilled trades are discussed as being more or less AI-proof and among the last to contend with serious AI adoption.

However, anyone actually on the ground will tell you that isn’t the case.

Mechanical contractors, builders and trade partners are already looking hard at where AI can reduce waste, speed up decisions and help teams spend less time fighting routine problems. Not only that, but where in other industries, AI adoption comes with a fear of “robots taking over our jobs,” in an industry as beleaguered as our by labor shortages and increased work, that same fear can look like opportunity. Maybe even survival.

Still, the rush to adopt can suffer from the same overall look and feel. Bring it in fast, push it through the org as fast as possible. Adapt or die.

Harris, on the other hand, is doing things differently. We’re not interested in putting a shiny label on a tool and calling it innovation. Instead, Harris has invested heavily in the development of an in-house technology team, which I am proud to lead.

We believe long-term advantages don’t appear because we started using Copilot. They show up when we get strategic and intentional about where the technology is suitable; where it’s going to really support the work. Then we take that same, strategic approach to its design, development and deployment. This is how leaders separate themselves from companies just adding to their tech stack.

Proof in action

The strongest case for a thoughtful AI strategy is simple: It has to work in the field, in the office and in the handoff between them. Our technology team focuses on building practical solutions to address real operating pain points, not hypothetical ones. How do we know they’re real? We get our teams together. We host committees. We put our questions to the people out there every day.

Here are a few “rockets” we’ve developed to cut repetitive administrative work, improve access to internal knowledge and help teams move faster without lowering the bar for quality or oversight:

Job finance agent: Digging for answers from outdated spreadsheets is a thing of the past. Our job finance agent brings financial information under one roof, keeping it clean and current. Ask a direct question, get a direct answer from live project data. 

We analyze questions from finance teams and use them to make the tool even smarter and more intuitive. It’s an active system that evolves with projects and keeps relevant information easily accessible. 

Tech talk: For mechanical systems and controls, critical knowledge is often scattered across loose documents, random emails and decade-old notes. It’s a disorganized, but essential library of project history, and can feel nearly impossible to navigate. 

We created a system that tags, indexes and searches every piece of information. Specific details and relevant points of contact are quickly discovered so jobs don’t stall. Years of personal, professional experience are now available via search. 

AI Hub: While the above are great examples of individual solutions to specific pain points, point solutions aren’t scalable. We knew we needed to develop a holistic approach to the development of AI solutions across our organization, one that puts team members in the driver’s seat and invites everyone to play a role in how work at Harris is shaped today and tomorrow. So, we’re building an internal platform, or AI Hub, to house agents that do the heavy lifting by managing training, providing open access, sharing subject expertise and giving real feedback. 

This environment will be a collaborative space that drives improvement. It gives our people tools and space to improve processes and identify opportunities. It’s an investment in our people, giving them the resources to innovate and refine how they work as individuals and how we work as an organization.

In this way, we’re building a platform that scales and evolves with Harris, continuing to deliver real value across the company and beyond it to our customers, vendors and partners. 

Lessons learned

Harris has long described itself as a company built on partnership, problem-solving and support across the full life cycle of a project. On our company site, we emphasize collaboration, practical solutions and long-term relationships, not just with customers, but across teams and disciplines.

That matters here because AI adoption is really an organizational change effort wearing a technical badge. If the culture underneath it is shaky, strategy won’t hold.

That’s also why Harris hasn’t treated AI as something to outsource now and bolt on later. Building internal capability sends a clear message. We want the people shaping these tools to understand our business, our teams and the day-to-day reality of the work. We also want the organization to see AI development as part of our future, not an experiment run by outsiders.

There’s another hard truth here that a lot of companies dance around. Employees can spot empty reassurance in about 5 seconds. Saying, “AI won’t replace you,” at all hands meetings doesn’t hold water if you’re freezing hiring, slashing budgets and laying people off. Smart employees see through it, and if they stop trusting leadership, companies will lose the very people most capable of helping them design useful systems.

People don’t care about slogans and flash. They want stability and they want honesty. Tell them what problems a tool is meant to solve, what expectations are changing and where their expertise fits into the future state. When leadership means what they say, teams can tell.

This is also where the phrase “human in the loop” falls short. It sounds catchy (clearly, since everyone uses it everywhere), but it does the human-technology partnership a huge disservice. Humans don’t just check the machine’s homework. People define the problem, shape the workflow, test the output, catch edge cases, set the guardrails and decide whether a tool belongs in production at all. In serious AI strategy, humans aren’t in the loop. We are the loop.

People-first AI strategy

A people-first strategy usually includes a few basics:

Clear use cases tied to real work.

Honest communication about goals and limits.

Training that respects different roles and skill levels.

Feedback channels that lead to actual changes.

Leadership that treats adoption as a shared effort, not a compliance exercise.

If this seems simple, good. It should be. That’s what “efficiency” is all about.

Good AI strategy falls apart when knowledge gets siloed. That’s true in construction, and it’s especially true in construction companies large enough to have multiple regions, business units and layers of specialization.

If one team holds all the technical knowledge and another team holds all the operational context, neither side gets the full picture.

Harris’ technology team understands this, which is why cross-training matters so much. We’re building curricula across the organization to help develop champions in different departments and regions. That creates a stronger network for adoption. It also makes the process more honest. The best ideas rarely come from one room, one title or one department.

The people building AI tools bring one kind of expertise. The people using them bring another. Project teams know where workflows jam up. Operations leaders know where delays and rework hit margins. HR teams understand how to meet employee needs. Safety leaders know what can’t be compromised. Finance teams can help separate a tool that sounds efficient from one that actually saves money. All that input matters.

Marketing and communications should be in that mix too. Marketing understands how messages are heard, how culture gets shaped and how people process change. They know when language is clear and when it sounds evasive. They can help frame rollout in a way that builds confidence instead of friction. In any company trying to drive adoption at scale, that’s not a side role. It’s core infrastructure.

Strategy matters more than hype

A measured approach takes more time. Construction is full of edge cases, real-world constraints and workflows that don’t fit neatly into generic software logic. A rushed rollout may create a short burst of internal excitement, but it’s often excitement about little things like find-and-replace bots. They’re novel, but they’re not game-changers.

Taking time and being strategic delays that initial dopamine rush, sure, but it compounds the inevitable gratification because it asks harder questions:

Does this tool solve a costly problem?

Can teams trust the result?

Does it reduce friction or add another layer?

Are we creating something that supports skilled professionals or something that interrupts them?

When you get a “yes” to these, you’re about to kick something into a new gear.

AI tools are helping solve issues this industry has been dealing with long before generative AI showed up, including fragmented information, inconsistent workflows, administrative drag and the constant pressure to do more with limited time.

The opportunity is real, yes, but so is the risk if we get it wrong. Harris is positioned to lead because we’re willing to do what it takes to get it right. That’s how Harris has done business for over seven decades.

In conclusion, AI adoption in construction isn’t hard just because the tools are new. It’s hard because useful adoption demands judgment, trust and follow-through. Harris’ approach starts from that reality. We’ve developed and invested in in-house resources because we believe the strongest results come from people who understand the work, respect the people doing it and know how to build for the long term.

We’re serious about a strategy that rolls out AI with care, defines the tools we’re developing, and draws power from the fact that we see culture as part of our technical stack, not separate from it. We cross-train, collaborate and share ownership across departments and regions.

AI isn’t coming to construction. It’s already here. Ask whether your organization is building the conditions to use it well. And as we do that, we carry another responsibility, one that, without question, is more important than anything else: As the first generation of workers in the trenches of a changing employment and industrial landscape, it is our responsibility to pass on what we’re learning to those training tomorrow’s workforce.

This field is changing in real time. Leaders in this space will understand success isn’t a matter of getting to the finish first. They’ll know it’s about getting over the line and then making sure a path exists behind them to carry the next generation forward. 

The gift of AI isn’t a boom of technology at the expense of humanity. It’s elevating the human experience on a foundation of better technology. Leaders, regardless of industry, will see that and apply it accordingly. 

Akira Togawa, vice president and chief technology officer, Harris, is no interloping tech bro stepping out of Silicon Valley and into the trades. He cut his teeth as a mechanical engineer designing power plant cooling systems, got hooked by the journey from paper to pipe, and transitioned into applied technologies where he watched BIM/VDC grow from a niche value-add into a necessity for total project delivery today. That nearly two decade foundation in the world of construction technology is what led him to where he is today.