In a world where top executives often get most of the credit for driving innovation, Tyler Smith of Johnson Controls says that you shouldn’t forget about colleagues who work closer to the customer. “Our sales teams and our technicians and our mechanics that are on the front line of jobs are really important inputs into our innovation process,” says Smith, Vice President of Global Lifecycle Solutions for Johnson Controls.
And collecting their input, he says, can help ensure that new offerings work well — and sell well.
Smith marked his 20th year with Johnson Controls last spring; the company makes a wide range of building-related technologies, from air conditioning systems to fire detection to security. Johnson Controls had $23 billion in revenue in 2024. Smith runs the digital services arm of the more than 100,000 person company, including its platform OpenBlue, which unites building data from different sources.
We spoke to Smith in late October 2025 to get his advice on driving innovation in the built environment.
• • •
In an increasingly digitized world, how does this technological innovation that we’re dealing with right now — including digital twins, cybersecurity, AI and more — impact our built world?
The pace of innovation right now is unlike anything that we’ve ever seen, and that’s being accelerated by what’s possible with AI these days. In the 20 years that I’ve been with Johnson Controls, and the 20-plus years that I’ve been a part of delivering solutions into the built environment, I’ve never seen it like this before.
Our challenge as we innovate for the built environment is to make sure that we’re doing so with customer outcomes in mind. We’ve got digital twins and cybersecurity and machine learning and things like that. How can we leverage those technologies to deliver outcomes that [our customers] care about, like decarbonization, the health and wellness of people in their building, or a return on some type of capital investment?
I’ll give you a couple of examples. None of what we can do with AI or digital twins would be possible without innovation in sensor technology, and we have several things to thank for that. One is wireless capabilities — being able to put sensors in creative places, because they can now communicate wirelessly or be battery-operated, or both.
The next one is digital twins, or a digital recreation of a physical asset. The sky’s the limit with how we can use that digital twin to do things like simulate operating conditions and predict outcomes before they happen by leveraging data. That then enables us to deliver things like risk mitigation, health and wellness, and productivity better than we were able to before.
We like to say that innovation starts with the customer… what that means for us is when we go initiate innovation…our goal is to use inputs from the market, the competition, or macroeconomic trends…
I find it interesting that you said you focus on delivering outcomes that your customers care about. I think this is an interesting point to focus on, because it contrasts the AI washing that you see sometimes, slapping a label on things just because you might find that it invites consumers. When you do proceed with these digital innovations, how do you make sure that core focus on delivering those outcomes that they care about remains at the forefront?
We like to say that innovation starts with the customer, and that’s a cool thing to say, but in practice, what that means for us is when we go initiate innovation — whether it’s in digital services, or it’s a physical product, or it’s a software-as-a-service offering like what we have in OpenBlue — our goal is to use inputs from the market, the competition, or macroeconomic trends that are going to shape the built environment, to make assumptions.
Very quickly thereafter, we’ve got to go validate those assumptions in the market. That’s how we continue to involve people that are currently customers or prospective customers of ours to say, “We understand you care about X, Y and Z. In order to help you solve for that better than you can today, we’ve created A, B and C. What do you think? Can we install it in your building and test it for a little bit? Can we demonstrate to you how it will solve for things that you care about better than you’re currently doing?”
In real-time, we collect feedback, guest test, and revise. We make some tweaks, and we do it all over again, until such time we’ve reached some state of that innovation, of that new product or that new service that we feel is market ready. It’s all about engaging, bringing the customer or prospective customer along with you, engaging them as often as possible to make sure that we’re headed in the right direction.
It sounds like that’s a process in which you’re building trust at the same time. You also talked about predicting outcomes by leveraging data. Can you talk about when that shift from break-fix to predictive maintenance happened?
I think we need to acknowledge that we’re in the middle of it right now. Much like lots of use cases of AI that we’re experiencing on a daily basis, it’s a capability that’s constantly evolving, and that was really accelerated with the technical leaps forward that we’ve seen in AI over the last few years.
We are convinced that we’re in a leading position with predictive maintenance capabilities because of the large data sets that we have. We’re not able to predict something without having some data that is similar to the situation that we need to predict. So we’ve got to be able to train models to say, “I’ve seen this condition before with a chiller that looks like this, and operating conditions that look like this, and some feedback that we’re getting from sensors that look like that, and that data shows that there’s an X percent chance that we’re going to see a failure.”
…We’re already in an advantaged position because of the age of our company, the amount of assets that we have out in the field, our global scale, and the collection of data that we have been doing for some time.
You spoke about how you test new solutions with these prospective customers. I would like to talk a little bit more about your in-house innovation strategies for software, AI and related technologies. Who’s involved in the process, and how do you get from this initial stage of ideation to experimentation to beta and beyond?
We talked a bit about the importance of bringing the customer or prospective customer along. I’ll just expand on that by talking about the various stakeholder groups here at our company that are involved in driving innovation forward. Of course, those that are closest to the customer end up being the most important means of getting feedback to us on what our needs are that aren’t currently being solved for today. What are new outcomes that are emerging in the marketplace that our customers value differently than they did previously?
Our sales teams and our technicians and our mechanics that are on the front line of jobs are really important inputs into our innovation process.
Then, we’ve got the best product management and engineering teams in the business that are receiving requirements and also monitoring macroeconomic trends and competitive activity and trying to make predictions on where things are going to go and how Johnson Controls can leverage our capabilities to get out in front of everything. Once those requirements are set, we start building. It’s through that building process — whether we’re building a new thing like a chiller or an air handler or control system or a fire panel, or a new service like what my team is responsible for, or a new OpenBlue capability that we have to very quickly iterate — it’s through that process that we can go test in the market, get feedback, scurry off with that feedback, make some enhancements, and keep doing that as we move forward.
Then we close the loop. We get to the point where it’s gone through the engineering and product management process, and you can imagine that we’ve got a formal process that we follow to bring new innovation to market. We close the loop by turning it back over to the front liners and enabling them to communicate effectively about it and position it with their customers, and connect the features of whatever that is with outcomes that our customers value. And then the cycle continues.
I think your focus on book-ending those frontline workers who have so much valuable insight and knowledge is interesting, because in having these conversations, I frankly don’t often hear people give them credit for what they’re doing as part of the innovation process.
The worst thing that we could do as a team that’s responsible for innovation is to go create something that our engineers or our product managers think is awesome, get really proud about that, go file a bunch of IP, make something — and then it doesn’t work, doesn’t sell, doesn’t connect as well as it needs to. The value proposition falls flat. The outcomes weren’t considered well enough.
Looking through some of the digital services that Johnson Controls offers, I see all of these updates about all these new capabilities, especially on OpenBlue. So I’m curious: You’re working with clients who want answers to their problems, not more questions. Where is the line between offering comprehensive, innovative digital solutions versus being overly complex? And how do you manage to toe that line?
It really is an important balance. From an innovation standpoint, it all starts with properly segmenting. Where do we want to compete? What segments of the market do we feel like we’re in a better position than others to compete and win? We can’t be everything for everybody. That’s not a good business strategy. With a company of our size and scale, it’s important for us to create platforms that meet a majority of the outcomes that we want to deliver, in markets that we want to deliver, and then that platform can be leveraged across multiple regions and multiple verticals.
It’s kind of hard to imagine what’s going to be the next AI, but there’s going to be something.
Then we can bring value-added features on top of those platforms to make sure that any nuanced, mission-critical environments like hospitals or data centers [have what they need]. But we can’t create tens of thousands of snowflakes every year. That’s not a good way to drive innovation, so we’re striking that balance.
You went from being a project engineer in 2005 to leading the development of digital services in 2025. At a high level, throughout your time at the company, how has Johnson Controls pivoted to remain ahead of the curve in digital innovation?
It’s been quite the ride. …It’s just been a really cool time to be a part of a company like Johnson Controls, and in the corner of the industry that we lead in. …It’s kind of hard to imagine what’s going to be the next AI, but there’s going to be something. And I’m convinced that we’re in a really advantaged position to take our history and our capability and existing relationships that we’ve earned and maintained to help our customers take advantage of whatever the next thing is, to operate their buildings in a better way, to meet their outcomes better than they can today.
You must be logged in to post a comment.