Experts around the construction industry have echoed for years that for artificial intelligence, data quality is king. Doug Harrison, vice president of corporate operations at Suffolk Construction, is one of those voices.
Harrison has spent the better part of a decade with Suffolk, where he started as a project engineer in 2018. Now, Harrison is heavily involved in how the company folds AI into its workflows, a process that aims to take scattered data and present it in a way builders and executives can understand.
“This is such a fragmented and disjointed business and environment that we work in here in construction, and the world of AI is truly bringing that together,” Harrison told Construction Dive.
To that end, Suffolk has established its Jobsite of the Future program, which embeds AI engineers with project teams to help solve key problems and build new tech tools for successful construction.
Here, Harrison talks with Construction Dive about how Suffolk is adding AI into its workflows, how it standardizes technology across its business and what it considers when it’s looking to buy a tech solution versus when to build the tech itself.
Editor’s note: This interview has been edited for brevity and clarity.
CONSTRUCTION DIVE: What initiatives has Suffolk started to embed AI into its construction process?
DOUG HARRISON: It starts with our Jobsite of the Future, which is an initiative to embed AI into our project teams.
It pairs an AI engineer with a project team in a specific sector, because most of our sectors have differences in the way that we deliver. A lot of the underlying principles of construction are still the same, but there's unique aspects of mission critical work versus healthcare versus higher ed, for example.
Pairing an AI engineer with our project teams really has two fronts.
It gives our project teams more of your typical licenses — Claude, ChatGPT — like you see across the industry right now. Our question is, how do you get day-to-day efficiencies right and train those individuals that are on the projects to actually utilize it? We want them to use that to ingest documentation and do comparisons for change order reviews, RFI creation, etc. That's a part of their role.
The bigger piece that is actually going to, we believe, create transformational change in the industry is the development we’re working on for tools and applications that go hand-in-hand with our project teams.
How do you think about deploying technology to your teams?
The biggest thing that we realized is that you really need to have consistent technology deployed to capture the data first.
It's incredibly important that before you say this is a required tool, you have the backend infrastructure set up to say, what data are we going to be pulling from this? Is this going to be a source of truth for us to manage our projects at a regional level?
Because all of this data flows up, not only for our project executives, but this same data goes into our dashboards that our COOs and our GMs and ultimately our presidents, use to manage their portfolio.
You want to think about consistent data collection and technology deployment in the same light.
That’s what has let us be successful in standardizing tools across our portfolio.
What goes into standardizing those tech tools?
It's got to prove long-term value. It has to change the process that we execute today for the better.
Right now, we're working on finding the right partner and platform for procurement and delivery, as an example.
Yes, all of these applications are figuring out ways to embed AI into their solutions, but some have better workflows for the way our company operates. Finding that match helps us manage procurement and delivery.
So, there's a big piece of functionality, right? You have to get the field users in there to actually test the tech. You have to make sure you get the results you want, and that the usability is blessed by those project teams. That's where you want that early adoption.
How you actually standardize that and then deploy it is really a training strategy, and how do you make sure that you have the right robust training for these individuals and these specific platforms.
Contractors today are wondering whether it’s better to build or buy their tech solutions. What’s Suffolk’s position on the build vs. buy calculus?
It's a constant debate before we go forward with any solution prototyping.
I think the relationship with Suffolk Technologies in the BOOST program affords us the opportunity to see the full view of the landscape and how this is shaping up.
The biggest thing we look at is, if it's on the forefront of a platform we currently use or there's something to test in the market, we are not going to build that application unless it is something that is really bespoke to the way we work. We understand for standard applications, we're not going to build and beat some of the models that are out there.
We're not a software company. We're a construction company at the end of the day, and what we want to do is deliver best-in-class value to our clients and to our trade partners. And ultimately, we want to make it the most productive solution possible for our team members to scale and grow in the organization.
