Why Every Construction Company Needs AI Right Now ( Real Examples Inside)

Construction Hot Takes: Leveraging AI in Construction

Host/Intro: We’ve helped hundreds of clients build better construction companies. We love this business. We love the people and we’ve seen what running a successful, profitable business takes. And we’re sharing it all with you here on Construction Hot Tapes.

(Teaser Snippet from Adam Cooper): Yeah, I don’t think there’s a reason not to get into AI if you’re a construction company. As I said, what I heard last week was if you’re not already on the AI train, you’re late and it’s coming in everywhere.

Adam Cooper: Alright. So, I was at this business leader conference out in Vegas last week. And the hot topic across all the different sectors that were there was artificial intelligence, how people are leveraging AI. And I’ve been dabbling with it. I’ve been playing with it. I know a lot of us here have. The marketing team is probably doing more with it than I have been doing. And I had some people showing me some things and it blew my mind how much it’s come, how far it’s come, how advanced it’s gotten, how superior it is to what I looked at just 12 or eight months ago. It has grown exponentially in its power and its usability. And I started thinking about ways that we could use it internally, but I was thinking about how construction companies could start leveraging AI. And I know that I don’t think I’ve talked to any of our clients in the last six months that have, you know, promoted how much they’re using AI.

Jeff Robertson: No, I could say zero for me.

Adam Cooper: Yeah. So, I think a great topic for us to talk about off the cuff. How could small to mid-sized construction companies start using AI in their business?

Jeff Robertson: So, I got a couple of thoughts on this. This is something that I thought of a couple of weeks ago. It’s not like I invented it because I did a little research and it turns out somebody’s out there trying to figure it out: using AI for predictive analysis of scheduling.

Adam Cooper: Mhm.

Jeff Robertson: So, you know, I don’t know, your P6 schedule or your Microsoft Project schedule, whatever. Plug it in, assuming you’re keeping it up, you know, you’re progressing it properly. Being able to analyze delays, forecast… you know, if you can lay in lead times, etc., etc. And, you know, putting together a delay claim is a pain in the ass. It’s really hard to do. It’s difficult to prove sometimes, despite the best, you know, the best dates and all the stuff you’ve kept. Showing it on paper sometimes just does not work. That would be huge. I wish I had that when I was, you know, in a trailer trying to figure out how to prove that we had a delay.

Adam Cooper: Right. Building on that, you know, one of the things that I find it’s really good for is just writing those emails, writing those letters, just crafting the right language. Like, tell it to write me an official letter noticing the owner that I’ve been delayed this many days. Here’s the contract. Please reference, cross-reference the contract clauses that apply to this. And it can do all of that and generate a letter in like 30 seconds. You can load in your contract and load in the weather reports and load in the schedule and have it generate all that for you.

Jeff Robertson: Procore has some predictive analysis built in now.¹

Adam Cooper: Yeah, the co-pilot feature.

Jeff Robertson: Uh, no, it’s… there’s something else. I can’t remember. I’m blanking on what it’s called. Anyway, according to them, I haven’t played with it yet. According to them, you can use it. It’ll look at RFIs, other correspondence, submittals, etc., etc., all your contract documents. It’ll scan all that and help you. It’ll tell you, “You have a problem over here. You have a cost problem over here. You might have a change issue here.” It’ll pull all that forward for you on a dashboard to say, “Okay, go chase the following 12 things down to either make sure they don’t happen or make sure they do, depending on what it is.”

Adam Cooper: Right. Uh, building on that, there’s a few third-party apps that are really leveraging AI that we’re familiar with. I’m going to name-drop a little bit. There’s a company here in Atlanta that we’ve gotten to know, Document Crunch. They use AI to read contract documents from a legal lens, was developed by construction lawyers. And then it gives you like an output inside Procore or as a standalone that tells you all the things you need to be worried about in your contract. So, it kind of pulls all that stuff out and highlights it for you. And then it can create documents around it. It can create RFIs. It can create submittals around that. It can really help you manage the legal side of your contracts.

Jeff Robertson: There’s a lot of cool stuff out there right now. I’ve seen… I don’t remember the name, but you and I saw them at a conference. We were at the, uh… the glasses.

Adam Cooper: Yes.

Jeff Robertson: So, it videos what you see.

Adam Cooper: Mhm.

Jeff Robertson: Uh, you can have the augmented reality stuff, but…

Adam Cooper: Right, with the model superimposed over reality.

Jeff Robertson: …but what they’re also doing now is they’re using AI to analyze the video and find quality and safety problems, right? So, you may not see it because you’re looking, you’re doing a punch, I don’t know, whatever you’re doing. You’re walking around…

Adam Cooper: …and you have a camera on your helmet.

Jeff Robertson: You have a camera on your helmet, or maybe it’s in your glasses or whatever, and it sees something over here that’s just not where you’re… it’s out of your peripheral vision a little bit. It’s a guy standing on a bucket, say for example. It will alert you.

Adam Cooper: Oh, that’s pretty cool.

Jeff Robertson: It’s… it’s freaking sci-fi. I don’t even understand it.

Adam Cooper: And then on the, uh, flipping over to another area, uh, we have… there’s a couple of companies out there, but we were working with a client with Field Materials right now.

Jeff Robertson: Yes.

Adam Cooper: Field Materials uses AI to generate purchase orders, receive invoices and delivery tickets from the field, matches them all, and helps you automate or semi-automate the purchasing and procurement and AP process.² Right. It’s not OCR. It’s AI that’s reading.

Jeff Robertson: It’s AI that’s reading everything and then making sure that what you ordered is what was delivered and it was invoiced at the proper quantities to match the delivery. And it’s helping people make sure that they’re not overpaying or underpaying or paying for material that wasn’t delivered.

Adam Cooper: Yeah, that’s really slick to be able to get that. I mean, it’s, you know, keying an invoice alone is labor-intensive, etc.

Jeff Robertson: Right.

Adam Cooper: Matching them up is even more labor-intensive. So, it can save a lot of money on the back end. I’ve heard about some of our clients using AI just to write proposals. Like, they feed it their proposals from previous projects and then say, “Write a new proposal that, you know, closes any gaps here or that writes out the scope in a more professional manner” because they don’t have a professional proposal, so they use AI to write a better proposal. Right? These are like some easy ways that clients could be using it.

Jeff Robertson: Well, you make a good point there. You know, some of the software solutions are, they could be pricey, they can be. And they’re maybe more designed for more of an enterprise kind of like, you know, if you’re going to have robot dogs walking around your job sites…

Adam Cooper: …which we’ve seen at the conferences as well.

Jeff Robertson: They’re creepy.

Adam Cooper: Yeah, it’s very creepy.

Jeff Robertson: Which is creepy. Um, but the same basic technology, it’s a robot dog, but they’re using… it thinks in some way or another. Um, but my point is, you know, on the scale of cost and availability and usability, that’s up here on the far end of, “You know, I got a couple tens of thousands laying around, let’s try a robot dog concept,” versus, “You know, I’d just really like my proposals to be a little bit more fine-tuned and I don’t have an expert to tell me how to do that.”

Adam Cooper: Right. There’s a pretty broad range there.

Jeff Robertson: Yeah. I think there’s just a lot of opportunity for even the small construction companies to start leveraging things like ChatGPT, which is what, $20 a month?

Adam Cooper: Yeah.

Jeff Robertson: To start. And I challenge them to start finding ways to use it because it’s the way that the world is going. Like, everybody’s going to be using it. And the phrase I heard a lot last week was, “If you’re not already in AI and using it, today’s the day to start because you’re already behind.” And you’re not going to get… it’s going to move away from you faster and faster and faster. You need to start finding the ways to use it.

Adam Cooper: And be creative. I think two things to say about that. One is the creativity piece. To the extent that I’ve played around and looked for ways for us to help ourselves and also leverage for our clients, the creativity and the imagination thing is the limiter. Which is to say, there is really no limit. How creative can you be?

Jeff Robertson: Yeah, you just have to open your mind and figure it out.

Adam Cooper: I mean, the way that you prompt it, I know, is a big thing. Like, it’s getting more and more intelligent. You can ask it longer and longer, feed it more information in the prompts. I was loading all sorts of stuff into it just to give it a background, and now it’s starting to think like we think. It’s almost like… Oh, something I came up with last week was, you know, training your AI is like onboarding a new employee. Teach it as much as you can about your business, about your projects, about your people, feed it job descriptions, feed it proposals, feed it drawings, feed it historical data.

Jeff Robertson: I think I must have read this or maybe I heard it in another podcast somewhere, but the concept of, you know, how could a project manager utilize this and, you know, “It’s just going to take our jobs” and this whole bit. Number one, I mean, we work in the built environment. Construction is the built environment. That’s not going anywhere.

Adam Cooper: Right.

Jeff Robertson: It’s… I think if you change your mindset to “it’s another tool in the tool belt.” It’s just like when we started, I mean, this is a long time ago, but at the beginning of our careers, a WYSIWYG scheduling concept was really new and revolutionary. We were scheduling by just plugging stuff into, you know, DOS. And then all of a sudden we could actually play with bars and Gantt charts and it’s right there on the screen and it blew everybody’s mind. Right now, that’s pretty darn standard. Uh, it just became another tool, right? The same, you know, any project management software we’re using was pretty revolutionary 25-30 years ago. Not anymore. It’s just a tool. So, I think if that creativity piece and just considering, “Well, it’s just a tool I could use to maybe help. Maybe it’s predictive, maybe it’s just helping me think a little bit better.”

Adam Cooper: Think about how creative people have gotten with Excel.

Jeff Robertson: Yeah.

Adam Cooper: Like, they use Excel for everything. It’s the most versatile program ever created, I think. Like, we use Excel for so much stuff as an industry. Think of AI like that. It’s just a tool like Excel and you can use it in so many different varieties, so many different ways. You’re only limited, as you said, by your creativity.

Jeff Robertson: I mean, think of it this way. You could, if you’ve… you’d have to string together a couple of things through some APIs, or maybe you’re doing a little bit of manual entry into an Excel spreadsheet that you really love that helps you figure out your WIP and your cash flow, and it’s just how you think. Well, you plug that in. You plug AI into that through either maybe through an API or you dump the whole thing in, whichever way works for you, and it’s spitting out dashboards for you.

Adam Cooper: Mhm.

Jeff Robertson: In a way that you’re like, “I couldn’t do that in Excel. I just know how to look at numbers.” So now I have some reporting that’s automated. Uh, maybe you could plug it in. It’s pushing numbers into your spreadsheet automatically. There’s all kinds of possibilities.

Adam Cooper: There’s a new thing in AI, or I shouldn’t say newer, called an “agent.” So, an agent is an AI that has a specific set of instructions to perform a specific task.³ It makes me think of Keanu Reeves in The Matrix and with the agents chasing him. Those were agents for different things, right? It was the Key Master; he was an agent that mastered the keys. And then there was Agent Smith and so on and so forth. But that’s what makes me think about it. They’re calling it “agentic AI.”

Jeff Robertson: Agentic AI. So that’s just an AI that you’ve basically taught to do specific tasks, right? You’ve taught it to do something very limited but to do it really well.

Adam Cooper: Well, the beauty of this is there’s, you know, a lot of those agents or some of the API plugins are low-code or no-code applications now. It’s dragging and dropping and giving it some basic mapping, or you could ask ChatGPT to write the code for you, and it will do that too. So, uh, you don’t need to go hire somebody to write that code necessarily. You can play around with it and figure it out yourself if you have that kind of…

Jeff Robertson: I’ve seen it generate really complicated math formulas, like statistician-type math.

Adam Cooper: Right.

Jeff Robertson: I’ve seen it write HTML code. I’ve seen it write a few other machine languages that I can’t remember or pronounce. JSONs and all sorts of stuff that I know. It can create images, graphs, charts, files, and all.

Adam Cooper: That’s true. It knows all those.

Jeff Robertson: Yes. So anyway, that was just some thoughts I had about AI, and I’m looking forward to continuing to explore that internally as well as with clients. How can we start leveraging more and more AI tools? Yeah. Not to reduce headcount, but to really maximize the efficiency and the capacity of headcount.

Adam Cooper: Yeah, totally agree. So, alright, thanks for listening to our Hot Takes podcast today. Uh, in case you missed it on our last episode, we’ve set up an email address: hottakes@ascentconsults.com. Send us your questions, send us your thoughts, send us your topics that you’d like us to cover, and we will pull them out of a hat and answer them on the spot on our Hot Takes podcast. Check us out on Apple Podcasts and Spotify. Subscribe to our YouTube channel. Send us your thoughts from the field. Thanks.

Host/Outro: And that’s all we have for this episode of Construction Hot Takes. Thanks for listening. And if you enjoyed the episode, don’t forget that you can follow us on Spotify or Apple Podcasts at Construction Hot Takes. And while you’re at it, please subscribe to our YouTube channel, too, at Ascent Consulting. If you want to drop a thumbs up or comments on your favorite video, that’ll go a long way in helping our channel. And lastly, if you want to schedule a call directly with Adam, Greg, or Jeff to talk about problems or challenges facing your own construction companies, you can schedule free consultation calls directly with them on our website, www.ascentconsults.com. Again, thanks for listening and we’ll catch you next time.