(This column originally appeared in Forbes)
In the construction industry, AI use cases are beginning to grow.
Construction Industries Slowly Adapting AI
At my clients’ jobsites, I’m seeing a number of surveillance drones that are keeping track of safety, job progress and weather conditions as well as more foreman equipped with augmented reality headsets to assess job status, identify open areas and point out potential safety concerns.
Robots won’t be replacing plumbers and pipefitters anytime soon, but robot dogs are already being used to send back data from hazardous areas or places considered too risky to send a human. 3D printers, autonomous material movers, stand-alone drilling and mixing machines and IoT-enabled protective clothing are now contributing to more productivity on the job.
What about the back office? Here I’m seeing new features in project management, estimating and Building Information Management (BIM) software systems that are leveraging AI to scan, capture and process data faster and more accurately and on a larger scale than a single person.
How Trimble Plays The Role Of Connector
This is all about better ways to collect and then use data. But who sits in the middle? As all this data is accumulated, how do you connect one system to another — and then analyze the information? One company, Trimble, is doing this part.
Trimble is a Colorado-based technology company that provides software and hardware to help their clients — many who are in the construction industry — plan, design, build, measure, and manage projects that connects what happens in the office (design, estimating, project management) with what happens in the field (surveying, machine control, scanning, site data capture, and 3D takeoff/data prep). Trimble says their technology streamlines the entire project lifecycle — from estimating and design to project management, field operations, and financials and use AI tools to unify data, improve collaboration, reduce errors and waste, and give real-time visibility into costs and progress.
Trimble, like others in the construction industry, is heavily leveraging AI to provide a better and more productive experience for its customers. In fact they’ve been doing so well before AI took off in the public narrative. According to Mark Schwartz, a Senior Vice President at the company., the company had AI invoice recognition and routing in their ERP system “before ChatGPT was a thing.”
“Old-school AI-machine learning and photogrammetry has been in our products for years,” said Schwartz.
For companies like Trimble, AI readiness is less about hype and more about data connection and workflow integration. which unintentionally prepared it for today’s AI moment. Schwartz says that his company has had a long time strategy to enable better connection and scaling of different software systems which set them up to use generative AI faster.
Getting Small Businesses On Board
But when will small and mid-sized businesses (SMB) bite? Sure, many are dabbling with generative AI chatbots. But the day will soon come when they will move more towards back office systems to improve their operations. For Schwartz there’s little question that the opportunity is there.
“We heavily serve the small and mid-market…because it’s so underpenetrated with technology,” he said. “Once they start using one or two things, it’s like the enlightenment comes.”
Two other factors will come into play: tech-enabled bigger companies that will subcontract down to smaller organizations and pull them into their tech enabled ecosystem, and a growing number of younger people trained on newer technology at an earlier age. As this plays out more SMBs will find themselves being drawn faster into using AI internally. But companies like Trimble know that this will take time because SMBs have little room for error.
“For SMBs, they want the Google Workspace version,” Schwartz says. “They’re like: give me the five tools that just work.”
And many SMBs remain dubious, particularly because of the less-than-reliable AI results many have seen in real life. Even Schwartz admits that while today’s AI can be “excellent for productivity” it’s also dangerous when deterministic accuracy is required.
“I don’t think any of us want to go to the 52nd floor of a high-rise designed entirely by an AI agent,” he said.
So when will there be a tipping point for SMBs? For Trimble and Schwarz, it’s all about patience. He believes it will be as trust in AI is built gradually through narrow, high-value workflows. But adoption doesn’t have to be aggressive.
“There’s plenty of productivity to be had right now without forcing the issue,” he said. “It’s important for SMBs to use it in the most relevant workflows and that will build trust over time. People will realize that for many tasks, the 90 to 95% answer is good enough when it can be done in seconds versus days.”
