Your Field Problems Start in Design: What the Suffolk and MIT AI Report Means for Contractors | Ep. 402

Your Field Problems Start in Design: What the Suffolk and MIT AI Report Means for Contractors

Read the report: Construction in the Age of AI, from Suffolk and MIT

https://suffolk.com/wp-content/uploads/2026/09/Suffolk_MIT_AI_Whitepaper_FINAL.pdf

Some of the problems your superintendent fights today were created during design, months before his crew showed up. In this solo episode, Eric Anderton reads the new Suffolk and MIT white paper, Construction in the Age of AI, with one reader in mind. You run a $25 to $500 million company. You’re interested in AI. You want to know what’s usable right now.

Suffolk has resources most contractors don’t. So Eric skips the hype and goes to the applications a contractor your size can test.

Listen or Watch the Episode

Three Kids, Three Different Messages About AI

Eric has three kids in college. They’re hearing three different things about AI.

One son is interning as a project engineer with a GC. On his own, he built an AI tool to help him draft RFIs and submittals. Nobody asked him to. He saw something he could make easier and figured out how.

Another son works as a laborer while he finishes a construction management degree. His professors encourage him to use AI on construction coursework. Eric’s daughter is studying something else, and she’s had to sign a pledge in some classes promising not to use it.

That’s the spread inside one family. Meanwhile, one of them already has a tool he uses at work.

The 17 to 20% Cost Savings Number and Where It Came From

The numbers in the report are big. Suffolk and MIT estimate 17 to 20% in potential cost savings and 22 to 25% on schedule.

Eric wants you to know how they got there before you expect it on your next job. Suffolk took a completed residential project in San Francisco and modeled what might have happened with AI and better coordination.

That’s one project, and it’s a model. The paper says more research is needed. Eric reads it as a call for contractors to test these tools on real work now.

AI in Design: Catching Clashes and Constructability Problems Early

The design section is where Eric spends the most time. The paper describes AI helping teams generate and compare design options, including structural concepts, and weigh them against cost, code and schedule. It also covers reviewing drawings and models for omissions, clashes and constructability problems.

You already know finding those early is worth money. What AI adds is the ability to look at more options and more information while there’s still time to change the design.

Eric kept thinking about what that means for the field.

“By the time your foreman gets the drawings, some of the decisions that will make his job difficult have already been made.”

And the foreman is left working around them.

“Before you criticize the field for poor production, you gotta look at what the field received in the first place.”

Connecting BIM Models to the Construction Schedule

The paper goes further. It describes structured building information models feeding scheduling tools, so less manual work sits between what’s designed and how it gets built.

A design change hits quantities. Then it hits purchasing and installation. If those effects reach the schedule sooner, your team has more time to deal with them.

You still need the trades in the room while their input can change the design. The paper makes that point too. Software can sort the information. The people who install the work need to be part of the discussion.

AI Scheduling: Sequences, Delays and the Three-Week Look-Ahead

Two scheduling applications will look familiar. One tests different construction sequences and combinations of crews and equipment. The other uses historical schedules to spot where delays are more likely.

In the paper’s expert rankings, scheduling came out first once people’s votes for their own specialty were excluded.

One example caught Eric’s attention. A contractor takes hundreds of schedule activities and uses a model to put together a three-week look-ahead, with a list of risks in a format the superintendent can use.

Your superintendent still reviews it with the trades and decides if it’s realistic. He knows things about the job that never made it into the schedule. If AI saves him some of the time it takes to pull the look-ahead together, it’s worth testing.

Linking Procurement and Lead Times to the Schedule

The paper also covers site imagery and computer vision to track installed work against the plan. That could help your team see where progress stands sooner.

Then there’s procurement. You already have people tracking long-lead items. The paper describes connecting the procurement log and current lead times to the schedule, so a delivery change flags the activities it puts at risk.

Your team still decides how to respond. It could spend less time tracing the consequences.

There’s a catch. If the trades buy the materials, you need their commitments and delivery dates.

“The tool can’t make those connections reliably if nobody has supplied the information.”

That runs through the whole paper. These tools need current, accurate information, and getting your data into that shape takes work.

AI Risks: Wrong Answers, Confidential Information and Accountability

The paper has sections on permitting, prefabrication and trade coordination. Depending on your business, those may matter more to you.

Eric also recommends the section on risks. The paper covers inaccurate answers, confidential information and accountability.

Someone has to check what the tool produces, and your people need to know what they’re allowed to share with it. The decision still belongs to a person. When you judge how much time an AI tool saves, count the hours spent preparing the inputs and checking the output.

“My son still reads every RFI before it leaves his desk.”

He’s learning what the technology can do while he builds his construction judgment.

How to Read the Suffolk and MIT Report

You probably have someone in your company who understands the technology. You probably have someone else who can spot a bad assumption about the work from across the room. Eric’s homework is to have those two read a section of this paper together.

Read it with one of your jobs in mind. Where could better information or an earlier warning have changed a decision you made?

Subscribe to Construction Genius wherever you listen, and leave a rating and review. If this episode made you think of another contractor or someone on your team, forward it to them. More at constructiongenius.com.

Subscribe & Review

If this episode helped you, the best thing you can do is leave a rating and review wherever you listen. It helps other contractors find the show.

Construction Genius is hosted by Eric Anderton.

👉 More from Construction Genius