Daily field briefing
POLR AIPOLR AI

AI + AEC Today

Daily AI intelligence for the built world

Weekly digest

Weekly AI in AEC stories

← Archive

Edition · Tuesday, August 11, 2026

AI & Construction

Weekly Pulse · Industry actionability

The field is piloting, not just experimenting.

70%
ExperimentalPilotingField-ready
  • Reality capture plus AI catches build errors before the field does
  • AI now lays out full building systems in a day, not months
  • Brick-laying robots move from Europe toward US job sites
  • Big tech agrees on one standard so AI tools work together

Practical Use of the Day

Catch scope gaps before you award a sub

It is easy to miss a hole in a sub's bid until the job is underway. Let AI do a first check. Paste in your scope and the sub's proposal. Ask the AI to list what the sub left out or marked as not included. Read its list, then ask the sub about each gap before you sign. You still make the call, but you catch the holes early.

Try this prompt

I am a general contractor reviewing a subcontractor's bid. Here is my scope of work: [paste scope]. Here is the sub's proposal: [paste proposal]. Compare them. List every item in my scope that the sub did not include or marked as excluded. List anything the sub added that I did not ask for. Point out any vague wording that could lead to a change order later. Keep it short and in plain language.

Tool — Claude (or any capable LLM)

Stories

4 curated
ConTechHigh impact

Buildots adds laser scanning to catch build errors early

Buildots makes an AI that tracks how a job is going. Now it can use laser scans too. It teamed up with NavVis, a company that maps sites with lasers. The AI lines up the scan, the site photos, and the 3D model in one place. It then finds spots where the real work does not match the plan. Intel is already using it on a chip plant build.

Why it matters to GCs

Small build errors get costly when you find them late. This tool spots them early, while they are still cheap to fix. Catching a wall or pipe in the wrong spot before the next trade shows up saves rework and time.

AEC Magazine·Aug 11·Read source →

LinkedIn
ConTechHigh impact

Endra's AI designs a building's systems in a day

Endra builds AI for MEP work, which is the mechanical, electrical, and plumbing systems in a building. Its AI reads your building model and lays out the systems for you. It routes the pipes and wires and makes the drawings. Endra says a job that took about two months can now take less than a day. The company just raised $50 million and opened offices in the US and UK.

Why it matters to GCs

MEP design is slow, and a late set of drawings holds up the whole job. Faster design means you can start work sooner and bid more jobs. You still check the AI's plan, but you skip weeks of drafting.

AEC Magazine·Aug 3·Read source →

LinkedIn
ConTechModerate

Robots that lay brick are coming to US job sites

Monumental makes small robots that lay brick and spread mortar on real job sites. They use cameras and AI to read the 3D plans and place each brick with care. The robots have built walls for more than 100 homes in Europe, plus schools and hotels. Now the company raised $32 million to bring them to the US. People still do the tricky parts while the robots do the long, straight runs.

Why it matters to GCs

Skilled masons are hard to find, and that slows jobs down. Robots that handle the repeat work can keep a wall moving. They will not replace your crew, but they may help you hit a tight schedule when hands are short.

The Robot Report·Jul 17·Read source →

LinkedIn
AI FrontierModerate

Five tech giants agree on one way to plug in AI tools

Five big tech firms just agreed on one shared way to build AI add-ons. They are Amazon, Microsoft, OpenAI, Vercel, and Cursor. The idea is called Agent Plugins. A maker can build a tool once, and it works across many AI apps. Before this, they had to build a new version for each one.

Why it matters to GCs

The AI tools you use should soon work better together. A plug-in built for one app may work in the next. That means less lock-in and more choice when you pick tools for your team. It is early, and the safety questions are not solved yet.

The Next Web·Aug 6·Read source →

LinkedIn