In partnership with
Builders have never rated site technology more highly. Yet fewer of them use it. In the latest BuiltWorlds benchmarking, positive evaluations of innovative equipment rose above 95%. Yet active robotics use fell to 46%. The tools work. Adoption is going backwards.
Reality capture means scanning what was actually built, then checking it against the design. It catches errors while they are still cheap to fix. The capability now comes in several forms, and teams from Oslo to Tokyo run it in production today. Most operators still meet it 18 months late, after the deviations have already become rework.
Scanner, camera, or robot: no vendor owns reality capture
Reality capture is a category, not a product. A site can be captured three ways, and each compares the as-built result to the building model.
A laser scanner gives millimetre accuracy. A 360-degree camera gives speed and reach. An autonomous robot removes the human from the walk entirely. All three feed the same goal: find where reality and the design diverge, while the fix is still cheap.
You no longer need a laser scanner. A 360 camera will do
The hardware barrier is falling. Cupix turns 360-degree video into a navigable 3D digital twin, with no laser scanner required. A worker records a site walk on a helmet camera. The platform builds the twin and lines it up against the BIM model.
Others take the same photo-first path. OpenSpace and Buildots map 360 footage to the plan automatically. Imerso compares a laser scan to the model and flags deviations within hours. The method varies. The principle does not.
Robots now scan the site themselves, and investors put £298m behind it
Capture is going hands-free. Field AI builds autonomy software so robots navigate a changing site with no GPS, no map, no set path. The robot generates 360-degree 3D scans, tracks progress daily, and compares the build to the BIM model.
The market is betting on it. Field AI raised £298m ($405m, €345m) and reached a £1.5bn ($2bn, €1.7bn) valuation. Deployments span the United States, Japan, and Europe. Backers include the venture arms of NVIDIA, Intel, and Samsung.
Iowa built its bridge twice and caught the flaw before cutting steel
The payoff shows up early. Iowa is building the Black Hawk Bridge twice, first as a 3D model, then in steel. Reviewing the model, fabricators found truss access holes too small for tools to pass. They fixed the specification in days, before fabrication began.
The digital-first approach saved an estimated £2.4m ($3.2m, €2.8m) and six weeks. In the old workflow, that flaw becomes field rework, found only after the steel is up.
Rework eats 5 to 15% of every build. Reality capture aims straight at it
The prize is large. Industry studies put rework at 5 to 15% of total construction cost. Much of it traces to bad data: crews build from drawings that no longer match the site.
Reality capture closes that gap. On one Danish hospital, early deviation alerts saved over £4.3m ($5.9m, €5m). Britain's HS2 feeds its construction data into a virtual 3D replica of the railway for the same reason.
On your projects, how does reality capture compare to the BIM model at handover?
The data outlives the vendor. 3D Tiles has been open since 2019
Open standards make the data portable across all these tools. 3D Tiles, an Open Geospatial Consortium standard since 2019, streams massive 3D models to any device. Bentley acquired Cesium, the standard's originator, in September 2024 and merged it into its iTwin platform.
The same logic reaches process plants. i3D AI converts raw point clouds into engineering-ready solid models for oil, gas, chemical, and power facilities. Open formats mean the scan you take today still opens in a decade.
Builders love it more, yet use it less
Here is the disconnect. The tools are production-ready, the money is flowing, and the standards are open. The Facility Lifecycle 3D Model Specification, or FL3DMS, is standardising the 3D facility model. It covers laser scans and photogrammetry, the very methods these tools use. The biggest operators go further. In an ARC Advisory Group working group, bp, Chevron, Shell, and ExxonMobil set guardrails for an open digital twin. The rule is simple: separate the data from the software, and keep it interoperable across the lifecycle. Yet adoption lags, usually until after a painful handover.
Nuclear megaprojects show the better path. During Barakah's construction in the UAE, laser scanning confirmed alignments and tolerances as the plant went up. Most projects do not scan until the end, to document what already went wrong.
Three habits bury reality capture in the survey folder
Three habits delay adoption. First, capture is filed under survey, not management, so leaders never see its value. Second, the scan is treated as a record, not a control. It proves the past instead of steering the present.
Third, the budget sits with construction, but the benefit lands in operations. The split hides the return. Engineers and managers, not the scanner, decide when to look.
Scan early, or read the post-mortem
Reality capture is not a future capability. It ships today, by scanner, by camera, and by robot, across hospitals, bridges, and plants. The operators who win capture early and often, and act on the deviations while they are cheap to fix.
The rest meet the same tools 18 months later and read them as a post-mortem. The difference is not the technology. It is when you choose to look.
Next week: ISO 18136 just became the passport for AI in industrial facilities.
The ones showing up in LLMs convert 3× better than Google
They optimized for LLMs, not just Google.
FAQs. Comparison pages. Transparent pricing. LinkedIn presence. These aren't vanity plays. They're what gets you cited in ChatGPT, Gemini, and Claude when your buyers are researching, your investors are looking, and your future hires are deciding where to work.
Download the free AEO Playbook for Startups from HubSpot and get the exact checklist. Five minutes to read.