Episode Content
Rework eats 5 to 10 percent of every construction budget. On a $100M project, that's $5M to $10M of pure waste, and most of it isn't a modelling failure, it's a translation failure.
Caner Dolas, CEO and co-founder of GAMMA AR, has spent the last few years building the layer that lives between the BIM model and the foreman holding a hammer. The conversation cuts past the usual AR demo-reel pitch and lands somewhere more useful: a working theory of why digital construction has matured everywhere except the final few metres of every project, and what the firms moving the rework number are actually doing differently.
The Real Bottleneck Isn't the Model
BIM coordination has matured. Models are getting cleaner. Clash detection is routine.
The work still falls apart somewhere between the model and the crew building from it. Even on projects with mature BIM standards, information arrives on site as PDFs, spreadsheets, and phone calls.
Caner's reframe is the one to steal: BIM is a language, not a model. The model is vocabulary. The value is in how clearly that language reaches the people doing the work.
Here's the failure mode. Someone on site calls the office: "top left corner of room C02, change needed."Which top left corner. From whose perspective. At what elevation.
A model with stable component IDs collapses that ambiguity. Issues, RFIs, and progress reports attach to the same object the design team references, in a form a foreman can act on. The model stops being a deliverable and starts being the shared vocabulary of the project.
That's where the 50 percent rework reductions Caner's customers are reporting on complex projects actually come from. Not from the AR overlay alone. From the component-based language underneath it.
The Pull-Not-Push Pattern of Adoption
Construction is famously slow to adopt new software. The default vendor playbook, top-down sales into a procurement department, almost always stalls. The people signing the contract aren't the people who would use the tool.
The pattern that actually works is the opposite. Organisations come to Gamma, usually because one internal champion already ran a pilot and became an advocate.
The champion is almost always a "digital native", often a recent graduate, who picks up the tool quickly and pulls the rest of the project crew along. Train the trainers, then watch the propagation happen inside the company itself.
Here's the part most vendors miss. In conservative industries, the unit of adoption is not the organisation. It's the individual willing to put their reputation behind the new thing.
That has implications for buyers too. If you're trying to roll out new site tech and you don't have a champion on the project yet, the rollout will fail.
Find the champion first. The procurement paperwork comes second.
Open Integration Is a Market-Entry Requirement, Not a Values Choice
Most construction software still tries to build a walled garden. Proprietary file formats, captive workflows, closed platforms.
Gamma went the other way and integrates with Autodesk, Procore, BIM Collab, and the rest.
The reasoning isn't idealism. Projects already involve many systems, and refusing to interoperate just removes the vendor from the project.
This is worth absorbing as a buyer. If a platform cannot connect to the systems your team already uses, the integration burden falls on the project. And the project is the wrong place to absorb that cost.
Openness should be a hard line in any procurement scorecard, not a nice-to-have feature comparison row.
The Strongest Objection Isn't About the Tech
Caner has heard every objection to AR on site. The strongest one isn't "the technology isn't ready." It's "I don't think my project team can implement it."
That worry is real, not resistance dressed up as a concern. Project crews are under stress, they're busy, and not everyone on a site is tech-savvy.
The fix isn't more demos. It's one or two digital natives embedded in the project. They learn the tool, they show the older site staff, and the dynamic shifts.
The same pattern played out with BIM itself a decade ago. The teams that adopted BIM fastest weren't the ones with the best procurement processes. They were the ones with a junior on the team who already knew Revit.
If you're a digital lead, the most valuable hires you can make right now aren't more BIM coordinators. They're the people you can put on site to act as bridges.
Key Takeaways
- Rework is the most measurable failure in construction. 5 to 10 percent of project cost. If you don't track yours, you don't know your headroom.
- Treat BIM as a language, not a deliverable. The value isn't in the model itself, it's in how cleanly the model reaches the people who build from it.
- Use component IDs from the BIM model as the shared reference for issues, RFIs, and progress reporting.
- In conservative industries, adoption only scales through internal champions. Find your digital native on the project and let them lead the rollout.
- Closed platforms become your problem, not the vendor's. Open integration should be a hard line in any procurement scorecard.
- The strongest objection to new site tech is rarely about the tech. It's about whether the project team can absorb it. Solve for capacity, not features.
- Hardware constraints often produce leaner, more durable software. Don't wait for the next device generation to start piloting.