The sim-to-real gap is a commercial question

Closing the gap between simulation and a real site has a price at every installation, and that price sets the unit economics.

By Harinderpal Hanspal on March 2026. Updated October 2026

Engineering teams put the sim-to-real gap on the technical roadmap. The business question is what closing it costs at each site, because that number decides whether the product earns product margins.

Closing the sim-to-real gap site by site turns a product into a services business Two lanes of five deployments. When the per-site adaptation cost is confirmed and priced in, engineering is a thin slice of each site and product work dominates. When bespoke engineering is needed at every new site, it fills most of each deployment and the product slice shrinks. Verified Per-site cost confirmed in a representative environment and priced into the model Assumption Bespoke engineering at every new site: a services business with product-company pricing Site 1 Site 2 Site 3 Site 4 Site 5 Product work Engineering to close the gap at that site Illustrative, not to scale
Closing the sim-to-real gap site by site turns a product into a services business

The gap belongs to the site as much as to the model: lighting, surfaces and dust, equipment of mixed vintage, the layout, and the undocumented workarounds the night shift relies on. More lab work cannot settle it. It can only be measured by running the system in the buyer's environment, which makes it a validation question with a commercial answer.

If closing the gap needs bespoke engineering at every new site, the company is a services business with product-company pricing. Margins shrink with each deployment and timelines stretch. No single deal shows it, because each one closes with "some integration work". It appears in aggregate, once deployments have consumed the engineering roadmap.

Adoption raises the stakes. A Manufacturing Leadership Council survey in early 2025, cited by Deloitte, found 9% of responding manufacturers using physical AI and 22% planning to within two years. Each of those deployments will meet its own gap, during validation or during the contract.

"We are working through site variability" is an Assumption. A per-site adaptation cost confirmed in a representative environment and priced into the model is Verified.

Go deeper: Validating physical AI before the capital is irreversible treats the gap as the second of three dimensions, and how we validate physical AI shows how each is graded.