From 3 months to 3 seconds: what a construction metaverse taught us
Years before 'industrial metaverse' was a category, we pitched a game that would let the world plan construction sites like Fortnite. The concept never shipped — but its ideas run through everything we build today.
In the early 2020s we designed a concept we called a construction metaverse: a free, game-engine-based simulation where anyone — a planner in London, a student in Kerala — could plan, test and optimise a real construction site the way gamers build worlds. Pre-planning that took three months in spreadsheets would take three weeks in a live twin, with any answer retrievable in three seconds. Play a game, disrupt an industry.
It was a grant pitch, and we'll say so plainly: it was never built. But the ideas in that document became the operating system for everything WINNIIO delivers now — and they're worth showing, because most of the industry is only arriving at them today.
First: the game engine as the data model. Construction drowns in file formats that can't scale; the pitch argued the twin should live inside the engine — one open environment covering design, review, simulation and operations, fed by the CAD models, spreadsheets and manuals a company already has. That is exactly the virtual-first pipeline we later delivered for a London heritage project, and it is how we build network and grid twins now.
Second: the Dark Panel Philosophy. Commercial airliners keep the cockpit dark — a light means something needs attention. A construction site, a telecom network, a power grid should read the same way: everything green says nothing; the one red icon tells you where to look, why it's red, and who to call. Status, in one glance, for every part in space and time — produced, in transit, delivered, installed.
Third: crowd-sourced optimisation. The pitch imagined gamers racing their own ghost — each project setting the baseline the next one beats, waste heat-mapped like a digital Kaizen. Swap 'gamers' for 'agents' and that is precisely where autonomous AI has taken us: fleets of software workers running simulations against reality's baseline, proposing the next improvement.
The honest lesson is about timing, not technology. Every capability in that pitch exists today — photoreal capture, real-time engines, agent swarms. What we underestimated was that the constraint is never the tool; it's how processes, people and incentives are arranged around it. Which is why we now start every engagement with the as-is — because when you truly understand what exists, you can re-arrange instead of re-invent.
One call is enough to know if we're a fit.