Living Digital Twin of Toronto
What if you could test-drive the future of your city before you vote?
Every election, citizens make critical decisions based on static PDFs and abstract promises. Our team at Flybits asked a different question: what if we agentified an entire city in near-real-time?
With Toronto’s mayoral election approaching, our team built a living Digital Twin of Toronto — 92,704 building footprints, 20,000 vehicles routing at once and a working flight simulator, running at full frame rate in an ordinary browser tab. It is powered by the new Flybits Spatial Agentic Intelligence, Perspective-Aware AI and AI Chronicles.
Every city publishes its own blueprint. Almost nobody can read it.
Every building footprint, road centreline, park, pier and runway in Toronto is already open public data, as it is in most cities. The information is not withheld. It is simply unreadable — shipped as coordinate arrays that answer none of the questions anyone actually has about the place they live.
Turning raw data into context people can act on is the problem Flybits exists to solve. So our team took the rawest, least-usable public dataset we could find and asked how fast we could turn it into something a person understands instantly — not a report about the city, but the city itself.
The answer was weeks, not quarters — and the result runs on a phone.
“The fastest way to prove a capability is to build it. So we built the city.”
Every system is a live agent
Every component of the city operates as a live, synchronized AI agent.
Traffic
Flights
Transit
Crime dynamics
Zoning
Emergency dispatch
What is actually running
Every number here is measured from the running application and the data it ships with, not estimated.
What it demonstrates
A spatial data pipeline
Raw open geographic data is ingested, cleaned, clipped at tile boundaries and reconciled into one coherent model — coastlines rebuilt from water rings, roads stitched across tile seams, landmarks recognised and handled specially.
Real-time simulation at scale
Two independent systems run live on that geometry: traffic, with up to 20,000 vehicles pathfinding across the real street graph on a background worker thread, and aviation, with twelve aircraft taking off and landing on Billy Bishop’s actual runways under runway contention rules.
Zero-footprint delivery
It ships as a static site and installs as a standalone app. The map, the traffic simulation and the rendering all run on the device, with a single open-source rendering library as the only runtime dependency — no bundler, no build pipeline, no vendor in the critical path.
What powers it
Spatial Agentic Intelligence
New from Flybits: agents that reason about place — where they are, what surrounds them and what is changing around them — grounded in real geometry rather than text descriptions of it.
Perspective-Aware AI
Our research into AI that understands a situation from a specific point of view, rather than averaging everyone into one answer.
AI Chronicles
User-owned AI graphs that let one person see a situation through another’s lens — here, a city seen through the eyes of the people who live in it.
Built in the Living Lab
Our Living Lab exists to be a flight simulator for enterprise decisions — a place where real ideas get built, stressed and measured before anyone bets on them. This build is that principle turned on a city instead of a balance sheet, and it was produced by the same squads, on the same cadence, that enterprise engagements run on.
It is also the substrate the next generation of our spatial work needs. Agents that reason about place first have to be given a place to reason about. This is that place, and we own every layer of it.
Where it goes next
Direction of travel, not shipped features.
Live municipal feeds
Swapping simulated movement for real transit positions and road conditions — turning a model of the city into a mirror of it.
Scenario comparison
Holding two versions of a neighbourhood side by side, so a proposed change can be walked through before it is approved.
About Flybits
Flybits delivers contextual AI technology, expert teams, and our Living Lab model to help enterprises turn knowledge into working capability.
Flybits Labs, the company’s applied R&D division, sits at the intersection of financial services innovation and cutting-edge technology. Partnering with leading institutions, global organizations and top academic institutions, Flybits Labs advances next-generation AI applications, sophisticated data models and intuitive interfaces that accelerate the journey from concept to real-world deployment.
Map data © OpenStreetMap contributors. Contact: marketing@flybits.com