One platform. One city model. Unlimited decisions. Monitor infrastructure, predict failures, optimise investment and give governments a defensible answer before the money is spent.
Every module reads one shared twin — the same buildings, the same roads, the same people. Change something in traffic and the water model sees it.
Every figure above is computed from the shipped data bundle, not estimated.
Buildings, roads and amenities from OpenStreetMap; building heights from
GHS-BUILT-H; population from the State Committee of Statistics of
Uzbekistan (resident population by district, 1 January 2025), cross-checked
against WorldPop.
Each module is a lens on the same twin. They share the demand model, the road network, the dependency graph and the timeline — so they can argue with each other.
Agent-based traffic with reinforcement-learned signal control, benchmarked against fixed-time and actuated baselines.
Demand derived from building use, floor area and SNiP norms — then routed through a network on real streets.
Recorded live against the Tashkent model — vehicles, signals, city gates and the AI controller reacting to real queues.
One temperature change, seven systems, a single simulation. This is the part no dashboard can do — because it requires knowing what depends on what.
Heatwave hits 46 °C
Electricity demand spikes
Substation overloads
Pump stations stop
Water pressure drops
Hydrants go dry
Fire response delayed
Investment AI returns the cheapest intervention
Hiring: GIS engineer · urban planning advisor.
Type: Space Grotesk headings · Inter body ·
JetBrains Mono numerals. The mark is a “U” whose crossbar carries a
skyline and a live network node — urban, and thinking.
UrbanMind