Overview
Rapid urbanization is degrading air quality across fast-growing cities, and the planning response is usually intuition: plant trees, somewhere, and hope. AirQualify set out to replace that guesswork with evidence — a quantitative link between a city's blue-green cover (vegetation and water bodies) and the air its residents actually breathe.
AwaitSol built the platform end to end: the computer-vision pipeline that measures urban land cover, the statistical analysis connecting it to air quality, and the web application that puts the findings in planners' hands.
The Challenge
The fundamental question — how much green and blue space does each district have, and how strongly does it correlate with the Air Quality Index — had no off-the-shelf answer. Land-cover data for these cities was either outdated, too coarse, or nonexistent, and AQI readings come from a sparse network of monitoring stations that don't align neatly with district boundaries.
Beyond measurement, the output had to be decision-grade. A correlation coefficient doesn't help a city planner; what they need is 'this district needs this much additional green cover, in these locations, to move its AQI by this much.' Translating model output into that language was as much of the challenge as the model itself.
Our Approach
We built a segmentation model that classifies urban land cover from imagery, trained and validated to reliably separate vegetation and water bodies from built-up area. Running it across the full urban landscape produced district-by-district blue-green coverage figures — a dataset that simply hadn't existed before.
With coverage quantified, we modeled its relationship to AQI readings across districts and seasons, then inverted the question: given the correlation, how much additional green cover would each under-served district need to reach target air-quality levels, and which specific zones are viable candidates for afforestation?
Everything ships through an interactive web application designed for non-technical users — planners explore districts on a map, compare coverage against air quality, and export the afforestation recommendations that the analysis supports.
What We Built
AirQualify turns satellite-scale imagery into a planning instrument: a city-wide inventory of blue-green space, a quantified link between that inventory and air quality, and ranked, location-specific recommendations for where new plantation would deliver the most measurable impact.
- Land-cover segmentation model for blue-green area detection
- District-level blue-green coverage inventory across the city
- AQI correlation analysis across districts and seasons
- Afforestation recommendations with required green-cover targets
- Interactive web application for urban planners
Key Results
Urban greening decisions that previously ran on intuition now run on evidence. Planners can defend afforestation budgets with quantified AQI impact, target the districts where green cover is both lowest and most beneficial, and measure progress against a baseline that the platform established.
Technology
Have a similar challenge?
Let's talk about what we can build together.
