Personal air-pollution exposure journeys rather than station readings

AIRAWARELABSbuyer room · privateManage buyer room

Personal air-pollution exposure journeys rather than station readings — the company's 2025 Annual Report states 250,000 real-world journeys, each a GPS track with per-segment pollution, pollutant levels and an inhaled-dose estimate, joined to activity type (run, ride, walk, commute) and transport mode, and synced from Strava, Garmin and Apple Health so that heart-rate and HRV-recovery physiology sits alongside the exposure record. Layered underneath is a 50-metre street-level pollution model updated hourly, built from NASA GEOS-CF, CAMS, emissions and land-use data, traffic, weather, 3D city models, UK and European regulatory networks and OpenAQ outside Europe, and validated against 145 Breathe London sensors with a holdout showing up to 35% accuracy improvement over a raw global baseline on R-squared and RMSE. Real-time indoor measurements arrive through MedicAir devices, making indoor and outdoor exposure for the same person at the same timestamp available. The non-obvious byproducts are: (a) dose-weighted exposure series in which route, mode and timing are the treatment variables and inhaled dose is the outcome; (b) a natural experiment of the same street coordinate experienced from a car, a bike and on foot, giving mode-attributed exposure contrasts no fixed network can compute; (c) exposure-physiology pairs linking pollution to HRV recovery, which the company reports as correlated; (d) revealed-preference labels from personalised alerts that measurably caused route and routine changes; and (e) a route-optimisation objective — the cleaner-route engine scores alternative paths against the direct route, which is preference data over named street segments.