Ground-level air quality, climate conditions, and AI-powered analysis across all five KCCA divisions, live, open, and actionable.
What you get
Temperature, humidity, pressure, dew point and apparent temperature from Tomorrow.io primary weather feeds during this web pilot.
CO₂, PM2.5, PM10, NO₂, CO, NH₃ and VOC/gas resistance, cross-referenced with OpenAQ station observations and Open-Meteo modelled data.
Ask plain-English questions about current conditions. Llama 3.3 70B on Cloudflare Workers AI returns structured risk analysis, likely outcomes, and mitigations.
Canonical reference points for all five KCCA divisions: Central, Kawempe, Makindye, Nakawa, and Rubaga, switchable in one click.
The GSS desktop application connects BME680, soil probes, photoresistors, PMS5003, MQ-series gas sensors and more. Packets queue to flash storage during outages, so nothing is lost.
Zero cold starts. Durable Objects provide live WebSocket streaming, D1 SQLite stores persistent history, and Workers AI provides on-demand analysis at the edge.
Spatial coverage
Each division has a canonical collection point. External intelligence data is collected and merged per area. Switch between them live in the dashboard.
How it works
A fully integrated pipeline from hardware to browser.
ESP32 nodes read BME680 and connected probes. Packets are queued locally and posted to the Cloudflare Worker over HTTPS.
Durable Objects deduplicate and broadcast via WebSocket. D1 persists the full history for queries and analysis.
On demand, the worker fetches from Tomorrow.io, OpenAQ, Open-Meteo and WeatherAPI and merges into the sensor timeline.
Ask a question. The AI advisor analyses the merged snapshot and returns structured risk levels, outcomes, and mitigations.
Get started
Sign in with your Google account. No password or credit card is required during the pilot period.
Sign in with Google