Hyperlocal environmental intelligence · JKUAT / Juja, Kenya

From environmental data to early action.

AFYA MAZINGIRA turns live Conduit observations into environmental states, short-horizon forecasts, activity-aware risk and a recommended best time to act, before conditions peak.

Current station situation

Connecting to the JKUAT station…
observations analysed
45,043
observations analysed
environmental states
4
environmental states
forecast horizons
+1/+3/+6/+9h
forecast horizons
languages
EN · SW
languages

The interpretation gap

Sensors tell you what. They don't tell you what to do.

A person does not need temperature, humidity and wind. They need to know whether conditions are becoming more demanding, and what to change.

Illustrative values

A conventional dashboard

Temperature29°C
Humidity48%
Wind1.5 m/s
Pressure850 hPa
Rain0 mm

The user still has to work out what these numbers mean.

Example output

AFYA MAZINGIRA

Rapid Warming

Transitioning toward the afternoon's peak heat.

Expected peak
12:00–14:00
Best window
16:30–17:30
  • Air temperature rising
  • Solar radiation remains high
  • Ventilation relatively weak

Situation → meaning → action → explanation.

Deterministic pipeline

Every recommendation is computed, not guessed

  1. Conduit data

  2. Quality control

  3. Feature engineering

  4. Climate Reflex state

  5. Forecast +1/+3/+6/+9h

  6. Risk interpretation

  7. Best-Time engine

  8. Grounded explanation

The AI communication layer sits strictly downstream of this quality-checked chain. It explains results, it never produces them.

Climate Reflex

The environment has four states

Discovered from real station behaviour, not imposed by hand.

Typical overnight

Cool & Humid Stable

Lower temperature, high humidity, weak wind and low radiation.

Morning transition

Rapid Warming

Temperature rising quickly while humidity falls and radiation climbs.

Daytime maximum

Hot / Peak Heat

Peak temperature and radiation with stronger wind and elevated exposure.

Evening recovery

Cooling / Recovery

Temperature falling, humidity recovering and radiation declining.

Product

Built for decisions

Data provenance

Every number has a source

Measured, predicted, satellite-derived and regional-model data are always visually distinct, never blurred together.

GROUND MEASUREMENT

Conduit

~15 min · JKUAT/Juja station

Local temperature, humidity, pressure, wind, light and rain. WBGT is computed from its wet bulb and air temperature.

REGIONAL MODEL

ERA5

~28 km · hourly · about 5 days behind

The regional background for the local-versus-regional comparison, and the 7-day and 30-day rainfall totals.

REGIONAL FORECAST

Open-Meteo

hourly · Risk Map counties

Temperature, humidity and rain for each county on the Risk Map, and the outlook used to plan tomorrow.

SATELLITE-DERIVED

Sentinel-2

10 m vegetation

Vegetation (NDVI) near each county centre, when the Copernicus service answers. Blank when it does not.

Scientific honesty

What AFYA MAZINGIRA does not claim

A smaller number of defensible capabilities beats a longer list of impressive-sounding claims.

  • A medical diagnosis system
  • A malaria, cholera or asthma predictor
  • A street-level flood predictor
  • One station speaking for all of Juja
  • A generative-AI forecasting engine
  • Universal “safe” or “dangerous” thresholds

See it working, right now

Live station data, deterministic forecasts and a recommended window, in English or Kiswahili.

Open AFYA MAZINGIRA