IoT Weather Station with Mobile Dashboard

Weather data is the classic first IoT project, and for a reason: it exercises the entire pipeline in one build — sensing, calibration, wireless telemetry, storage, visualization and alerting. This weather station does it properly. A DHT22 measures temperature and relative humidity, a BMP280 adds barometric pressure, a tipping-bucket rain gauge counts rainfall in 0.3 mm tips, and a cup anemometer with a reed switch reports wind speed. An ESP32 samples all four, converts raw counts to physical units, and publishes to a cloud dashboard over Wi-Fi using MQTT, where live values, 24-hour graphs, daily rainfall totals and heat-index calculations are displayed. Threshold alerts (heat, storm-level rain rate, high wind) fire push notifications. The project is honest about sensor accuracy: datasheet values are stated, and the calibration procedure against a reference thermometer is part of the deliverable.

1. The ESP32 samples the DHT22, BMP280, rain-gauge tip counter and anemometer pulse counter on a fixed schedule.
2. Raw counts are converted to physical units: tips × 0.3 mm for rainfall, pulses-per-interval for wind speed, with debouncing on the reed contacts.
3. Readings are packed into a JSON payload and published over Wi-Fi via MQTT to the cloud broker.
4. The dashboard backend stores the series and renders live values, 24-hour graphs, daily rainfall totals and heat-index values.
5. When a reading crosses a configured threshold (heat, rain rate, gust), the alert service sends a push notification to the phone.
6. A solar + Li-ion option with ESP32 deep-sleep between readings supports unattended outdoor deployment; the accuracy note in the report compares each sensor's datasheet tolerance with the buyer's calibration check.

Tech stack:

• ESP32 development board
• DHT22 temperature/humidity sensor
• BMP280 barometric pressure sensor
• Tipping-bucket rain gauge (reed-switch)
• Cup anemometer (reed-switch pulse)
• MQTT broker + cloud dashboard (web)
• Arduino IDE (C/C++ firmware)

• [Multi-sensor node] DHT22 (temperature/humidity), BMP280 (pressure), tipping-bucket rain gauge and reed-switch anemometer — four real sensing channels, each converted to physical units in firmware.
• [Live web dashboard] Current values, 24-hour trend graphs, daily rainfall total and computed heat index, served as a working dashboard alongside the hardware.
• [MQTT telemetry] ESP32 publishes JSON readings over Wi-Fi via MQTT with a configurable reporting interval (design target: 60 s) and reconnect handling.
• [Threshold alerts] Heat-index, rain-rate and wind-gust thresholds trigger mobile push alerts; thresholds are configurable from the dashboard.
• [Tipping-bucket rain gauge] Self-built or sourced tipping bucket with 0.3 mm-per-tip resolution; firmware debounces the reed contact and accumulates daily totals.
• [Heat-index computation] Apparent temperature is computed from temperature and humidity using the standard heat-index formula — a real derived quantity, not a raw sensor dump.
• [Solar/battery power option] The node runs from USB or a documented solar + battery option with deep-sleep between readings to extend field life.

• Working weather-station node prototype (ESP32, DHT22, BMP280, rain gauge, anemometer)
• Complete firmware source code (sensing, unit conversion, MQTT telemetry, deep-sleep)
• Cloud pipeline + working dashboard (live values, graphs, alerts) demonstrated with the prototype
• Circuit and wiring documentation with sensor calibration notes
• Component list with ratings
• Buyer-run calibration procedure (reference thermometer comparison)
• Project report PDF (meteorological sensing background, MQTT/IoT architecture, methodology, test procedure)
• PPT presentation for final review
• Viva Q&A preparation document (sensor principles, MQTT, heat-index derivation)
• Setup and demonstration guide

• Sensor accuracies are datasheet values (DHT22 ±0.5 °C, ±2% RH typical); this is a teaching station, not a meteorological instrument — readings are compared against a reference during the buyer's calibration, not claimed as lab-grade.
• The rain gauge resolution is 0.3 mm per tip by design; drizzle below that per event is under-counted.
• Telemetry needs a stable 2.4 GHz Wi-Fi link; long outdoor runs depend on signal quality and are not part of the base build.
• The anemometer measures wind speed only; wind direction needs an additional vane sensor (listed as future scope).
• Push alerts need internet on both the node and the phone; local Wi-Fi outages queue nothing.