2022 A classroom that tells you when to open the window. archived

Klimostat 2.0

School project · 6 months · team of four

2: sensors
5: tiers, end to end
~100 h: my share

A milled copper board, four connector groups marked in red.

What it does

Stale air makes people slow, and nobody notices until the lesson is already lost. Klimostat 2.0 measures it: a Raspberry Pi station with an MH-Z19 NDIR carbon dioxide sensor and a DHT22 for temperature and humidity, reporting into Azure IoT Hub. An Event Hub function forwards the readings to an Express and MongoDB backend, and once pollutant levels cross the threshold the system issues a ventilation recommendation.

The whole chain is there, from a sensor on the wall to a Vue dashboard — including the parts that usually get skipped: a printed enclosure, JWT authentication, request validation and rate limiting on the API.

Board layout with blue traces between labelled pin headers. — The layout the board was milled from — the traces run between the Raspberry Pi header, the CO₂ and temperature connectors and the status LED.