Tech behind the sensors
We’re building a Raspberry Pi based weather station to help increase the amount of granular weather data in and around Durban, and to help data literacy in schools by enabling students to interact with the data from the stations to analyse, draw conclusions and create data stories about their local environment.
The stations can measure temperature, humidity, wind-speed and direction, the presence of certain harmful gases and the amount of microscopic dust particles in the air. The gas and dust measurements will help us to determine the quality of breathable air in various parts of the city and have the potential to provide some very interesting results.
Components
The weather station will be centered around a Raspberry Pi 3 Model B single-board computer. We have chosen this board over something cheaper (Arduino for example) because this board is
basically a mini computer complete with an operating system and storage space. This makes the board far more accessible, both in terms of networking and operation as well as in terms of computer skills and experience. The added advantage of storage space is that the data is not lost if the internet connection breaks down. One advantage that the Arduino has over the Raspberry Pi
is that the Arduino can read both analog and digital signals, while the Pi can only read digital. This is problematic as some sensors have an analog output. In order to be able to read these sensors, we need to include an analog-to-digital (ADC) converter such as the MCP3008 microchip — a low cost, 8-channel, 10-bit ADC.
We measure the temperature and humidity with a DHT22 sensor. We chose this sensor over the lower cost DHT11 because it is more accurate and more precise without being much more expensive.
We determine air quality by looking at the amount of harmful gas and dust particles in the air. We measure the levels of Benzene, alcohol, and smoke in the air with an MQ-135 gas sensor. The amount of dust is measured by a Nova PM SDS011 High Precision Laser sensor. This sensor measures particles at 2.5 and 10 micrometers in diameter in mg/m3
We measure wind speed and direction using a Maplin anemometer and wind wind vane.
We combine all of this onto a “breadboard” to produce our first prototype!
Our next steps will be to house this system in an enclosure and to begin collecting data for a trial-phase analysis of performance.
- Heiko Heilgendorff
Originally published at Open Data Durban.
