Over 5.5 million people world-wide are estimated to die every year because of atmospheric pollution, and although the developing world has greater problems than the west, we still are not great. The UK has some of the worst pollution in Europe, and the south-east is particularly bad. It is estimated that almost 6% of deaths in the Reading area are due to particulate pollution alone.

The project uses LORAWAN to send data about a dozen different gases to the Things Network. This includes carbon monoxide, carbon dioxide, methane, nitrogen dioxide, propane, isobutane, ammonia, hydrogen, ethanol, and total volatile organic compounds (TVOC). The system also records temperature, humidity and atmospheric pressure.

There are numerous applications for this project including internal and external air quality monitors, agricultural monitors (looking at gases like ammonia), traffic exhaust monitoring, greenhouse monitors looking at assisting growth with enriched CO2 atmospheres, and gas leak alerts. Custom level triggers allow emails to be sent if parameters go out of range.

We used a Sodaq Explorer board, and a variety of sensors (BME280, CCS811, MiCS-6814). The embedded software was written using the Arduino IDE. A custom javascript decoder was written for The Things Network, microshare.io was used for storing and manipulating the data, and angular was used for the GUI.

Getting all of the sensors working well was a challenge. There were a number of hardware problems like non-standard I2C addresses and physical wiring problems. One of the more challenging issues was a non-working driver supplied with the most significant sensor we were using. It hung on I2C operation and took many hours to track it down. It was eventually traced to timing issues on the low-level I2C operations that required delays in the order of microseconds to be inserted to make it operational.

We were pleased with the way the project came together - we had no experience of Sodaq, Lorawan or any of the sensors before starting and other than purchasing the sensors the day before, all work was undertaken over the course of a day and a half. Solving the technical I2C issues and getting the GUI, data processing systems and email alerts all working took a significant amount of work.

We learnt about the LORAWAN technology, interfacing to gas sensors, the microshare.io platform, the Sodaq bard, The Things Network, and how to connect all of these together.

It would be great to develop this further with a custom hardware build complete with processor, radio, sensors, battery and enclosure, as a way to make this design accessible to a wide variety of interested parties.

Built With

Share this project:

Updates