Goleta Air Monitoring Project

Trusted Data From The Physical World

Air quality in our own back yard impacts how we feel today and our long term health. In this project we built five custom air quality stations using commercial grade sensors, Raspberry Pi and Arduino and connected them through zymbit.connect software to generate real time charts. The trusted data streams are further integrated into environmental analysis software from Groundswell Technologies. Air Quality in My Back Yard - Graphic 1

System overview

The complete system is mounted in a modified solar radiation shield crowned with an IP67 enclosure. All sensors are mounted on a custom motherboard which takes care of power supply and mechanical mounting of these tiny parts. Sensor data is acquired using an Arduino Zero mounted within the top tier of the radiation shield. From there, data is packaged and sent to a Raspberry Pi via a serial connection, external to the enclosure. This way heat is properly dissipated and does not affect sensor measurements. The Raspberry Pi acts as the connection gateway and publishes the packaged data to zymbit.com/console. For immediate purposes data flow is unidirectional, meaning the unit is not subscribing to any outside streams, though this could easily be integrated. Additionally, with room to spare in the Raspberry Pi enclosure, we added a POE (Power Over Ethernet) splitter for versatility. This makes installation simple and improves overall reliability since the unit only requires a single cable connection and POE can handle wide line voltage variations. The user can then choose either a Wi-Fi or Ethernet connection. Of course, a simple USB cable will also work to power the unit.

Custom sensor board

At the heart of the Air Quality station is a custom design sensor board that integrates multiple sensor types:

  • Particulate matter. - Telaire SM-PWN-01A(2.5 micron and 10 micron channels)
  • CO2. - Telaire Mini Module 5713
  • Relative humidity. - SHT1x
  • Temperature. - SHT1x
  • Barometric pressure. - SparkFun Altitude/Pressure Sensor Breakout with MPL3115A2

The particulate sensor was the primary driver for the board’s design; it uses a small convection heater to circulate air and this requires the module to be oriented vertically. The particulate matter sensor has two channels; one for ~ 2.5 micron particles and one for ~ 10 micron particles. Each channel produces pulses which are measured and accumulated over thirty second intervals. The other sensors are managed via I2c bus and are all averaged over 15 second intervals.

Mounting components inside the solar radiation shield.

We get plenty of sun in southern california so we need a solar shield to protect the sensors while letting air flow freely to them.

Waterproof enclosure for RPi.

And sometimes we get rain too, so we protect the RPi by putting it in a waterproof IP65 enclosure.

Data work flow

As mentioned above, the Raspberry Pi acts as the connection gateway for the Arduino Zero acquisition unit. Installing and running the Zymbit Connect software on the RPi very easy. Here is a step-by-step tutorial. Once data is published it is stored in the Zymbit Cloud and is easy to interact with using instant dashboards or the Zymbit API. For this application the engineers at Groundswell used the Zymbit API to pull in the raw data into their analysis and visualization software. You can see cool real time results below.

Air Quality in My Back Yard - Graphic 1

Air quality stations installed in private back yards

Five Zymbit air quality stations are now deployed and active in the area of interest. Each unit took less than an hour to install and start receiving data. Data streams from each unit are now being integrated into Groundswell's geospatial software and we expect everything to be operational in early November 2015.

GAM 1to5

Zymbit hardware is self contained and designed to operate outdoors in a nominally shaded area. For this initial project connection to the internet is made via WiFi or ethernet connection to host building gateway/router. For subsequent projects (2016) Zymbit has plans to provide options for solar power and cellular connection.

Real-time charts keep the commmunity in the know

Real time data charts provide a simple and powerful way for community members to review and interact with data. Custom dashboards can easily be built for specific projects.

AQIMBY Charts 2

For more information

The Goleta Air Monitoring project is now operational and delivering real-time data to the local community. You can follow along as the project develops and share what residents learn about the air they breathe.

  • bboyes

    Might any of these sensor boards be available either bare or assembled? I’d be interested if so. Also curious why the choice of SHT1 vs what appear to be better TMP102 temp and HIH613X temp/humidity? Interesting project. Any problems using the Arduino Zero, or is support for it pretty stable? Thanks for posting this.

  • zymber1

    @bboyes thanks for your comments. Yes we will be offering a machine-assembled sensor board in our store early in the new year. If you need something sooner send us a private email to hello@nullzymbit.com. Regarding the choice of Humidity/Temp sensors, the SHT1 was used for this proof of concept because of the superior specs and ease of hand assembly. For a machine assembled board we would likely use the SHT2 which is generally has better specs than the Honeywell HIH61310 device and is about 30% of the cost. See attached comparison based upon manufactures published datasheets. The TMP102 is a great if you only want temperature.

    Re the Arduino Zero, we had some teething issues, but nothing fundamental. We’ve added some pretty neat over-the-wire code that we will talk about in a later blog.

  • Pingback: Measure the air quality in your backyard | Atmel Bits & Pieces()