Public Lab Research note

  • 8

MACA - Open Air Quality Monitor project in Latin America

by nanocastro |

nanocastro was awarded the Basic Barnstar by xose for their work in this research note.

What is MACA?

The MACA Open Air Quality Monitor project (Monitor Abierto de Calidad del Aire in spanish) started a few years ago in Mendoza, Argentina to tackle some problems related to local air pollution monitoring (no public data, high costs of reference instruments, etc, etc) by developing an open, small and low cost air quality monitoring station.You can find more info about the MACA project in our wiki.

What are we doing?

Our first prototype was capable of measuring ozone and it worked ok as a proof of concept but we are currently working at Nodo39 Fablab on a second prototype that combines an optical particle detector (Shinyei) and metal oxide gas sensors for CO, NO2 and O3 (MICS-e2v). It's a remix and adaptation of previous citizen air quality monitors (AQ Egg, Smartcitizen, Dustduino and Airbeam). It is based on an Arduino Mega 2560.

Second prototype

This second version is still a work in process. We designed and assembled the first version and we are just starting to get data from the sensors. Below a description of this prototype.

Gas sensors

A general survey and a lot of resources of different gas sensors and sensor boxes for air quality monitoring can be found at Snuffle.We chose metal oxide (MOx) gas sensors due to its really low cost compared to electrochemical sensors and because we can remix with/from previous open source monitors that use this kind of sensors (AQ Egg, Smartzitizen, AirCasting, EveryAware).

The MOx sensors consist of a sensing layer, composed of a metal oxide, generally SnO2, that is heated by a heater structure. When chemicals are absorbed on its surface, its electrical resistance change. Ozone and NO2 are oxidising gases and cause an increase of the sensing resistance. On the contrary, reducing gases (CO, VOC) cause the resistance to decrease.

Metal oxide sensors generally show temperature dependence and cross-sensitivity to other trace gases. They are also sensitive to fluctuations of relative humidity and show baseline drift with time. Using MOx sensors quantitatively requires that a model be developed which not only characterizes the relationship between sensor resistance and gas concentration, but also includes the impacts of these other variables and sensor characteristics.

All the sensors used are MICS and were purchased from SGX Sensortech Limited (fromerly e2v).

List of sensorsimage description

They are all through hole mounting but there seem that are no longer produced by this manufacture, only SMD mounting sensors. The MICS-OZ-47 used in the first prototype is no longer available. All sensors are exposed like in a passive sampler, that is, in a vertical tube with the MICS suspended near open end.

image descriptionimage description

More details about MOx sensors can be found at:

  • Fine, George F.; Cavanagh, Leon M.; Afonja, Ayo; Binions, Russell. 2010. "Metal Oxide Semi-Conductor Gas Sensors in Environmental Monitoring." Sensors 10, no. 6: 5469-5502.
    1. Eranna. 2011. "Metal Oxide Nanostructures as Gas Sensing Devices". CRC Press, Boca Raton, FL,316 pp.,ISBN 978-1-4398-6340-4.

Some other references were MICS sensors are used and evaluated

  • Piedrahita, R., Xiang, Y., Masson, N., Ortega, J., Collier, A., Jiang, Y., Li, K., Dick, R. P., Lv, Q., Hannigan, M., and Shang, L. 2014. "The next generation of low-cost personal air quality sensors for quantitative exposure monitoring". Atmos. Meas. Tech., 7, 3325-3336, doi:10.5194/amt-7-3325-2014, 2014.
  • Spinelle et al. 2015, "Field calibration of a cluster of low-cost available sensors for air quality monitoring. Part A: Ozone and nitrogen dioxide". Sensors and Actuators B: Chemical, 215, 249-257.
  • Spinelle et al. 2017. "Field calibration of a cluster of low-cost commercially available sensors for air quality monitoring. Part B: NO, CO and CO2". Sensors and Actuators B: Chemical, 238, Pages 706-715, ISSN 0925-4005,
  • Air Quality Sensor Performance Evaluation Center (AQ-SPEC) by California's SC-AQMD

Particles sensor

We are using the Shinyei PPD42NS, the same sensor used in the Dustduino. The are other really interesting laser based particles sensors and open source monitors like MyPart but we chose it because is the most documented and easier to get PM optical sensor. The are a lot of comments and Research Notes about this sensor.

Some of the known issues about the Shinyei:

  • From LPO to concentration [pcs/0.01cf]. This has been done by fitting a cubic polynomial curve to the digitized Spec Sheet characteristic curve of Cigarrette smoke vs LPO.
  • In order to read both outputs of the sensor simultaneously (P1 and P2) interruption routines are needed.
  • Conversion to mass concentration (ug/m3) is done by making assumptions about PM10 and PM2.5 shape, size, and density.
  • There is an intrinsic limitation to the use of optical techniques as proxies for gravimetric measurements. When a difference in measured values is observed, one cannot be certain whether it is attributable to a difference in the total mass, size distribution, or optical properties -- or some combination of all three -- of the measured aerosols (Holstius et al., 2014)
  • The window in the front of the sensor must be closed!
  • Response improvement of the sensor can be achieved by connecting a fan and by incrementing the integration time of the reading.

We designed a shield and a fan duct that can be 3D printed and attached to the front of the Shinyei in order to close the front window of the sensor and to increase response by increasing the airflow across the sensor.
image description

Temperature and relative humidity

We are using the widely known and documented DHT22.

Communications and data logging

A WiFi shield is mounted on the Arduino to send data to an IoT platform (we have not chosen one yet, recommendations are welcomed) and to log data on an SD card


The ongoing documentation about hardware and firmware can be consulted on Github.

Next steps

  • We want to start getting data to verify that the sensors are working ok.
  • Redesign the enclosure in a way that can be placed outdoors.
  • For calibration purposes we will conduct an in-field calibration by placing MACA along with Reference instruments (There is only one monitoring station that works like 6 weeks in the whole year)
  • Evaluate the response of MOx sensors and the influence of T y RH.

We will show some real world data and updates on following research notes.

All comments and suggestions are welcomed.

Thanks for reading.

air-quality monitoring dustduino particles gas-sensor shinyei ozone south-america metal-oxide latinamerica carbon-monoxide nitrogen-dioxide argentina



Hi @nanocastro, thanks for cross-posting from your excellent wiki! Some open data friends mentioned this government open data portal for Buenos Aires -- have you worked with their air quality data before? CC @nshapiro

Hi @liz. De nada, I think this data portal is quite new. Two years ago I contacted with the person in charge of the BsAs monitoring stations to calibrate my sensor and at least she was pretty jealous about validating the data. I don't know if anyone is using this data but I can ask around. Saludos .

thanks @nanocastro! if have any questions for the PL community for the next steps please let us know :)

Awesome!! Soon we'll update the Leptos station, a spanish cousin of your beautifull MACA ;) Would be great to meet digitally to exchange impressions. Documentation here is outdated. We're working with the Multichannel gas Sensor by Seeedstudio in the place of the MQ2

Reading your documentation right now ;D

@xose awards a barnstar to nanocastro for their awesome contribution!

Hi @xose. Muchas gracias.The multichannel looks good and it has more or less the same sensors (except for NH3). The documentations is still incomplete, I will try to update some of it by next week. Yes we should meet you can write me to ferhcastro on gmail. Saludos .

You must be logged in to comment.