PM2.5 Device Comparison

There are so many devices available right now on the market that measure PM2.5 (as we can see on this post) but there aren’t all of them the same. In my opinion PM2.5 is the key for air quality monitoring. Different sensors, different manufacturers and also different software can give different results.

Let’s try to analyze them and see which of them are accurate and which ones aren’t. In order to do that firslty I will present some the sensor manufactures and models. Prices are in USD:

Novafitness SD011 ∼23$
Plantower PMS1003 ∼17$
Plantower PMS3003 ∼17$
Shinyei PPD42NS Made in Japan ∼10$ (R² ∼.83)
Shinyei PPD60PV (More expensive than the PPD42 because it allows smaller particle detection) Used by AirBeam
Samyoung DSM501 Korean copy of the Shinyei PPD42 ∼5$ Used by Speck
Sharp GP2Y10 ∼15$ (Used by Awair and probably by Foobot) (R² ∼.74)
TZOA RD02 135$-600$ Used by TZOA Wearable (R² .9253)
Dylos DC1100 ∼300$ (R² ∼.85)
DustTrak TSI DRX 8533 ∼3000$ (R² ∼.97)
Turnkey Instruments Dustmate ∼3200$

Here we can see some graphs of different sensors cheap ones and expensive ones. In general the expensive ones are more accuarte and industrial graded but some new companies have managed to build low cost sensors with high accuracy.

Screen Shot 2016-01-29 at 22.49.01 — aqicn.org/sensor

Screen Shot 2016-01-30 at 19.12.23.png — Shinyei R-squared .83 correlation value

Screen Shot 2016-01-30 at 19.12.59 — Sharp R-squared .74 correlation value

gm4z989eevgqiw7g3uy7 — TZOA R-squared .9253 correlation value

On the table below there are some of the devices that measure PM2.5 among other stuff and have been reviewed by the Air Quality Sensor Performace Evaluation Center aka AQ-SPEC www.aqmd.gov. Focus on the R² (correlation) values on the last column “Sensors vs FRM/FEM Method”.

An R² of 1 indicates that the regression line perfectly fits the data, while an R² of 0 indicates that the line does not fit the data at all. So, closer to 1 then better results.

summarytable