Hey there,
I have performed some "calibration myself".
I have taken readings from different sensors:
Notes:
- Every row in the table is a different location
- Every reading is the average of the last 5 readings, after 3 minutes of running to have the value settle.
- The Smart Air (which is boasting 99% accuracy and is about $120, has been left to run for 15 minutes to settle.
- For the pressure I used an Airport Weather Station that is close to me.
And then used Linear Regression to calibrate it to the accurate values (which is the Smart Air Monitor):
Temperature:
Humidity:
The pressure only needs a 0.7% flat increase.
So the final formulas to have your output correct are:
- correct_temp = 1.0095 * nicla_temp - 4.8051
- correct_hum = 1.4383 * nicla_hum -2.5628
- correct_pressure = 0.718/100 * nicla_pressure + nicla_pressure
Furthermore, the Air Montior rounds up temperature within the next .5, and the both the humidity and VOC are rounded to the next integer. So those values were used for Linear Regression.
Mind the fact that the IAQ is calculated differently on the Air Monitor by using the Particulate Matter and Carbon Monoxide, with different weights. The Nicla calculates it with temp, hum, voc, estimated co2(not actual).


