wow that’s a lot of answers, thank you!
Of course low cost means different values for different people, however, as we all have to think economically usually this means “as low as possible while doing what it’s supposed to do”. I don’t need scientifically accurate temperatures but it should be consistent.
The reason I’m asking is: A while ago we tested CO2 devices from a) Globalsat and b) Zane.hu - the sensors that zane used drift extremely, the Globalsat devices are very consistent. We compared 10 devices each, same location and setup. While drift is expected and in general only a trend canbe meaningfully used, it comes down to the auto calibration they use.
The thing is, if you show both charts to a customer - it’s obvious that he will choose the one where all the measurements are aligned nicely - no matter if this is clever auto calibration, fake data or accurate measurements.
Since the LoRaWan market is quite new, there are huge variances in both price and quality, that’s why I prefer to ask about other experiences before testing.
As @BoRRoZ mentions you cannot sell Arduino/Homebrew/3D printed sensors to cities, no matter how cheap and accurate they are, it’s simply no option.
In the first phase it will be tested with around 20 sensors, expanding to 100-200 at a later stage.