For more and detailed info about my reply see: https://github.com/teusH/MySense (full documentation and software) and (Dutch) http://behouddeparel.nl/samen-meten (global info and measurements results).
MySense is an Open Source Python framework for air quality measurements (sensor kits) and data proxy server. The framework supports all types of PM sensors (Dylos, Nova SDS011, Plantower PMS7003, Shiney PPD42NS, GPS, gasses (NOx, SO2, O3, NH3) sensors from Spect, meteo (DHT22, BME280/680). Comminucation: LAN, Wifi and LoRa.
Together with national health institute we had quite some (more as a dozen…) sensorkits equipt with SSD011/DHT22/Marvin LoRa for two month now running in several locations in Holland.
Experiences: SSD011 differ amongs each other, operable life time is limited for fan and laser, DHT22 differ too much (BME280 too) to be reliable for calibration of measurements, Marvin signal strength is limit is 1 km. The good story is: probably with more tests some type of calibration can be done. Personally a favor BME280 or BME680 (more reliable), and PMS7003 (more PM sizes, less energy, but no tube air inlet) and probably the PyCom LoRa controller (Python, antenna).
PM sensors have a fan: need more energy. I have my doubts on using (small) batteries in sensing air quality. E.g. a small TTN temp/light/motion/ESP LoRa sensor empties the battery in 3 weeks time.
See the github for the map “statistics” for the calibration reports.