Our first gen units are using PLANTOWER PMS5003 PM sensors along with a Winsen MQ131 Ozone sensor. The PLANTOWER PM sensors have performed admirably through the course of a Michigan winter and the data generated from them aligns nicely with the federal reference instruments that they are positioned next to. I wish I could say the same for the MQ131 - that sensor is nearly useless, it uses variable voltage output that, once settled, never changes at all. I could have saved myself some money by replacing it with a resister for all the good it has done.
I’ve spoken to Spec directly about their sensor line. They have been in the gas sampling business for a while and are just now selling sensors for IoT use. They have datasheets in English (!) and distribution through actual resellers like DigiKey, all of which I wasn’t getting from Winsen. Further, they offer it in a factory-calibrated package which shoots out digital readings over a UART connection, making them dead-simple to integrate.
So, thus far I LOVE THEM, however - they have not been placed in the field, and I have no idea (yet) how well they actually perform, so I’ll withhold judgement until I see the data.
One substantial drawback is the price - at $75, that one module clocks in at 3x the price of the rest of the package combined, and has resulted in the BOM going from ~$30 to ~$100 which is a little annoying.