I’m also looking into this at the moment. The PM sensor choice (SDS011) of Luftdaten.info seems pretty reasonable to me. This community project was developed in a great test environment with lots of smog sponsored by the german car lobby in Stuttgart… Speaking of testing, operation of these sensors in closed rooms next to a 3D printer should be interesting.
Did anyone try that?
Nice comparative measurements of SDS011 (17 USD on Aliepxress) with a much more expensive reference instrument (Grimm 1.108) and the same sensor that Safecast is using (Alphasense OPC-N2, 400-500? USD) can be found here: http://www.opengeiger.de/Feinstaub/CalibDocu.pdf
A few more interesting articles in english are linked at the bottom of http://www.opengeiger.de/Feinstaub/feinstaub.html