Over the last two or three weeks I went to a forrested area far from home and captured smells of many objects via "clean air smell", "object in clean air smell", "clean air smell", "object in clean air smell again" each for 30 minutes. I was driving back home, through NYC and decided to get an Air Quality reading and lost my microsd card in the traffic.
That was fun, I lost all my data, but each specimen I took wasn't examined to determine the best temperatures & settings for capturing the presence of these objects. I will tape the microsd card into the board as I start again in a slightly different ecosystem.
I am very interested in equipment and setups that people deploy and if measurements should be taken at different humidities, temperatures, barometric pressures, seasons and concentrations. I am VERY concerned with contaminating the BME688 featherwing sensors and the esp32 featherboard with the objects being measured. Two very large stainless steel tea strainer that can hold the esp32 huzzah featherboard, featherwing and battery would be a good find.
I know I have a lot of work to learn how to take good measurements and many measurements to apply them to my neural network. I am using tensorflow. I am going to test AI studio to see how useful it is. My final goal is to measure a small wooded environment and detect the presence of specific mammals with video and collect measurements. I have video equipment setup that currently detects these animals via a ML model. I hope to duplicate that work via the BME688 sensors. I have 2 BME688 sensors and an 8x BME688 featherwing, python, keras and I'm not araid to use them.