With regards to BME688 I guess there are already several known gases for which there are well-tested configurations available, apart from the Co2 in the default bsec library sources?
You could refer current supported BSEC output. If the gas you want to detect is not in the BSEC output list, you can use BME AI-Studio to collect data, train data and generate configurations to detect it.
The above is obvious from the datasheet. However, since there are many gases out there it seems unnecessary that each one of your customers should develop the data for common organic compounds, for instance CO (some of which are not so easy to handle). But should I take it as a no?
It is extremely difficult to train the sensor on standard VOC gases without a sophisticated and somewhat dangerous lab environment. I imagine most of the clients are invested in the software development and computer hardware, not in labs. If the BME688 has already been tested and working with certain VOC gases with known documented ranges, can these configurations be shared? That would be very beneficial to all clients.
Except for the configuration in the example code, we have no configuration that can be disclosed. You can use the algorithm trained by BME AI-Studio to identify the type of gas you want to detect.
