03-07-2023 11:51 PM
Hello,
I got a new BME688 development kit, configured and its is recording the data. I used the recorded data to create an AI algorithm using the AI studio. Now,
1 - Is there any way I can test the algorithm with further recordings?(other than the app, the app has only a demo feature not a live testing) It need not be on realtime, but like recording the same and then analyzing the algorithm I created earlier. (I don't have much experience with IDE or arduino in general)
2 - How can I read the raw data of sensor output? Like resistance vs time, a CSV file or data log of the all 8 sensors?
05-04-2023 08:41 AM
Hi atsaleel,
Do I have to re-input the library into the arduino after replaceing the 'hand_sanitizer' wiht the config file I created?
Reply: BME AI-Studio only generate config file, not generate library, replace "FieldAir_HandSanitizer/FieldAir_HandSanitizer.h: with your config file. But remember to keep same BSEC version with BME AI-Studio, see the dependencies in release note under BME AI-Studio package.
Relase note under BME AI-Studio-v2.0.0-win package:
Also, How should I define the classes I created in the basic_config_state.ino, this line :- const String gasName[] = { "Field Air", "Hand sanitizer", "Undefined 3", "Undefined 4"}; ?
Reply: These classes are determined by the classes you created in BME AI-Studio, have you created any classes?
2 weeks ago
Hi, i would like to continue this discussion since i have the same question:
I have successfully compiled same example and then added my own data exported from the AI studio, where Classes are defined as:
ClassA = "Coffee"
ClassB="Fresh Air"
I would guess that the order of the Classes should be the same:
const String gasName[] = { "Coffee", "Fresh Air", "Undefined 3", "Undefined 4"};
However, predition on the open air says that Class 1 is 98%, which is what i expect for the fresh air, not for Coffee
Can you please expalin the logic of defining class name's order here?
Thank you,