BSEC 2.0 + BME688 to detect BSEC_OUTPUT_GAS_ESTIMATE_1

[AkmalSQ]

Hi, I'm using BSEC 2.0.6.1 and BME688 sensor with a config generated using BME AI Studio. The config is a simple coffee detection algorithm trained using BME AI studio.

Unfortunately I am having issues using the BSEC 2.0 library. I have implemented my code according the bsec integration guide and the bsec example. 

Problems:

1. I am unable to use a sample rate other than BSEC_SAMPLE_RATE_HIGH_PERFORMANCE when calling bsec_update_subscription() with BSEC_OUTPUT_GAS_ESTIMATE_1 enabled.

2. If BSEC_SAMPLE_RATE_HIGH_PERFORMANCE is used, I am not getting new information for requested outputs that relate to gas measurement. The aqi, gas_percentage and gas_estimate_x values are not changing.

Questions:

1. Do you have sample code for using bme-ai solution with bsec2.0? 

2. What limitations apply to the sample-rates requested using bsec_update_subscription?

It seems as if others are also facing this issues.

Thanks,

Akmal

[BSTRobin · Accepted solution]

Hello AkmalSQ,

Is there any error or warning code on your side?

The example code in BSEC2 release package is verified and can be used for start up.

Please make sure you used the bsec_serialized_configurations_selectivity configure string. And for selectivity solution, when gas estimated is requested, we can only use BSEC_SAMPLE_RATE_HIGH_PERFORMANCE, otherwise gas estimated will not be updated.

Could you give feedback?

[RuiMPCosta]

Hello,
I am using the bsec_serialized_configurations_selectivity configure string as well as BSEC_SAMPLE_RATE_HIGH_PERFORMANCE, however I keep getting unexpected values for GAS_ESTIMATE1 and GAS_ESTIMATE2, that value being equal to one for gas_estimate2 and zero for estimate1. This value never changes no matter the environment inserted, being that the values of bsec_serialized_configurations_selectivity are from a C02 test algorithm.
Regards

[BSTRobin]

Hello RuiMPCosta，

Do you mean you choose other config in your code? If yes, which config file you used?