BME AI Studio

The BME AI Studio is designed to work seamlessly with the BME690 & BME688 development kit and related software (BME AI Studio desktop and BSEC) to fasten up and enhance the proof of concept with the BME gas sensor. The BME690 is further developed on BME688 which includes the gas sensor with Artificial Intelligence (AI) and integrated high-linearity and high-accuracy pressure, humidity and temperature sensors. The improvements of BME690 over the BME688 include reduced power consumption in IAQ modes, with increased robustness for use in environments with high condensation levels and enable adherence to WELL & RESET building standards. It is housed in a robust yet compact 3.0 x 3.0 x 0.9 mm³ package and especially developed for mobile & connected applications where size and low power consumption are critical requirements. The BME AI Studio tool (desktop and mobile App) enables customers to train the BME690 & BME688 gas scanner on their specific application, including home appliances, IoT products or Smart Home.

The BME AI Studio App connects to the BME690 & BME688 Development Kit using Bluetooth. Once connected two modes of operation can be choosed:
1. Record & Label Data: raw gas sensor data is recorded with the BME690 & BME688 Development Kit and the specimens can be labelled through the App. Prerequisite: board configuration file (.bmeconfig) must be copied on the SD card or flash drive (based on the development kit connected).
2. Live-Test Algorithm: the exported algorithm from BME AI Studio desktop can be or demonstrate live with real specimens. Prerequisite: a BSEC configuration file (.config) and an AI configuration file (.aiconfig) must be copied on the SD card or flash drive (based on the development kit connected).

In the Record and Label mode user can:
• see the current sensor state
• label the specimens
• see the recorded raw data
In the Live-Test mode user can:
• see the current prediction of each sensor (currently only four sensors at a time)
• set a class as current "Ground Truth" (for Classification Algorithms)
• open a full screen demo view

Please note that the complete BME690 Development kit includes BME690 8x Shuttle board and Application board 3.1.