How microchiptech/e-nose works
E-nose is a six-sensor gas-classification system targeting an Adafruit Feather M4 Express. Its documented workflow collects sensor readings, logs them as CSV for optional Edge Impulse training, then deploys an Arduino inference sketch that reports class probabilities. The repository tree contains only README.md, so implementation paths cannot be verified; nodes therefore have null paths. Relationships shown here are limited to those stated in the README.
Components 10
Data collection
- Six-sensor array
- Sensor firmware
- Python logger
- CSV dataset
Model development
- Edge Impulse
- Classification model
Embedded inference
- Feather M4
- Inference sketch
- Class probabilities
Other components
- Air sample
Connections 11
- Air sample → Six-sensor array: sensed by
README.md - Six-sensor array → Sensor firmware: read by
README.md - Sensor firmware → Python logger: serial output
README.md - Python logger → CSV dataset: writes
README.md - CSV dataset → Edge Impulse: optional upload
README.md - Edge Impulse → Classification model: trains
README.md - Classification model → Inference sketch: exported for
README.md - Feather M4 → Inference sketch: runs on
README.md - Inference sketch → Six-sensor array: reads analog data
README.md - Inference sketch → Classification model: runs inference
README.md - Inference sketch → Class probabilities: prints
README.md