QuickFeather is good development platform for AI applications. First install the tool and test the hardware.
For arm-m4 for tinyfpga-programmer shall be installed according to documents.
2 Flashing the binThen Flash the bin into the hardware.
python c:\Users\north\envDL\TinyFPGA-Programmer-Application\tinyfpga-programmer-gui.py --port COM14 --m4app output\bin\qf_helloworldsw.bin --mode m4
Then, the bin is flashed
Press reset, enter into another UART port for user interface and send command like
diag red
The red LED is on
Other command shall be like this.
Connected Sensirion SGP-40
The I2C shall be used as ,
In Jumper J8,
Since this sensor use same I2C port as on-board sensors, most of the code can be shared. Only sensor shall be added to Data Collection Lab and SensiML,
The Data stream, train and prediction both can be shareed.
4 Download the pack and predictAfter training the data , model shall be download to make it works.
Due to limited the data, the train process could not converged to valid model.
More training have to be done.
5 SummaryThe QuickFeather is versitile, but need some time to familiar with the dev-environment. But SensiML and Data Collection Lab are innovative products, worth further study and application.
Comments