John Nordby, a machine learning engineer, set out to create a development board suitable for tinyML applications with a price tag of one dollar.
The miniaturized board developed is based on the Puya PY32F003 microcontroller with an Arm Cortex-M0+ processor. It may not be too powerful, but it is enough for many applications. A Holtek BC7161 Bluetooth Low Energy Holtek BC7161 has been included for communication, and a LIR1220 button cell battery and charger to complete the basics of the build.
For the device to be useful, it must gather information for analysis. To accomplish this, sensors were installed on the board. Specifically, an STMicroelectronics LIS3DH three-axis accelerometer to collect motion and acceleration data, and a MEMS microphone (sound sensors) to record sound.
Programming the custom boards should be fairly easy as they have been designed to be compatible with emlearn. This framework allows developers to train their models in Python and then run them on any device, even low-power microcontrollers like the Puya PY32F003 with the C99 compiler.
If you want to build your own copy of this board to experiment with tinyML, be sure to check out Nordby's project description: https://hackaday.io/project/194511-1-dollar-tinyml. The design files and source code are also available on GitHub: https://github.com/jonnor/embeddedml/tree/master/projects/dollar_tinyml to get you started.
For the device to be useful, it must gather information for analysis. To accomplish this, sensors were installed on the board. Specifically, an STMicroelectronics LIS3DH three-axis accelerometer to collect motion and acceleration data, and a MEMS microphone (sound sensors) to record sound.
Programming the custom boards should be fairly easy as they have been designed to be compatible with emlearn. This framework allows developers to train their models in Python and then run them on any device, even low-power microcontrollers like the Puya PY32F003 with the C99 compiler.
If you want to build your own copy of this board to experiment with tinyML, be sure to check out Nordby's project description: https://hackaday.io/project/194511-1-dollar-tinyml. The design files and source code are also available on GitHub: https://github.com/jonnor/embeddedml/tree/master/projects/dollar_tinyml to get you started.