Alex Jagger
Published

Build a Smart Dumbbell with the SAMD21 Machine Learning EVK

Turn your dumbbell into a smartbell with the Microchip SAMD21 ML EVK and an embedded ML classifier built in the Edge Impulse Studio.

IntermediateFull instructions provided2 hours1,552

Things used in this project

Hardware components

SAMD21 Machine Learning Evaluation Kit with Bosch IMU
Microchip SAMD21 Machine Learning Evaluation Kit with Bosch IMU
The SAMD21 ML Evaluation Kit is also available with a TDK IMU sensor
×1
USB-A to Micro-USB
At least 4' long to allow freedom to exercise
×1
Dumbbell
Chose any weight that your are comfortable with, but keep in mind that you will need to collect at least 10-15 minutes of training data
×1
Double Sided Tape
×1
Rubber Bands
×1

Software apps and online services

MPLAB X IDE
Microchip MPLAB X IDE
This project utilizes several tools within the MPLAB environment: MPLAB XC32/32++ Compiler, MPLAB Data Visualizer, and the ML Plugin
Edge Impulse Studio
Edge Impulse Studio
The embedded ML development platform

Story

Read more

Code

SAMD21_ML_Kit_Datastreamer.zip

C/C++
MPLAB X project for the SAMD21 to stream 6 axis IMU data over UART to MPLAB Data Visualizer for upload to the Edge Impulse Studio. Both the IMU2 (Bosch) and IMU14 (TDK) are supported by separate project configurations.
No preview (download only).

Smartbell-SAMD21-BMI160.zip

C/C++
MPLAB X project with integrated Edge Impulse library for the SAMD21 ML Kit with IMU2 (Bosch).
No preview (download only).

Credits

Alex Jagger

Alex Jagger

1 project • 5 followers
Applications Engineer at Microchip Technology

Comments