EGM Muscle Tracking

Muscle activity sensors that can track your exercise progress.

Outside of my design work, I'm actually really into traditional weight training. However, lifting weights requires a certain degree of mind-muscle connection and a firm grasp of form. If you aren't doing the movements or flexing your muscles correctly, you won't get good results. This is something you grasp over time or figure out with a personal trainer. But I wanted immediate feedback training on my own. To that end:

I wanted to see if I was using my muscles correctly in real time.

IDEATION

How might we visual our muscle activity? I needed a way to detect when I was using my muscles, determine the intensity of the muscle flexion, and give feedback to the user that's more easily interpreted than our own senses.

TECHNOLOGY

While the basic idea is simple, translating the sensor readings into a useful GUI is somewhat complicated to execute. Further, the method of detection and the sensor readings need to be carefully considered.

The best method of detecting muscle activity are EMG readings which, using electrodes or probes, can detect electricity running through our nerves and muscles. The intensity of the current while the muscle is flexing or under load is what allows the sensor to produce interpretable readings. However, these electrodes are one-time use and need to be applied correctly in order to get a reliable connection.

HARDWARE

MACHINE DIAGRAM

TESTING

Myoware 2.0 Muscle Sensor

My initial attempts at soldering didn't the Myoware 2.0 to the Arduino logic board didn't end well, so I decided to use the link shield and Arduino shield allowing me to use a 3.5mm audio cable.

Electrodes

Attaching the electrodes proved rather difficult. There's a specific orientation that the Myoware 2.0 Sensor needs to be attached at in order to get a reliable connection. I chose the bicep as my test muscle and I failed to attach it correctly, resulting in nonsense readings.

Once I figured out the orientation however, I got reliable readings performing dumbbell bicep curls.

Bench Tests

GUI

I used a muscle diagram template from Figma Community and imported it into Unity, which with a simple plugin can receive inputs from the Arduino R3. Additionally, by individually exporting each muscle layer in Figma and importing them into Unity, I can highlight each muscle group and make changes without affecting the others.

DEMO

Reflections

I ran into a lot of issues during this project just trying to get the hardware right. Strictly speaking, I only needed the sensor, electrodes, and a logic board but I ended up with a whole host of accessories to avoid the need for soldering. Not to mention replacement electrodes. However, I really loved the core concept, and I'm quite happy with the results.

A real application of this prototype would need to be far more convenient and involve more than one EMG sensor, but I feel that there's solid product potential in this.

Next Steps

More user testing
Write an algorithm for processing the readings
Product sketches and prototyping
System design and app integration