Our analytic activewear is designed for active, fitness-oriented individuals. The hope is that individuals can leverage our activewear to make well-informed and data-driven decisions of the performance of an exercise routine. The system is built to reinforce positive training habits such as good form and help prevent over-exertion of muscles.
This is accomplished by (1) visualizing muscle contractions on the activewear via LEDs that leverage EMG bio-signals, and (2) machine learning and AI to adapt exercise routines and predict good performance of an exercise - such as curls with a dumbbell.
How we built it
We used EMG sensors to measure electrical signals elicited by upper-limb muscles. The data from the sensors, which correspond to force exertion, are shown as different colors attached to the activewear, while all information is visualized in real time and leveraged by our machine-learning data model to predict good performance of an exercise.
Challenges I ran into
HARDware is HARD. Hardware and software integration is even harder. Darn you Python3 and no-backward compatibility with Python2.7
Accomplishments that I'm proud of
Teamwork, proud of our team dynamics and execution. This was our first hackathon!!!!!!!!!!!!!!!! Made new friends.
What I learned
Technology sucks, fashions rocks. Jk, we <3 both. We learned how to analyze EMG signals along with hardware.
What's next for Analytic Activewear
Plan to raise VC by next month, launch by the second quarter of 2018.