Our team wanted to hack something that would promote public safety. A discussion on the length of the bus ride from UCLA to Cal introduced a discussion on how to safely drive long distances. There were two big problems to tackle: staying awake, and safely communicating.
What it does
The MyoMessenger offers two safety features. The first harnesses Myo technology and allows you to send common text messages to any person on a predetermined emergency list (e.g. text "I'll call you back later" to Mom) using hand gestures. This allows you to communicate your whereabouts safely with friends and family without needing to deviate your eyes from the road.
The second safety feature uses the Muse headband's accelerometer to detect when a driver's head starts bobbing in exhaustion or fatigue. When the driver is detected to be falling asleep, it will send an audiovisual signal for the driver to wake up.
How I built it
For the Muse accelerometer, we obtain the accelerometer's position points and calculate the acceleration at those points. If there is a peak in the acceleration curve, surpassing a threshold, it indicates the head is bobbing, suggesting the driver has fallen asleep.
Challenges I ran into
We were trying to connect many components together (Myo, Twilio text messaging, Muse headband, etc) and this proved difficult to provide a smooth continuous user experience between all devices.
What's next for MyoMessenger
We want to eventually upgrade the MyoMessenger to be able to send one of the predetermined text messages to whomever called or texted last. The limitation of the current model is the user must take initiative to select which contact to send a message to first. We would also like to incorporate a speech to text functionality as well. Lastly, we want the Muse accelerometer to ping the Myo when it receives a spike in acceleration to create a seamless user experience rather than outsourcing the signal.