Pulse

step 1 of cpr process through spectacles
step 2 of cpr process through spectacles
step 3 of cpr process through spectacles
step 4 of cpr process through spectacles
logo

Inspiration

Out-of-hospital cardiac arrests affect over 350,000 people annually in the US alone. However, less than 20% of the U.S. population actually understand how to conduct life-saving CPR, and less than 3.5% of people are actually trained. In a cardiac emergency, every second counts. But when panic sets in, precious time is lost. Seconds matter. However, less than half of the U.S. population actually knows how to conduct life-saving CPR - including us.

Inspired by our lack of first aid administration knowledge and the incredible potential AR/VR has in the medicinal world, we believed Pulse provided the most practicality and usefulness out of the simplicity that the Snapchat Spectacles created.

What it does

Through Spectacles, Pulse activates an AR-guided CPR assistant through a "pulse":

a visual overlay highlights the correct hand placement on the body
activation of "Staying Alive" when hands are in correct form to the body - guides correct compression rhythm (~100 bpm)
call to emergency services directly to 911
the entire incident will also be auto-recorded for documentation and medical review

How we built it

We built Pulse mainly through Snap's Lens Studio platform, creating custom scripts to catch audio cues and also recognize where body and hand tracking was located to perform the necessary first-aid functions.

We integrated audio pattern detection through simplified real-time audio analysis pipelines from Snapchat's microphone recognition. For the CPR rhythm, we synchronized a music playback trigger with hand placement confirmation

Given the hardware limitations, we also simulated the emergency text alert and recording functionalities as visual prompts to showcase the full potential of the system.

Challenges we ran into

We had a hard time bringing a lot of our ideas to life - although the Spectacles were extremely powerful devices, it was relatively limited compared to general software structures or smartphones. There was also a decently steep learning curve as the software of Lens Studio was very similar to Unity but with little documentation available, which was not something we were totally familiar with.

Accomplishments that we're proud of

We're extremely proud of building a working prototype of Pulse that is useful at its current state despite the limitations (and limited documentation) we faced.

What we learned

A lot! The biggest part of course was understanding the complex and intricate structure of Lens Studio, being able to make the most cohesive Spectacles project through the simplistic open-source models that were presented to us, and coming up with good use cases throughout time during the hackathon.

What's next for Pulse

In the future, Pulse aims to expand into broader emergency responses, including pulse monitoring, Heimlich maneuver guidance, and field-first-aid procedures.