Stop! Don't Go On!

Stop! Don’t Go On senses when you start gooning, challenges your focus, and keeps you accountable with alerts, snapshots, and a "splash" of reality.

Comment

Inspiration

As avid Go Oners who needed to lock in for November, we realized that no current solutions for quitting go oning was actually daunting enough for us to care. That’s why we built Stop! Don’t Go On, a more accessible and high-stakes way to quit go oning once and for all.

What it does

Stop! Don't Go On detects and tracks you through computer vision so that when you go on, it will know. And it will stop you.

Once the app detects gooning audio, it flashes an alarm while the elixir of life fills the screen. Then, Stop! Don’t Go On will spray you with our Professional Sprayer™ and prompt you with a math question, which you will have to answer within 6-7 seconds. If the math is not completed in time, an alert will be sent to an email contact of your choice with a live picture of you stroking it. Use your mom, your wife, or your kids as your contacts for maximum effectiveness!

How we built it

Hardware

Arduino Uno is connected to a servo motor. A rope is attached to the spray bottle and the servo motor so that when the motor rotates, the spray bottle sprays. A separate motor is connected to the base. Both motors are connected to batteries through a breadboard.

Software

The app listens for gooning by sampling microphone input for rhythmic clapping. Python OpenCV backend tracks your face, making sure the camera and the spray bottle stays locked on the user at all times. If the user is unable to solve the math problem in time, a photo is captured mid-goon, which is handed off to EmailJS so every address the user adds will receive an alert with the photo.

Challenges we ran into

One of the biggest challenges was connecting the frontend (Next.js + TypeScript) to the Python backend. Managing communication between these two different environments posed a significant hurdle. Another one of our roadblocks was that the spray bottle was too stiff, so the motor we had did not generate enough force to squirt us. This was largely due to difficulty with converting the rotational force of the motor to a pulling force on the spray bottle.

Accomplishments that we're proud of

We’re proud about implementing OpenCV into our project seamlessly, successfully connecting multiple different environments, and finally getting Python, Node.js, and Typescript to work together in harmony through a plethora of technical roadblocks. Additionally, we are proud of the fact that we were able to bridge the gap between our software and our hardware, being successful in our first time working with Arduino kits. As people newer to hardware, the satisfaction with every fixed bug made the whole process worth it.

What we learned

We learned that backend development is far more complex than it seems, and that AI tools alone can’t magically solve integration or communication issues between systems. Debugging, managing concurrency, and maintaining reliable data pipelines all require a deep understanding of both sides of the stack.

What's next for Stop! Don't Go On!

We plan to implement facial expression detection, to allow for more accurate goon detection. Additionally, we plan to implement user login to encourage saved sobriety streaks using a saved database that collects data per user. Furthermore, a more high-pressure squirting system can be implemented in the future for better range and accuracy.

Share this project:

Updates