Inspiration

We were inspired by how easy it is to get distracted while working on a computer and how passive most productivity tools are. Notifications, reminders, and timers are easy to ignore. We wanted to create something that physically forces accountability, something you can’t just dismiss and keep procrastinating.

What it does

Get Back to Work is an interactive productivity system that detects when you’re distracted and reacts immediately. The app analyzes screen captures using AI to recognize distractions. When detected, a cardboard man named Charlie physically slaps you with a servo-powered arm until you verbally apologize. An LCD display reinforces the message by repeatedly showing “GET BACK TO WORK.”

Text-to-speech and speech-to-text allows the system to talk to the user and listen for an apology, turning productivity into a comedic and interactive experience.

How we built it

AI & Software:

  • Desktop app written in Python
  • OpenCV, OpenRouter, and Google Gemini was used for detecting distractions from screen captures
  • ElevenLabs AI for text-to-speech and speech-to-text, allowing user interaction

Hardware:

  • An Arduino-controlled servo motor powering the cardboard arm
  • An LCD display to show messages and feedback
  • A menacing cardboard man (Charlie) made from materials scavenged from the hackathon space

Integration

  • Serial communication connects the AI detection system to the Arduino, triggering physical actions in real time

Challenges we ran into

Synchronizing software-based AI detection with real-time hardware actions was challenging, especially managing serial communication and avoiding repeated triggers. We also ran into issues with hardware reliability, servo timing, and finding materials to build Charlie. Additionally, balancing responsiveness without making the system overly aggressive required careful tuning.

Accomplishments that we're proud of

We successfully built a system that:

  • Combines AI, speech processing, and physical interaction
  • Reacts in real time to user behavior
  • Turns productivity enforcement into a humorous experience rather than a boring reminder

Most importantly, we created a fully functional hardware–software loop within a limited time frame of 24 hours.

What we learned

We learned that being resourceful can get you far. We built a working hardware prototype on from borrowed motors, sharpie markers, cardboard chip boxes and wooden forks from Steve's Poke Bowl.

We learned that building something ambitious in a short time means constantly changing plans, fixing mistakes, prompting Gemini, and having a positive attitude when things break at the worst possible moment.

What's next for Get Back to Work

Next, we want to improve distraction detection accuracy, add customizable sensitivity levels, and introduce more physical feedback options. We also plan to refine the voice interaction, add usage analytics, and explore safer but equally effective alternatives to slapping.

Built With

Share this project:

Updates