Skip to content
Notifications You must be signed in to change notification settings

emma-w1/PennApps

Repository files navigation

Athena Award Badge

Sollis

https://docs.google.com/presentation/d/1kRaQyKPKHA3sn7sRTVwOSdsWiQJeeMlTRAeg5QPmgUQ/edit?usp=sharing (see the hardware slide for specifics)

IMG_1729 IMG_1762 Screenshot 2025-10-02 at 3 26 41 PM (2)

A wearable armband and iOS app that helps you track UV exposure in real time and receive personalized sun safety recommendations.

Inspiration

Too much sun exposure is the leading preventable cause of skin cancer, but most people don’t realize when they’re at risk. We wanted to create a simple, wearable way to monitor UV exposure and make sun safety more accessible—especially across different skin tones.

What It Does

Sollis is a wearable + mobile system that:

  • Uses a UV sensor to measure sunlight intensity.
  • Processes data in a backend to calculate a risk score.
  • Leverages Cerebras AI to generate personalized insights (e.g., when to reapply sunscreen or step into the shade).
  • Displays results in a Swift-based iOS app with actionable guidance.

How We Built It

Hardware: Arduino board + UV sensor programmed in C++ to capture exposure data. Backend: Python service that processes UV readings, calculates risk, and communicates with the Cerebras API. Mobile App: An iOS app written in Swift to visualize exposure data and deliver recommendations. Database & Cloud: Firebase + GitHub integration for data handling and collaboration.ion.on.

Challenges We Ran Into

  • Compiling issues while integrating Arduino code into the broader system.
  • Designing a UV-to-risk formula that was both accurate and meaningful.
  • Getting each layer of the stack (hardware → backend → AI → app) to communicate reliably under hackathon time constraints.

Accomplishments We’re Proud Of

  • Built a full-stack prototype in under 36 hours.
  • Successfully integrated hardware, backend logic, AI inference, and mobile UI into a working system.
  • Created a user-friendly experience that turns raw sensor data into clear, actionable insights.

What We Learned

  • How to connect hardware, backend, and frontend into a cohesive system.
  • Debugging integration issues under tight deadlines.
  • Refining formulas for real-world health accuracy.
  • How AI can make technical health data feel personal and actionable.

What’s Next for Sollis

  • Sleeker, more comfortable wearable design.
  • Better sensor calibration for improved accuracy.
  • Expanded inclusivity by tailoring recommendations across a wider range of skin tones.

Built With and BY

Arduino, C++, Cerebras, Firebase, GitHub, Python, Swift By: Emma Wong, Adishree Das, Vijeta Garg, & Skyler HallHalll

Athena Award Badge