SunShiftAI

I’m Ayushmaan Bellum, a sophomore from Folsom, CA, and I built SunShiftAI to help communities save energy and reduce carbon emissions by running appliances at the cleanest, lowest-impact times. Inspired by California’s grid stress and the hidden carbon cost of timing electricity use, I learned how solar irradiance, cloud cover, and TOU pricing interact and designed a SunShift Score:

SunShiftScore[h] = 0.4 × solarScore[h] + 0.3 × priceScore[h] + 0.3 × gridScore[h]

I built the app with Flutter and Dart for iOS, Android, and web, using OpenWeather API for hourly weather, SharedPreferences for local storage, Provider for state management, and Google Gemini API for AI explanations and a chatbot. The biggest challenges were normalizing scores, sliding variable-length appliance windows, and making AI explanations beginner-friendly, but in the end, SunShiftAI transforms complex grid data into simple, actionable recommendations that anyone can follow to save energy and reduce emissions.

Built With

• dart
• flutter
• geminiapi
• openweatherapi
• provider
• sharedpreferences