GrowEasy - Urban Gardening

• 

  GrowEasy

Inspiration

Lockdowns turned our balconies into makeshift jungles—yet most of us still
 killed the basil. We wanted a no-jargon, no-stress way to help city-dwellers grow food and greenery in tight spaces while nudging them toward more sustainable habits. GrowEasy was born to merge AI plant care, gamified motivation, and ultra-simple UX into one pocket companion.

What it does • Snap & Diagnose: Take a photo; our on-device model detects species and common issues (pests, yellowing, nutrient deficiency). • Smart Care Calendar: Auto-generates watering, fertilizing, and repotting tasks; syncs across devices and calendars. • Eco-Score & Badges: Completing tasks earns points, levels, and shareable badges to keep motivation high. • Offline-First Logging: Record tasks anywhere; Supabase sync reconciles once you’re online. • One-Tap Sharing: Export stunning before/after shots to social media with branded overlays.

How we built it • Frontend: React Native (Expo) for rapid cross-platform delivery; Tailwind-style tokens via NativeWind. • Backend: Supabase (Postgres + Auth + Realtime) for user data, plant library, and achievement tracking. • AI Layer: Low-code n8n workflows pipe each user photo to the OpenAI Vision API for one-shot species recognition and disease diagnosis. • Notifications: Expo Push + cron jobs in Supabase Edge Functions for daily reminders. • Dev Workflow: GitHub Actions CI/CD → Expo EAS builds → TestFlight Internal.

Challenges we ran into 1. Dataset quality: Many public plant images were mislabeled—manual curation took hours. 2. Model size vs. speed: Keeping inference under 150 ms on older phones forced aggressive pruning and quantization. 3. Calendar logic: Accounting for climate zones and daylight savings in recurring tasks was trickier than expected. 4. Motivation psychology: Finding the right reward frequency to feel fun, not spammy, required multiple user tests.

Accomplishments that we’re proud of • End-to-end prototype shipped in under 72 h. • App size under 35 MB, fully functional offline. • First-time users completed an average of 3 care actions/day during beta.

What we learned • How to prune and quantize TensorFlow models for React Native without breaking WebGL. • Supabase Edge Functions can replace a traditional Node API for many real-time use cases. • Micro-copy and subtle haptics dramatically boost task completion rates. • Hackathon speed demands ruthless scope management—less really is more.

What’s next for GrowEasy - Urban Gardening 1. AR “Plant-Vision” to suggest optimal placement based on sun exposure. 2. Sensor Integration: Bluetooth soil-moisture probes for real-time alerts. 3. Community Challenges: Monthly “Zero-Waste Herb” quests with leaderboard prizes. 4. Marketplace: Curated seeds, pots, and eco-friendly supplies delivered by local partners. 5. Open API: Let other green-tech apps plug into our care calendar and achievement system.

Built With

• expo.io
• react-native
• supabase