One Earth

Every voice, every place, speaking up for our planet

Comment

🌍 Inspiration

In a world facing escalating climate challenges, most people care — but feel helpless or disconnected from meaningful environmental action. Meanwhile, local pollution, illegal dumping, and deforestation often go unnoticed or unreported. We asked:

“What if everyday people could easily report environmental issues, learn about global sustainability, and take part in real campaigns — all in one beautiful, empowering space?”

One Earth is our answer — a platform that transforms awareness into action, built around the idea that every voice, every place, speaks for our shared planet.

💡 What it does

One Earth is a mobile-friendly web app where users can:

  • 📸 Report local environmental issues with images, tags, and optional geolocation
  • 🌐 Explore global sustainability events, climate news, and campaigns through our curated EarthSpace tab
  • 🧠 Learn about climate concepts via interactive bite-sized modules
  • ✅ Auto-generate tags and issue types using Gemini image analysis
  • 📍 Enrich reports with context like environmental impact zones or nearby regulations

It’s not just about reporting — it’s about reflection, learning, and activating real climate action from the ground up.

🛠️ How we built it

  • Frontend: React + Tailwind CSS with clean UX and tabbed layout for navigation (Feed | Report | Map | EarthSpace)
  • Database: MongoDB Atlas for storing user reports, campaign data, and Learn tab interactions
  • AI Integration:
    • Google Gemini API for:
    • Generating image descriptions and initial tags
    • Categorizing issues into types (Air Pollution, Dumping, etc.)
    • Severity scoring on a 1–5 scale
    • Future-proofing for before/after comparisons and hazardous material detection
    • Contextual enrichment: Gemini is also used to identify local environmental sensitivity (e.g., protected wetland, nearby water sources) and potential applicable regulations
  • Design Tools: Vercel v0 for frontend prototyping and iteration

🧱 Challenges we ran into

  • Designing a non-generic interaction flow that makes people want to come back even if there’s “nothing to report”
  • Making AI integration feel native and valuable, not bolted on
  • Balancing scope — we had more ambitious features like co-observation, Earth journals, and gamified climate quests, but prioritized core use cases for clarity
  • Sourcing localized environmental data with enough richness for Gemini to contextualize intelligently

🏆 Accomplishments that we're proud of

  • Built a fully working frontend with 4 functional tabs (Report, Feed, Map, EarthSpace)
  • Seamlessly integrated Gemini API in a value-driven way — auto-generating tags and enhancing each report with real-time intelligence
  • Used MongoDB effectively to store and manage real report data
  • Created a platform that feels more like a movement than just an app — deeply human, engaging, and accessible to anyone

📚 What we learned

  • AI tools like Gemini shine when you align them with real user friction, like: “I don’t know how to describe this photo” or “Where do I even start?”
  • Simplicity of tab-based design with emotionally grounded copy makes people want to engage longer
  • Sustainability tech should focus not just on awareness, but action, community, and continuity
  • Reporting alone isn’t enough — you need reflection + education + activation

🌱 What’s next for One Earth

  • 🧠 Crowd verification: Let others confirm or dispute reports (like "I’ve seen this too")
  • 📊 Severity aggregation: Auto-prioritize issues getting more validation or severe scores
  • 🧠 Vectara integration: Semantic search across Learn + EarthSpace content
  • 🌍 Global Feed with region filters and environmental health overlays
  • 🧑‍🤝‍🧑 Partner with NGOs to allow real-time volunteer & donation opportunities inside the Campaigns tab
  • 📦 Offline-first reporting for regions with poor connectivity

Built With

Share this project:

Updates