EnviroLens – Smart AI Assistant for Environmental Awareness

An AI assistant that analyzes images and videos to detect environmental issues like pollution, waste, and deforestation — helping people take action for a cleaner planet.

Comment

Inspiration

Environmental damage is happening around us every day—polluted rivers, burning trash, deforestation—but most people don’t know how to report it or what actions to take. I wanted to build something that empowers anyone to identify environmental issues instantly using only a photo or video. The inspiration came from the idea that AI shouldn't just generate images—it should help solve real-world problems.

What it does

EnviroLens is a multimodal AI assistant that:

  • Detects environmental issues (pollution, waste, deforestation, fires) from images, videos, and audio descriptions
  • Generates an easy-to-understand environmental impact report
  • Suggests practical eco-friendly solutions
  • Visualizes a Before vs After restoration preview using AI

Its goal is to make environmental awareness accessible—no technical knowledge required.

How we built it

  • Frontend: React + TypeScript + Tailwind CSS
  • Backend: Node.js + Express
  • AI Engine: Google Gemini (multimodal for image/video/audio analysis)
  • Deployment: Docker + Google Cloud Run

Workflow:

  1. User uploads media (image/video/audio)
  2. Backend sends content to Gemini API
  3. AI returns detection + recommendations
  4. Frontend displays interactive reports + restoration preview

Challenges we ran into

  • Handling large video/image uploads and keeping performance fast
  • Structuring prompts to get consistent environmental insights from Gemini
  • Integrating multimodal input (image + text + audio) into a single analysis pipeline
  • Deploying the full app on Cloud Run with environment variables and Docker

Accomplishments that we're proud of

  • Built a working end-to-end multimodal AI system
  • Achieved accurate pollution/deforestation detection from real images
  • Deployed the project to production on Google Cloud Run
  • Received positive feedback from people in environmental clubs/NGOs

What we learned

  • Prompt engineering for complex AI workflows
  • How to build scalable backend infrastructure with Docker + Cloud Run
  • How multimodal AI models interpret and reason using images + video + audio
  • The power of AI when used for real-world social impact projects

What's next for EnviroLens – Smart AI Assistant for Environmental Awareness

  • Build a mobile app for environmental reporting in the field
  • Add a global map dashboard showing user-submitted environmental issues
  • Partner with NGOs and communities to use the tool for real projects
  • Automatic report submission to authorities from the app (in exploration)

Built With

Share this project:

Updates