About EcoVision AI
Building the next generation of intelligent waste segregation infrastructure to close the loop between consumer habits and industrial circular recycling.
Project Mission: Eliminating Recycling Contamination at the Source
Every year, human civilization generates over 2.1 billion tons of municipal solid waste. While millions of citizens genuinely wish to recycle, the reality of modern packaging is rife with ambiguity:
30%+ Contamination Rates
Well-intentioned people toss soiled pizza boxes, composite chip wrappers, and incompatible plastic resins into recycling bins, causing entire truckloads to be condemned to landfills.
Landfill Greenhouse Impact
Organic waste decomposing anaerobically in landfills is one of the single largest anthropogenic sources of methane—a greenhouse gas 28× more potent than CO₂.
Virgin Resource Depletion
Producing aluminum from recycled scrap consumes 95% less energy than mining bauxite ore. Recycling plastics saves up to 70% of crude petroleum energy.
The EcoVision AI Solution
EcoVision AI places a precision environmental sorting assistant in every pocket. By leveraging multimodal AI computer vision, the platform identifies material compositions, flags non-recyclable contaminants, outputs crystal-clear step-by-step disposal procedures, and audits avoided carbon metrics in real time.
How the Technology Works
Engineered with modern full-stack web standards, multimodal vision intelligence, and an AWS cloud architecture roadmap.
Client & Ingestion
Next.js 15 App Router with React 19. Client-side drag-and-drop image normalization, instant previews, and zero-auth local session state.
Edge API Route
Serverless route /api/analyze handles payload validation, extracts base64 streams, and checks API credentials.
Multimodal Vision AI
Google Gemini Vision model extracts resin identification codes (PET #1, HDPE #2), contamination alerts, and structured JSON output schema.
AWS Cloud Infrastructure
Cloud roadmap: Amazon S3 for encrypted image lake, AWS Lambda for image pre-processing, and DynamoDB for aggregated municipal metrics.
WeMakeDevs × AWS Environmental Hacks Hackathon
Track 03: Waste and Energy • Innovating for planetary resilience
Why Track 03: Waste and Energy?
Track 03 challenges developers to engineer high-impact technical interventions that tackle industrial and domestic waste while optimizing energy recovery. EcoVision AI addresses both fronts:
- Waste Minimization: Drives blue-bin contamination below 5% through instant point-of-disposal guidance.
- Energy Conservation: Recovers embodied energy by ensuring aluminum, plastics, and glass re-enter circular manufacturing streams.
Hackathon Compliance & Design Ethics
EcoVision AI was architected strictly according to competition requirements:
- Zero Simulated/Fake AI: If an API is offline or unconfigured, the app transparently communicates setup needs rather than hallucinating outputs.
- Frictionless UX: No Firebase or mandatory logins—inspectors can test the app instantly on any device.