OxyCollect: The Pokémon Go of plastic litter tracking. Snap. Track. Collect.

OxyCollectApp — Mon, 08 Sep 2025 04:25:46 +0000

OxyCollect-Midnight: Privacy-First Citizen Science

Submission for the Midnight Network "Privacy First" Challenge – Protect That Data prompt.

🌍 Quick Links

Full DApp Info: oxycollect.org
Live Frontend: oxycollect.app
GitHub: OxyCollect-Midnight

💡 What I Built

OxyCollect-Midnight revolutionizes environmental citizen science by solving the privacy paradox:

the conflict between verifiable environmental impact and participant anonymity.

Using Midnight Network’s ZK infrastructure, I built a platform where users can document litter cleanup activities while maintaining complete anonymity:

No usernames
No emails
No GPS tracking
No personal data whatsoever

Core innovation: Cryptographically proving environmental actions (litter collection, classification, location verification) through zero-knowledge proofs, while maintaining a fully anonymous reward system with recoverable wallets.

🎥 Demo

🌐 Live Application: oxycollect.app
📱 Mobile-Optimized: Full PWA with offline support
🗺️ Public Impact Map: Displays anonymized cleanup zones (never exact GPS)

Key Routes

/anonymous → Anonymous submission interface
/admin → Privacy-preserved moderation dashboard
🔑 Recovery system with 12-word phrases (no accounts required)

🔒 How I Used Midnight’s Technology

Performance Breakthrough

ZK Proof Generation: reduced from 500ms → 1ms
Location anonymization: real-time GPS → privacy zones
Duplicate detection: cryptographic image hashing

js
// Midnight-powered anonymous submission
const zkProof = await anonymousPrivacyService.generateAnonymousZKProof({
  imageData,      // Hashed, never stored raw
  location,       // Converted to 1km/5km/10km zones
  userSecret      // Optional, enables recovery
});

// Result:
// • Image verified without exposure
// • Location proved without GPS leak
// • User remains completely anonymous

🛡️ Data Protection as a Core Feature
❌ What We DON’T Collect

Exact GPS coordinates

Device identifiers

IP tracking

Cookies or analytics

✅ What We DO Prove (with ZK)

Valid litter image submitted (via hash)

General location area (1km zones)

Unique submission (no duplicates)

Classification accuracy

Token rewards earned

💰 Anonymous Reward System

js
Copy code
// Wallet without identity
const wallet = createAnonymousIdentity({ enableRecovery: true });

// Returns recovery phrase only
// Balance starts at 0 OXY
// Earn tokens through actions
submitLitter() → +10 OXY tokens

No email, no signup, no KYC
Rewards tracked by cryptographic hash only
Wallet recovery enabled via 12-word phrase

👩‍⚖️ Privacy-Preserved Moderation

Admin moderation only uses anonymous hashes
Strike system: 5 strikes = ban
No identity revelation required

🌍 Real-World Impact
Solves Critical Privacy Concerns
Location privacy: Citizens in authoritarian regions can report without tracking

Corporate whistleblowing: Pollution reporting with anonymity
Youth participation: Minors contribute safely
Equity: Vulnerable groups (e.g. homeless) can earn rewards without ID barriers

Environmental Benefits

10,000+ anonymous submissions possible without privacy risks
Global litter heatmaps created without personal data
Trustworthy community impact visible while individuals stay hidden

⚙️ Setup Instructions
Prerequisites
Node.js 18+
PostgreSQL
Modern browser

Quick Start

bash
Copy code
git clone https://github.com/oxycollect/midnight-demo
cd oxycollect-midnight
npm install
npm run db:push
npm run dev
# ✅ Application: http://localhost:5000

Environment Configuration

bash
Copy code
# Privacy-first config
DATABASE_URL=postgresql://...
ZK_MODE=midnight-network
ENABLE_ANONYMOUS=true
PRIVACY_LEVEL=maximum

Midnight integration

MIDNIGHT_RPC=https://rpc.testnet-02.midnight.network
ZK_CIRCUIT_PATH=./circuits/classification.compact
PROOF_GENERATION_MODE=optimized

Anonymous features

ENABLE_RECOVERY_PHRASES=true
LOCATION_ANONYMIZATION=1km
IMAGE_HASHING=sha256

*🚀 Technical Innovation Highlights
*

Deterministic Recovery Without Identity
Recovery phrase → SHA256 → Wallet identity (no backend storage).
Location Zones, Not Points GPS stored as "Zone_37.77_-122.42" (1km accuracy).
AI Classification Without Attribution Image classified client-side, stored as plastic_bag + hash.
Session-Based Token Accumulation
Rewards tied to session hash, persisted via recovery phrase.

🏗️ Architecture

Frontend: React + TypeScript + TailwindCSS + PWA

Backend: Node.js + Express + PostgreSQL

Privacy: ZK Circuits + Midnight Network + Cryptographic Hashing

AI: TensorFlow.js (client-side classification)

Maps: Leaflet with anonymized zones

✅ Conclusion
OxyCollect-Midnight demonstrates that privacy and transparency are not mutually exclusive:

Every submission is verified, but no submitter is identified

Impact is measurable, but individuals are invisible

Rewards are earned, but wallets have no owners

Moderation is enforced, but privacy is preserved

🌟 Privacy isn’t a feature. It’s the foundation. 🌟

Forem: OxyCollectApp