SafeRide: DePIN Driver Safety Network 🛡️🚗

SafeRide is a Decentralized Physical Infrastructure Network (DePIN) that incentivizes safe driving using AI, IoT, and the Solana Blockchain. It detects driver fatigue and distraction in real-time, logging critical safety data immutably on-chain.

---

🌟 Key Features

• 👁️ AI-Powered Monitoring: Real-time Computer Vision (MediaPipe) detects Drowsiness (Eye Aspect Ratio) and Distraction (Head Pose).
• 🏎️ IoT Telemetry: Raspberry Pi Pico W captures vehicle dynamics like Rash Driving and Impact Detection.
• ⛓️ Immutable Ledger: Safety incidents are cryptographically signed and logged on Solana Devnet using the Memo Program.
• 📊 Live Dashboard: A modern SvelteKit UI displaying real-time driver status, health metrics, and blockchain verification.
• 💰 Token Rewards: Drivers earn points for "Safe Streaks," redeemable for insurance discounts (Mock).

---

🏗️ System Architecture

The system follows a robust "Edge-to-Chain" architecture:

1. Edge (IoT/AI):
   • Computer Vision Agent: Python + MediaPipe (runs on host/laptop).
   • Vehicle Unit: Raspberry Pi Pico W (MicroPython) sends MQTT telemetry.
2. Ingestion:
   • MQTT Broker: Eclipse Mosquitto (Docker).
   • Backend: Go (Gin) Service.
3. State & Storage:
   • Hot State: Redis (Docker).
   • Cold Storage: Solana Blockchain (Devnet).
4. Presentation:
   • Frontend: SvelteKit + TailwindCSS (Docker).

---

🚀 One-Click Demo Setup

We have containerized the entire stack for a seamless "One-Click" launch.

Prerequisites

• Docker Desktop (Running)
• Python 3.10+ (For the Computer Vision Agent)
• Webcam

Step 1: Launch the Cloud Stack ☁️

Run the following command to start the Backend, Frontend, Redis, and MQTT broker:

docker-compose up --build

Once the build completes:

• Dashboard: http://localhost:3000
• API: http://localhost:8080

Step 2: Launch the AI Agent 🦅

The Computer Vision agent runs directly on your machine to access the webcam.

First Time Setup: If you encounter any issues with the auto-setup, or if you are on a new machine, run the setup script first to install uv and dependencies.

• Windows: setup_cv.bat
• Mac/Linux: ./setup_cv.sh (Make sure to run from the project root)

Run the Agent:

• Windows: run_cv.bat
• Mac/Linux: ./run_cv.sh

⚠️ MacOS Users: You must grant Camera Permissions to your Terminal/IDE (VS Code, iTerm, etc.) in System Settings > Privacy & Security > Camera for the agent to work.

Ensure your webcam is active. The agent will start publishing "Fatigue" or "Safe" signals to the dashboard based on your face.

---

🧪 Simulation / Manual Testing

If you don't have a webcam or want to force specific states, use our simulation scripts found in scripts/windows (for .bat) or scripts/linux (for .sh):

Status Category	Script Name	Description
Driver (AI)	send_safe_driver	Resets Driver Status to Green.
	send_fatigue	Triggers Red Alert (Fatigue) & Solana Transaction.
	send_distracted	Triggers Distracted Alert.
	send_drowsy	Triggers Drowsy Alert.
Vehicle (IoT)	send_safe_vehicle	Resets Vehicle Status to Green.
	send_harsh_turn	Simulates dangerous turning.
	send_hard_braking	Simulates emergency braking.

---

📂 Project Structure

• backend/: Go (Gin) API & Solana Service.
• frontend/: SvelteKit Dashboard (Svelte 5).
• cv/: Python Computer Vision Logic.
• iot/: MicroPython code for Pico W.
• docs/: Detailed documentation & Dev Logs.

---

📜 Documentation & Resources

• Development Blog: Read about our 4-day sprint journey.
• Walkthrough Guide: Detailed feature explanation.
• Insurance Model: The math behind the safety score.

---