DataDrive: Unified Insights for Data & Fuel Optimization

DataDrive is a full-stack platform that combines AI-powered insights with fuel optimization and data observability. It leverages machine learning models to monitor inefficiencies, predict fuel consumption, and calculate environmental impacts, while offering actionable recommendations via a dynamic and intuitive dashboard.

Features

• Predict fuel efficiency based on vehicle parameters.
• Cluster vehicles into efficiency categories using K-Means and visualize with D3.js.
• Provide SHAP-based explanations for ML model predictions.
• Real-time insights with API endpoints powered by Flask.
• Interactive 3D car model and PCA-based visualizations.
• IPFS-backed secure logging and reporting.

---

Tech Stack

Frontend

• Framework: Vite (React + TypeScript)
• Styling: TailwindCSS, D3.js
• 3D Modeling: React-Three-Fiber
• Visualization: Chart.js, D3.js

Backend

• Framework: Flask (Python)
• ML Models: scikit-learn, TensorFlow
• Job Scheduling: APScheduler

---

Local Setup

Prerequisites

Ensure the following tools are installed:

• Python (3.8 or later)
• Node.js (16.x or later)
• Yarn (or npm)

Step 1: Clone the Repository

git clone https://github.com/kushagrabharti/DataDrive.git
cd DataDrive

Backend Setup & Local Env Initialization

cd backend

# Virtual Env. for Python
python3 -m venv venv
source venv/bin/activate  # macOS/Linux
.\venv\Scripts\activate   # Windows

# Install dependencies
pip install -r requirements.txt

# Running FLASK server
python app.py

Frontend Setup & Local Env Initialization

cd frontend
yarn install
yarn vite dev