Aluminium Wire Rod Property Prediction System

Overview

The particular problem was faced by the NALCO(National Aluminium company Limited) , We tried to tackle this problem in the big stages of Smart India Hackathon 2024 during the Grand Finale and achieved runner up among 130 Teams. The Aluminium Wire Rod Property Prediction System is a modern AI/ML-powered solution designed to predict and optimize the physical properties of Aluminium wire rods, such as UTS (Ultimate Tensile Strength), Elongation, and Conductivity. The system improves productivity and quality control by analyzing the critical parameters of the casting and rolling processes.

This project leverages cutting-edge technologies in machine learning, data visualization, and web development to provide actionable insights, streamline workflows, and enhance the manufacturing process.

---

Features

Key Features

• AI/ML-Powered Predictions: Predicts critical properties such as UTS, elongation, and conductivity.
• Dynamic Parameter Analysis: Analyzes parameters like casting temperature, cooling water pressure, and emulsion properties.
• Visualization: Intuitive charts and dashboards for real-time monitoring and result visualization.
• User-Friendly Interface: React-based UI for easy parameter input and interaction with predictions.
• Optimization: Provides parameter recommendations to optimize the manufacturing process.

Achievements

• Developed and integrated multiple machine learning models for property prediction.
• Built a robust backend API for efficient data handling and user authentication.
• Designed an intuitive frontend for seamless interaction and visualization.
• Enabled real-time parameter monitoring and feedback.

---

Project Structure

Backend (backend/)

The backend handles server-side operations, database management, and API integration.

• Key Files and Directories:
  • app.js: Main server file.
  • config/: Configuration files for database and authentication (db.js, passportConfig.js).
  • controllers/: Logic for handling authentication, data display, and predictions.
  • middlewares/: Custom middleware for validation.
  • models/: Database models for storing attendance, predictions, and user data.
  • routes/: API endpoints for authentication, data display, and predictions.

---

Frontend (frontend/)

The frontend provides an interactive and user-friendly interface for system interaction.

• Key Components:
  • Inputs: Forms for entering and optimizing parameters (Dashboard.jsx, Optimize.jsx, etc.).
  • Outputs: Displays prediction results and their visualizations (PredictOptimize.jsx, Visualization.jsx).
  • Realtime Dashboard: Monitors live parameter changes (RealtimeDashboard.jsx).
  • Charts: Graphical representations of data trends (LineGraphs.jsx).
  • User Guide: Step-by-step instructions for using the system (Guide.jsx).

---

Machine Learning Models (mlmodel/)

This directory contains machine learning models and APIs for prediction services.

• Key Files:
  • app.py: Flask API serving prediction requests.
  • Model Files: Pretrained .pkl files for predicting UTS, elongation, and conductivity.

---

Technologies Used

Backend

• Node.js, Express.js: For server-side logic and API handling.
• MongoDB: Database for storing user and prediction data.

Frontend

• React, Vite: For building a fast, responsive UI.
• Chart.js, CSS3: For data visualization and styling.

Machine Learning

• Python, Flask: For building and serving ML models.
• Scikit-learn, Pandas, NumPy: For data preprocessing and modeling.

---

How It Works

1. Input Parameters: Users provide critical casting and rolling parameters via the frontend interface.
2. Prediction: The backend integrates with ML models to predict UTS, elongation, and conductivity.
3. Visualization: Predictions and trends are displayed in interactive charts and dashboards.
4. Optimization: The system recommends optimized parameter configurations for quality improvement.

---

Authors

This project was developed by AluMinds Team to enhance Aluminium wire rod production using AI/ML technologies.