Skip to content

AjayRauniyar/SIH-1725

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

27 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

KaaryaDrishti - Construction Stage Monitoring Using AI

Problem Statement

Smart India Hackathon 2024 - Problem Statement ID: SIH-1675

Team Name: Visionaries United
Theme: Smart Automation
Problem Statement Title: Utilization of images for monitoring the progress of construction activities for building construction projects.

Project Overview

KaaryaDrishti is an AI-powered solution for real-time monitoring and progress tracking of construction activities using images. It enables stakeholders to automatically detect the construction stage and track progress through machine learning algorithms, reducing manual intervention and improving efficiency.

Key Features

  • Automated Stage Identification: Upload images of the construction site, and our system will automatically identify the construction phase using machine learning models.
  • Progress Tracking: Real-time tracking of project progress by comparing current and past images, with options to generate time-lapse visuals.
  • Error Detection: Alerts users about blurry or irrelevant images (e.g., images of people or vehicles) and incorrect uploads.
  • 3D Model Comparison: Integration with 3D models to compare actual progress with planned design, ensuring alignment with project goals.
  • Dashboard & Reports: Visual reports, live updates, and alerts for delays or errors via an interactive dashboard.
  • Geotagging: GPS verification ensures that uploaded images are from the correct site.
  • AR/VR Visualization: Augmented reality (AR) overlays for real-time visualization of progress on the physical site, and virtual reality (VR) for exploring current and future project stages.
  • Safety Compliance: Detection of protective equipment on workers to ensure safety standards are met.

Technology Stack

  • Programming Languages: Python, JavaScript
  • Frameworks & Libraries: TensorFlow, OpenCV, React, Node.js
  • ML Models: Convolutional Neural Networks (CNN), YOLO
  • Cloud Services: AWS, Google Cloud Platform
  • Database: MongoDB, Firebase
  • AR/VR Technologies: Unity, WebXR

Methodology

  1. Image Capture and Upload: Users upload images from the construction site.
  2. Preprocessing & Classification: Images are processed and classified according to the construction stage.
  3. Progress Analysis: Current progress is analyzed and compared against the project timeline.
  4. Error Detection: Any discrepancies or errors are flagged.
  5. Data Storage & Reporting: Processed data is stored and visual reports are generated for stakeholders.

Feasibility

  • Technical Feasibility: With 1,000-2,000 labeled images per construction stage, our model can achieve over 90% accuracy. Development time is estimated to be 1-2 months.
  • Market Feasibility: Initial development costs range from ₹70,000 to ₹3,00,000, with potential savings of up to 30% in labor costs annually for large-scale projects.
  • Operational Feasibility: The solution can reduce site visits by 80% and improve project tracking efficiency by 40%.

Impact

  • Cost Reduction: Up to 30% reduction in monitoring costs, translating to over $60M in savings annually.
  • Increased Efficiency: 80% fewer site visits and 50% faster reporting.
  • Improved Accuracy: 90% increase in monitoring accuracy with 40% fewer human errors.
  • Enhanced Transparency: Real-time updates on 300,000+ projects, leading to better project management and public trust.

Challenges & Solutions

  • Data Quality: Image quality checks and object detection filters for non-construction elements.
  • Limited Dataset: Data augmentation and transfer learning to overcome limited image datasets.
  • GPS Accuracy: ML-based site verification for accurate geotagging.

Links

License

This project is licensed under the MIT License - see the LICENSE file for details.

Releases

No releases published

Packages

No packages published

Languages