SMART INDIA HACKATHON

2024
Problem Statement ID- 1725
Problem Statement Title- Utilization of images
for monitoring of progress of construction
activities for building construction projects.
Theme- Smart
Automation PS
Category- Software
Team ID- 387
Team Name: Aanya
 Team
InfraSna
LINK TO PROTOTYPE:
Aany Click Here

Idea/Solution/ p
Prototype Innovation & Uniqueness
Automated Construction Monitoring: machin learningto
automate the Utilize of e via 1.Street View: Offers street and site views for
monitoring analysis. construction site progres image comprehensive progress assessment.
s
Stage Identification: Determine the specific stage of construction 2.Raw Materials Prediction: Predicts and manages materials
(e.g., foundation, superstructure) from site images. to optimize resource use and minimize waste.
User-Input Driven: Software accepts user inputs on the type of
activity to guide the analysis process.
3.Budget Management: Monitors and manages budgets
across multiple projects.

Solution Flow of construction activities & ML model

Approach used -
Remote Monitoring for Enhanced Oversight: The software enables
ULBs, state, and central agencies to remotely monitor
construction progress through images, minimizing the need for field
visits.
Efficient Tracking for Timely Project Completion: It ensures efficient
and timely tracking of construction projects, enhancing oversight
and reducing delays.
Interactive Dashboard: Centralized dashboard for tracking project
progress
and downloading detailed reports.
2
 Team
Aany TECHNICAL
a
APPROACH
ARCHITECTUR
INPUT E TECH STACK
Blueprint of Site React JS

3D View Raw Materials

No. of Buildings Site Created Activities

Current Stage of
Construction
Structural Super Electrical &
Foundation Interiors Facade
Works Structure Plumbing
Budget
METHODOLOGIES
Select Activity
Prediction YOLOv5 (Foundation Stage): Detects foundation-related
Each activity uses
Pannellum (WebGl)
different
algorithms
machine
to analyse
learning
the
objects like trenches and concrete mixers to confirm
Generates 360 degree
image and predict the
progress of the construction Upload Image street view
uploaded images
using
foundational work.
activity

Mask R-CNN (Super-Structure Stage): Segments

structural elements like beams and walls, ensuring
Model’s Used progress in building the main structure.
Foundation Super Electrical & Interiors Facade ResNet-50 (Facade and Finishing): Classifies images to
Structure Plumbing
verify facade work like window installations and exterior
YOLO V5 Mask R-CNN DeepLabV3 EfficientNet ResNet-50
finishes.
EfficientNet (Interior Work): Identifies interior features
such as flooring and ceiling work to track interior progress.
DeepLabV3+ (Electrical, Plumbing): Segments pipes and wires
to confirm correct electrical and plumbing installations.
3
 Team
Aany FEASIBILITY AND
a
VIABILITY
Potential Overcomin
Feasibility
Challenges g
Analysis
& Risks Challenge
Technological Feasibility: Utilize Accuracy Concerns: Ensuring Diverse s Model Training:
s advanced AI and precise analysis across diverse Continuously train models with a wide
machine
technologies forlearning
precise image construction environments and stages. range of datasets to improve
analysis. Image Quality Issues: Variations in accuracy.
Integration
be with Existing
seamlessly Can
Systems:with
integrated image quality and potential Quality Assurance: Implement
existing construction management obstructions that may affect analysis. strict quality checks for images before
tools. Data Privacy Concerns: Handling analysis.
sensitive
construction requires stringent
Scalability: Designed to scale User Training: Provide training to
data
privacy measures, data posing
easily through cloud-based users for accurate input and image
infrastructure, accommodating ensuring compliance with global selection.
multiple projects across different data
protection challenges in
locations simultaneously. regulations.

Revenue Sources
Subscription Fees: This model generates recurring revenue by charging customers monthly or Link to
yearly for access to the software.
Consultation and Customization: This revenue stream involves charging one- t ime fees for
Buisness Model
providing consulting services.
Data Analytics as a Service: Off ering premium data analytics and insights as an add- on service
Click Here
creates an additional revenue stream

4
Team
Aany IMPACT AND
a
BENEFITS
POTENTIAL IMPACTS BENEFIT
Target Audience S SOCIAL
ULBs & Government Agencies:
Streamlined processes, improved monitoring, reduced Enhance transparency and accountability
costs, data-driven decisions. d in construction
projects.
Construction Companies: stakeholder
Greater engagement and feedback opportunities
Increased efficiency, coordinatio reduce s. for
better compliance and n, d ECONOMIC
inspection costs.
Cost savings from fewer site visits and efficient
Residents & Communities: resource allocation.
Greater transparency, more engagement, fewer Potential for increased investment in infrastructure.
disruptions.
ENVIRONMENTAL
Environmental Agencies:
Lower carbon emissions, betterenvironmental Reduced carbon footprint from minimized travel
compliance tracking. for inspections.
Decrease in paper-based documentation and waste.

5
Team
Aany RESEARCH AND
a
REFERENCES
DATASET
Data Extracted from google images

RESEARCH PAPERS
Zollmann, S., Hoppe, C., Burgess, J., Sutter, C., Koch, C., & Lindner, P. (2022).
"Utilizing UAVs and Photogrammetry for Progress Monitoring in Building Construction."
Automation in Construction. Retrieved August 18, 2024, from
https://www.sciencedirect.com/science/article/pii/S0926580520301888
"Sparse R-CNN: End-to-End Object Detection With Learnable Proposals". Peize Sun, Rufeng Zhang, Yi Jiang, Tao Kong,
Chenfeng Xu, Wei Zhan, Masayoshi Tomizuka, Lei Li, Zehuan Yuan, Changhu Wang, Ping Luo; Proceedings of the IEEE/CVF
Conference on Computer Vision and Pattern Recognition (CVPR), 2023, pp. 14454-14463

Existing Systems
Global Time-Lapse Solutions: Their system uses high-resolution time-lapse cameras and AI analytics for real-time job
site tracking, automated safety checks, and risk management. It integrates with platforms like Procore and BIM for
seamless data sharing and enhanced construction analytics(Global Time Lapse).
EasyFlow's Computer Vision Solutions: EasyFlow uses computer vision for construction tasks like safety
inspections, process optimization, and detecting structural defects. It applies deep learning to find issues like cracks
and supports automated measurements for project planning and safety management(EasyFlow).