Construction RoboticsActive

Inspection - Legged Robot and ML for Window Frame Inspection

Legged robot system with pan-tilt-zoom camera using computer vision and deep learning to inspect window frames and detect surface defects

legged-robotwindow-inspectionconstructioncomputer-visiondefect-detectionquality-control
Inspection - Legged Robot and ML for Window Frame Inspection
Inspection - Legged Robot and ML for Window Frame InspectionInspection - Legged Robot and ML for Window Frame InspectionInspection - Legged Robot and ML for Window Frame InspectionInspection - Legged Robot and ML for Window Frame Inspection

Project Overview

Construction quality control faces significant challenges in window frame inspection, where manual assessment on unfinished floors with uneven surfaces leads to inconsistent evaluation, safety risks for inspectors, and costly defects discovered only after project completion. Our research addresses this critical construction industry problem by developing a legged robotic platform equipped with advanced computer vision and machine learning capabilities for autonomous window frame inspection and defect detection. The core technical innovation combines legged locomotion for stable navigation on challenging construction terrain with a sophisticated pan-tilt-zoom camera system and adaptive lighting that captures high-resolution surface imagery from multiple angles. Our deep learning-based defect detection algorithms automatically identify scratches, dents, misalignments, and surface imperfections with accuracy exceeding human inspectors, while SLAM navigation and precise localization enable comprehensive geo-tagged documentation that integrates seamlessly with construction management systems.

This technology transforms construction quality assurance by replacing subjective, time-consuming manual inspections with objective, comprehensive automated assessment that dramatically improves accuracy and consistency while reducing safety risks. The system enables real-time quality monitoring during window installation, automated compliance verification against building standards, and detailed digital documentation for warranty and lifecycle management. Applications extend beyond window inspection to comprehensive building envelope assessment, façade quality control, and preventive maintenance programs that identify potential issues before they become costly problems. The broader impact includes significant improvements in construction quality standards, reduced rework costs, enhanced worker safety, and the development of data-driven quality control protocols that can be standardized across the industry. Our team brings expertise in construction robotics, computer vision, and quality assurance systems, actively seeking partnerships with construction companies, window manufacturers, building inspectors, and technology firms to revolutionize this critical aspect of construction quality control.

Team Members

PKS
Professor Kenji Shimada
JV
Jorge Vasquez

Project Details

Started

February 28, 2023

Category

Construction Robotics

Status

Active

Interested in Joining?

We're always looking for passionate researchers to join our team.

Contact Us