Video
Embedded Files
Overview
Replaces manual inspection
AI-based drone inspection system that converts images into 3D models to accurately detect and locate damage on railroad bridges.
Core Competencies
Purpose
Analyzing 3D point cloud data manually is time-consuming, with low reliability and efficiency, making real-time analysis of tunnel shapes impossible. LiDAR-based automated 3D shape analysis aims to automatically analyze LiDAR data to evaluate the status of tunnel shapes in real time.
Advanced Data Modeling & Integration
Advanced Data Modeling & Integration
• 3D modeling of target facilities as reference databases
• Conversion of drone-acquired images into 3D coordinates and alignment with facility models
Intelligent Damage Detection
Intelligent Damage Detection
• Deep learning-based analysis to automatically identify cracks, corrosion, and other defects
• Mapping of inspection results onto 3D models to create a digital visual inspection map
Autonomous Flight & Operational Control
Autonomous Flight & Operational Control
• Fusion navigation algorithm for precise autonomous drone flight along pre-set paths
• Consistent and high-resolution image acquisition through pre-defined flight and imaging parameters
Expected Benefits
Precise identification of damage location and type
Significant reduction of false detections compared to conventional methods
More than 60% reduction in inspection time, enhancing efficiency
Consistent inspection quality equivalent to skilled human inspectors
Safer inspection of inaccessible or hazardous facilities such as railroad bridges
Lower risk of accidents and minimized exposure of inspectors
Certifications
Demonstration at Miho Bridge (’22.05)
Test-bed on 10 bridges (’22.08–’23.07)
Accredited testing & K-Mark certification (’23.01)
PPS Excellent Procurement designation
TRL 7
Contact
Korea Railroad Research Institute (KRRI)
Global Business Department
Dr. Hyukjin Yoon
E-mail: scipio@krri.re.kr
Page updated
Google Sites
Report abuse