Application of Collaborative Robotics and Artificial Intelligence for Adaptive Path Planning in Welding Inspection
Faculty Mentor
Hossein Taheri
Location
Russell Union Ballroom
Type of Research
On-going
Session Format
Poster Presentation
College
Allen E. Paulson College of Engineering & Computing
Department
Mechanical Engineering, Manufacturing Engineering
Abstract
Structural inspections are essential for ensuring the safety and longevity of critical infrastructure, yet conventional nondestructive testing methods often involve labor intensive procedures, limited accessibility, and predefined scanning patterns that may not optimally capture complex or evolving defects. Building upon an ongoing work integrating robotic inspection, machine vision, and artificial intelligence for welding inspection, this study advances an automated inspection framework centered on collaborative robotic (cobot) systems and ultrasonic testing. The proposed project employs robotic controls to perform precise and repeatable ultrasonic testing, while also utilizing AI driven data analysis implemented in MATLAB for real time scan interpretation, flaw detection, and sizing. Unlike traditional predefined inspection approaches, the system incorporates adaptive robotic path planning based on calculated flaw locations, allowing the robotic platform to dynamically modify its scanning trajectory as defects are identified. This approach enables targeted scanning in regions of interest rather than pure reliance on predefined patterns, improving inspection efficiency and increasing the probability of detecting critical welding flaws. In addition, the AI framework considers material properties and ultrasonic wavelength selection relative to target flaw size, enabling optimization of inspection parameters for improved detection sensitivity and sizing accuracy. By accounting for ultrasonic wave behavior within different materials and weld geometries, the system enhances accuracy across varying structural conditions. The integration of conventional ultrasonic testing, collaborative robotics, and MATLAB based AI analysis demonstrates strong potential to improve repeatability, consistency, and reliability in welding inspections while reducing operator dependency and field related challenges. This research presents a pathway toward intelligent and autonomous nondestructive evaluation systems capable of real time decision making, adaptive inspection strategies, and scalable deployment for infrastructure quality assurance and structural health monitoring.
Program Description
.
Start Date
4-23-2026 10:00 AM
End Date
4-23-2026 12:00 PM
Recommended Citation
Oubre, Julia, "Application of Collaborative Robotics and Artificial Intelligence for Adaptive Path Planning in Welding Inspection" (2026). GS4 Student Scholars Symposium. 82.
https://digitalcommons.georgiasouthern.edu/research_symposium/2026/2026/82
Application of Collaborative Robotics and Artificial Intelligence for Adaptive Path Planning in Welding Inspection
Russell Union Ballroom
Structural inspections are essential for ensuring the safety and longevity of critical infrastructure, yet conventional nondestructive testing methods often involve labor intensive procedures, limited accessibility, and predefined scanning patterns that may not optimally capture complex or evolving defects. Building upon an ongoing work integrating robotic inspection, machine vision, and artificial intelligence for welding inspection, this study advances an automated inspection framework centered on collaborative robotic (cobot) systems and ultrasonic testing. The proposed project employs robotic controls to perform precise and repeatable ultrasonic testing, while also utilizing AI driven data analysis implemented in MATLAB for real time scan interpretation, flaw detection, and sizing. Unlike traditional predefined inspection approaches, the system incorporates adaptive robotic path planning based on calculated flaw locations, allowing the robotic platform to dynamically modify its scanning trajectory as defects are identified. This approach enables targeted scanning in regions of interest rather than pure reliance on predefined patterns, improving inspection efficiency and increasing the probability of detecting critical welding flaws. In addition, the AI framework considers material properties and ultrasonic wavelength selection relative to target flaw size, enabling optimization of inspection parameters for improved detection sensitivity and sizing accuracy. By accounting for ultrasonic wave behavior within different materials and weld geometries, the system enhances accuracy across varying structural conditions. The integration of conventional ultrasonic testing, collaborative robotics, and MATLAB based AI analysis demonstrates strong potential to improve repeatability, consistency, and reliability in welding inspections while reducing operator dependency and field related challenges. This research presents a pathway toward intelligent and autonomous nondestructive evaluation systems capable of real time decision making, adaptive inspection strategies, and scalable deployment for infrastructure quality assurance and structural health monitoring.
