Enhancing Aviation Safety: An Automated System for FOD Detection and Removal in Support Vehicle Tires

Abstract

Foreign Object Debris (FOD) detection is critical for maintaining the safety and operational efficiency of airport runways. This study presents an innovative approach utilizing advanced image processing techniques and machine learning algorithms to enhance FOD detection accuracy in tires of vehicular Ground Support Equipment (GSE). Aiming to reduce significant safety and economic risks, estimated by the FAA at $22.7 billion annually, this work focuses on debris removal from tire treads, an often overlooked area in past studies. Central to this system is a test bed with a dual-roller setup for comprehensive tire surface inspection. The detection process employs a high-resolution camera and the Segment Anything Model (SAM) for precise segmentation, followed by contrast enhancement to delineate tire grooves—the primary regions of interest (ROI). Gaussian blur and thresholding convert the images to a binary format, where contours are identified, and the longest contours, indicative of FOD-free grooves, generate a standard contour mask. This dynamic threshold mask allows for the detection of foreign objects regardless of tire type, size, or groove configuration. FOD identification is achieved through the SAM algorithm and grayscale conversion for contour analysis, with a dynamic thresholding technique that adapts to different tire sizes, enhancing detection accuracy. Upon detection of FOD, a linear actuator and a robotic servo-driven end effector are activated to remove the lodged FODs with the force generated by tire rotation. This study fills a significant gap in aviation safety and sets a new standard for airport operational efficiency, leading to safer and more efficient aviation environments.