Utilizing ADS-B and Computer Vision for Runway Status Lights
TL;DR
Evidence is provided that such an approach to runway surveillance can be effective and the proposed cost-effective surveillance methodology can provide numerous benefits to airport safety and operations.
Abstract
A technology-based runway incursion prevention system applicable to non-towered and general aviation airports and enabled by automatic dependent surveillance–broadcast (ADS-B) and computer vision is presented. The proposed system was designed to be a broadly applicable solution to prevent runway incursion accidents. A proof-of-concept surveillance prototype was developed and tested at the Purdue University Airport using simple and inexpensive hardware. The prototype was evaluated by comparing human observations with live and recorded operations for ADS-B and computer vision validation, respectively. Utilizing ADS-B, 94% of all operations observed during the study, including all operations involving ADS-B-transmitting aircraft, were detected and able to sufficiently provide timely runway status information despite being in airspace where ADS-B is not required. Meanwhile, using object detection and filtering algorithms, computer vision software designed to run on solar-powered modules was able to detect 110 out of 110 approaching aircraft while providing sufficient time to indicate potential traffic conflicts. Furthermore, the computer vision software was able to detect more than 94% of surface operations correctly. This study provides evidence that such an approach to runway surveillance can be effective. The proposed cost-effective surveillance methodology can provide numerous benefits to airport safety and operations.
