1665115 - Leveraging Computer Vision for Transportation Infrastructure Health Monitoring

Transportation infrastructures, particularly bridges, are subject to aging, contamination, fatigue, unexpected service excitations, extreme environmental conditions, and accidental loads. Comprehensive and timely structural evaluations ensure structures' safety and adaptability to meet contemporary service performance standards. Departments of Transportation continuously seek near-real-time, accurate data about the health and condition of structures to guide maintenance and replacement decisions. However, the current challenge lies in the lack of advanced automated techniques for continuous Structural Health Monitoring (SHM). This research explores the role of computer vision in SHM of such infrastructures. It details how computer vision techniques can extract vital vibration data from structures to report potential vulnerabilities. The latest advancements in computer vision, its practical applications in SHM, and the barriers to its widespread adoption will be presented in this study. A specific focus will be given to the extraction displacement responses of full-scale structures. The capabilities of video displacement tracking algorithms to discern inherent dynamic properties within a structure and thereby enable deeper structural behavior insights will be highlighted. In addition, the pros and cons of integrating computer vision into SHM will be addressed. The presentation will conclude with the acknowledgment of existing challenges with an emphasis on the promising potential of computer vision in enhancing SHM and provides details on advanced methodologies for assessing transportation infrastructure conditions.