US-Ireland Partnership Programme: Bridge Health Monitoring Using Cameras and Computer Vision Methods

The University Of Central Florida Board Of Trustees, Orlando FL

Investigators

Abstract

This project is a US-Ireland partnership program pursuing research in structural health monitoring of bridges. The goal of this collaborative research project is to use cameras and video-image processing to identify loss of load carrying capacity of bridge structures. The camera images can be captured continuously with minimum human intervention. The image processing algorithm can be automated to assess significant change in structural health and load carrying capacity of the bridge. Conversion of camera and computer images for assessment of traffic loads and evaluation of load carrying capacity of the bridge are major research contributions of this project. Early detection of change in load carrying capacity will allow timely repair and thus extend life of the bridge. Investigators from three institutions, one each from three countries of Ireland, Northern Ireland (United Kingdom) and United States will collaborate to develop recognition of traffic loads, changes in vibration characteristics of the bridge, computer vision output, and algorithms to predict load capacity and remaining life of the bridge. The developed algorithm will be validated by observations of a bridge in Northern Ireland in the field. The project is funded by two divisions in the National Science Foundation, the Civil, Mechanical, and Manufacturing Innovation and International Science and Engineering. The intellectual merit of this project will be the development of algorithms that transforms the recorded video images of a bridge into damage characteristics for the bridge structure. The structural health will be assessed within the context of structural identification with input and output characterization. This development will require new adaptations and transformations that can manage computer vision that are continuous in space for structural engineering applications. The novelties in this project are: vehicle load modeling using computer vision; bridge response using image processing; image- based structural identification using input-output signals; and framework for remaining bridge life evaluation. The development will be validated with laboratory and field experiments. The international activities among the three collaborating partners in the areas of research, education, and outreach will support the intellectual contributions and broader impact of this research.