Use of Statistical Analysis, Computer Vision, and Reliability for Structural Health Monitoring

Structural health monitoring (SHM) of civil infrastructures is becoming more feasible with the help of recent developments in sensing and computing technologies being more available and affordable. The SHM system may contain various types of measurements including, but not limited to, vibration, strain and image data. In this paper, the authors provide a general discussion of the two critical aspects of SHM: assessment of the current condition and future performance prediction from their recent studies at the University of Central Florida. First, SHM data can be used to track and evaluate the current condition of the structure with the help of statistical pattern recognition algorithms and computer vision techniques. Statistical analysis of these types of data can provide rapid extraction of information about the changes in structural behavior whereas the use of the computer vision technologies in a monitoring system offers to detect events visually. Subsequently, the available information obtained can be used for decision-making about the future performance of the structure. Prediction of the future performance is a very crucial step in better managing the life cycle safety, serviceability and costs.