CAIT project no.: CAIT-UTC-031
Fiscal Year: 2012/2013
Status: Final
Rutgers-CAIT Author(s): Trefor Williams, Ph.D., Patrick Szary, Ph.D.
External Author(s): Frank Otero
Sponsor(s): Paco Technologies, FHWA - RITA
Recently, new data derived from non-destructive testing, structural health monitoring, and text mining of nontraditional data sources have evolved as new techniques to analyze infrastructure assets. The purpose of this research is to explore how different types of data mining algorithms can be employed to provide useful predictive information for asset management decision making. Our research will focus on the analysis of numerical data from monitoring of bridge structures, and the use of text mining and data to monitor complex infrastructure projects. The researchers at Utah State University have access to several streams of data from actual highway bridges in service currently. These bridges include a steel girder bridge in Salt Lake City, Utah a precast-prestressed concrete girder bridge in Perry, Utah and a concrete box girder bridge along I-5 near Sacramento, California. These three bridges collect data using a variety of sensors. The sensors include strain gages, accelerometers, tilt meters, temperature gages, and environmental conditions (such as wind speed/direction, precipitation, ambient temperature, etc.)