| ||Format||Pages||Price|| |
|PDF (192K)||12||$25||  ADD TO CART|
|Complete Source PDF (11M)||580||$109||  ADD TO CART|
The massive condition inventory for highway systems requires continual inspection of pavement surface condition to evaluate pavement performance and make management related decisions. Typically, this requires intensive data collection in short periods of time, to make long-term predictions. Automated imaging techniques have been developed to obtain visual data quickly and efficiently. The automation process creates large amounts of data, most of which are removed during processing to highlight features of interest.
The pavement management process uses the results of measurement from these automated techniques for prediction and decision making. This process is very much affected by the accuracy of the measurements and the continual loss of accuracy at each processing stage. This loss of accuracy, in the most part, has been ignored in existing management systems.
This paper describes and illustrates error estimation techniques for automated data collection processes that model errors in the imaging system, data processing, and subjective input to processing. These accuracy estimation techniques can be used to assess the quantity of data needed, the appropriate quality of data, and the frequency of data collection. Finally, this analysis can be incorporated into the pavement management process as an error diagnostic system to aid in estimation and prediction of pavement performance, maintenance and rehabilitation expenditures, and to recommend or select data collection technologies.
surface distress, errors, pavements, automated methods
Research assistant, Massachusetts Institute of Technology, Cambridge, MA
Assistant professor, Carnegie Mellon University, Pittsburgh, PA