SYMPOSIA PAPER Published: 01 January 1985

Accuracy of Calibrated Roughness Surveys


The accuracy of road roughness measurement that can be achieved by calibrated response-type measuring systems is evaluated using Mays meter data from a major road costs study in Brazil. Variations in the vehicle response over time, quantization and hysteresis errors in the sensor, loss of damping due to high shock absorber temperatures on unpaved roads, repeatability across runs, reproducibility across different vehicles over an extended period, and variations in the reference profilometer system are quantified. The measurement error is typically 14% (68th percentile) across the calibrated vehicles and across time. The author shows that this error is approximately four times greater than the typical half-year increases in roughness observed on highway pavements. Studies to develop roughness prediction models, therefore, require a period of at least four years of observations. More accurate measurement methods, such as profilometry, are recommended for future research on pavement deterioration. Calibrated response-type systems, however, are adequate for monitoring roughness for regular pavement management and for measuring unpaved roads.

Author Information

Paterson, WDO
The World Bank, Washington, DC
Price: $25.00
Contact Sales
Reprints and Permissions
Reprints and copyright permissions can be requested through the
Copyright Clearance Center
Developed by Committee: E17
Pages: 66–88
DOI: 10.1520/STP34595S
ISBN-EB: 978-0-8031-4948-9
ISBN-13: 978-0-8031-0428-0