The Indiana Department of Transportation (INDOT) is using a pavement management system to identify roads for periodic maintenance and reconstruction. The present serviceability index (PSI), pavement riding comfort index, is one of the major factors in selecting roads for rehabilitation. This study searched statistically realistic models for PSI and international roughness index (IRI) correlation. Ten randomly selected subjects rated 1-mi-long test sections at three roughness levels for both concrete and bituminous pavements. Two nearly identical cars were used, and each subject rated the 20 test sections as a driver and as a front seat passenger. Each rater assigned a PSI value between 0 and 5 (0 for worst, 5 for best) and also marked whether the ride on the section was acceptable. The IRI of each test section was measured by a van equipped with noncontact laser sensors. The statistical analyses indicated that the PSI rating observations were normally distributed, the variances were homogeneous, and the position of the rater in the car was not significant. Then the average PSI ratings and IRI values of the test sections were used for model searches. Simple linear and exponential models were obtained to fit the data with r-squared-values ranging from 0.80 to 0.95. The acceptable service level IRI values were obtained by a logistic regression model using the average IRI and acceptance-rejection data of the test sections. Now INDOT can predict PSI values from collected IRI data. Using the IRI data and acceptable service level values, INDOT can identify roads for rehabilitation.


  • English

Media Info

  • Features: Figures; References; Tables;
  • Pagination: p. 27-37
  • Monograph Title: Pavement and traffic monitoring and evaluation
  • Serial:

Subject/Index Terms

Filing Info

  • Accession Number: 00667661
  • Record Type: Publication
  • ISBN: 0309055121
  • Files: TRIS, TRB
  • Created Date: Oct 13 1994 12:00AM