Factors Contributing to Differences Between Profiler and the International Roughness Index

One of the data elements included in a pavement management system is the roughness of each highway segment. State highway agencies use inertial profilers to collect profile data of the pavements and then use these data to compute the International Roughness Index (IRI) of each highway segment. An issue concerning the use of inertial profilers is whether the IRI values computed from the profile data are accurate. This issue is usually investigated by the establishment of several test sections, the collection of profile data on these sections with a reference device as well as the profiler, and then a comparison of the IRI values. In many cases, a given profiler will have excellent agreement in IRI with the reference device at some sections but noticeable differences in IRI at others. Data collected for Long-Term Pavement Performance profiler comparison studies were analyzed to identify factors that could contribute to differences between profiler and reference device IRI. The analysis indicated that the following factors can contribute to differences in IRI values: sampling qualities of the reference device, variability in the path followed by the profiler, averaging effects of profiler data, reference device data errors, IRI computational procedure of the profiler, and operational procedures during profiler data collection. Procedures that can be followed to minimize the difference in IRI and the application of roughness profiles to identify the cause for such differences are presented.