Use of Reidentified Vehicles to Evaluate Differential Calibration Accuracy between Weigh-in-Motion Stations

Abstract

The objective of this paper is to examine the relationships between weight and length measurements for two separate weigh-in-motion (WIM) stations based on individual vehicles that cross both stations. Technologies and methods to reidentify vehicles crossing multiple WIM stations are improving; this change increases the need for and applicability of this analysis. The analysis in this paper is based on WIM data for vehicles in Oregon that include automatic vehicle identification information for identifying the specific vehicle matches with minimal error. Although the differential calibration accuracy alone cannot be used to determine which station, if either, is accurate, it can be used to trigger further investigation or recalibration, which is useful for network-level analysis and monitoring. To evaluate the accuracy of the individual WIM station, Class 9 front axle weight (A1W) and Class 9 axle 2 to 3 spacing are investigated. Then each axle weight, gross vehicle weight, axle spacing, and bumper-to-bumper vehicle length are compared for the common vehicles crossing pairs of WIM sites with regression lines fitted to estimate the percentage difference. The best fit regression lines are observed for all attributes except for A1W and drive tandem axle spacing. The slopes of the regression lines fit for the station pairs provide an estimate of the difference in accuracy between WIM stations. This differential accuracy (calibration) information might enable agencies to perform routine field calibration at fewer sites and then monitor and adjust the adjacent sites' calibration with differential calibration based on reidentified vehicles.