Estimation of Vehicle Speed from Pavement Stress Responses Using Wireless Sensors

Abstract

Vehicle speed is a crucial input for mechanistic-based design and maintenance decision making of asphalt pavement. It can be measured using direct methods and indirect methods. Direct methods, such as video camera and inductive loops, have difficultly achieving real-time data processing and have a sophisticated installation process. Indirect methods use the mechanical responses collected by the traditional mechanical sensors to estimate speed. Although effective, these traditional mechanical sensors usually have a low survival rate and require highly sophisticated installation procedures. Recently, a new type of wireless micro-electromechanical sensor (MEMS), SmartRock (Sensor Technology Research Development and Application Laboratory, Hong Kong), has been promising in pavement research to monitor contact stress and acceleration at the particle level. This paper aims to explore the potential and reliability of using the SmartRock in acquiring vehicular speed and to evaluate the effect of loading amplitude and wandering effect (i.e., the loading location relative to the sensor or monitoring-point location) on speed estimation. In this study, SmartRock sensors were embedded in a semirigid pavement, and an accelerated pavement testing (APT) facility was used to apply loadings. Vehicle speed was estimated according to the dynamic stress responses (i.e., stress waves) measured by the SmartRock. In addition, a finite-element model was developed. After being calibrated by SmartRock stress data, this model, combined with SmartRock data, was used to evaluate the effect of load amplitude and wandering on the speed estimation. Findings indicate that SmartRock can reasonably estimate the vehicle speed using both time-domain and frequency-domain methods. The loading amplitude and wandering may have little effect on the speed estimation according to the results of this study, but future work is recommended to verify such findings.