Precision Bounds of Pavement Deterioration Forecasts from Connected Vehicles

Abstract

Transportation agencies rely on models to predict when pavements will deteriorate to a condition or ride-index threshold that triggers maintenance actions. The accuracy and precision of such forecasts are directly proportional to the frequency of monitoring. Ride indices derived from connected vehicle sensor data will enable transformational gains in both the accuracy and precision of deterioration forecasts because of very high data volume and update rates. This analysis develops theoretical precision bounds for a ride index called the road impact factor and demonstrates, via a case study, its relationship with vehicle suspension parameter variances.