Pavement Condition Monitoring by Fusing Data from a Mobile Multi-sensor System

Presenter: Salar Shahini Shamsabadi

Research Category: Interdisciplinary Topics, Centers and Institutes
PI: Ralf Birken

Statistics suggest Federal Government spends about 50 billion dollars each year on roadways, a considerable portion of which goes to road rehabilitation and maintenance. This number drops significantly if cities conduct preventive maintenance, the axiom of “pay a little now or a lot more later”. Current pavement inspection methods, typically done manually by inspection teams, do not provide sufficient information to fulfill this purpose. Consequently, an affordable framework that offers a continuous network-health monitoring of roadways is yet to be defined. Within these paradigms, a multi-modal mobile sensor system has been developed and mounted on a prototype vehicle. This system includes a microphone, a tire pressure sensor, an axle-accelerometer, and mm-wave radars installed underneath the vehicle in addition to a camera mounted on the back. While driving on roads, this system is able to collect pavement related features of friction, roughness and distresses. First, sensors’ data streams have been combined to provide one pavement indicator entailing all the information each of them carry. Then, a Web-based PAVEment MONitoring system (PAVEMON) has been designed to visualize these overview ratings along with all the information from these sensors and camera. In addition to visualization, PAVEMON provides front-end users the capability of accessing and analyzing TeraBytes data of road features in segments as small as 10 meters. Pavement performance, condition and distresses can be monitored frequently using PAVEMON which provides actionable information for making the right repairs, in the right place, at the right time, an ideal test-bed for planning cost-effective maintenance strategies.