Monitoring Urban Freeway Incidents by Wireless Communications

Abstract

The efficacy of the driver-initiated incident detection system was assessed by modeling the use of new communications technologies to mitigate problems that beset current cellular call-in programs, problems such as insufficient information on incident location, false reporting, and duplicate calls. An analytical model supported by field data for several important variables was used to determine the system performance measures, that is, the detection rate, mean time to detect, and false alarm rate. The results indicate that with the implementation of new technologies that lead to the automatic geolocation of the driver initiating an incident call, the driver-initiated detection system is effective. The system achieved good detection rates and faster detection times across all simulated incidents occurring under variable traffic flow rates that ranged from light traffic to congested traffic. For instance, even when the proportion of drivers with cellular phones was 1 out of 10, with only 10 percent of them willing to use their phones to report the incidents, close to 80 percent of the incidents were detected within 2 min. This result is consistent with the results of the field studies and shows that even better detection performance can be achieved if current levels of cellular phone ownership in the United States are considered and if drivers’ reporting propensity is raised. It is noteworthy that the Cellular Telecommunications Industry Association reported that there was one cellular phone user per two licensed drivers in the United States at the end of 1999.