Using Reidentification to Evaluate the SCOUT Traffic Management System (TMS)

The Kansas City SCOUT Traffic Management System (TMS) is a cooperative effort funded by the State Departments of Transportation of Kansas and Missouri and involving the DOTs, local and state law enforcement, Mid-America Regional Council (MARC—KC MPO), local municipalities and townships, and others. The system integrates dynamic message signs, video surveillance, traffic detection, computer networking and analysis, and a Traffic Operations Center. In the evaluation of SCOUT, incident-related delay was chosen to be the primary measure of effectiveness, because that is where the greatest benefits of the system are expected to occur. To measure this delay, an approach was developed that utilizes video observation and postprocessing to obtain travel times throughout the peak periods. The travel times were obtained by employing characteristic-based reidentification, aided by a custom computer application developed at The University of Kansas, ReID. Using ReID, temporal profiles across the peak periods can be generated of the average travel time for each corridor under study. Profiles are generated for baseline (i.e., non-incident) conditions and for each period during which an incident occurred on the study segment while it was under observation. The baseline profile is subtracted from the incident profiles to obtain the delay caused by the incident. Comparing the delay caused by incidents before the implementation of SCOUT with those after the implementation of SCOUT will show any improvements caused by the SCOUT system. Data was collected over a total of approximately seven months during 2002 and 2003.