Incident Detection Optimization and Data Quality Control (0-4745-3)

Project 0-4745 developed a geographic information system (GIS)-based approach for the determination of patterns in the spatial and temporal distribution of incidents along freeway corridors. The research documented incident detection and data archival at several transportation management centers (TMCs) in Texas, a process to develop a data model and geodatabase of intelligent transportation system (ITS) equipment and archived ITS data, and a process to determine patterns in the spatial and temporal distribution of freeway incidents. This report describes the procedures and activities completed during the second phase of the project. The analysis included an evaluation of incident detection procedures at a sample TMC (Trans Guide ) and an assessment of the feasibility to modify/calibrate alarm threshold values to help optimize incident detection practices at that TMC. The research involved the use of two performance measures (detection rate and false alarm rate) and the development of a prototype off line tool to evaluate automatic incident detection algorithm performance. The analysis also included an evaluation of archived loop detector data completeness and quality control.