Research Project:

Secondary Accident Data Fusion for Assessing Long Term Performance of Transportation Systems

Final Report

PDF version 537k
HTML version 226k

Principal Investigator

Carlos Sun
University of Missouri – Columbia
(573) 884-6330
csun@missouri.educsun@missouri.edu

Project Objective

To develop a robust, readily deployable methodology for extracting secondary accident data that can be used for analyzing the safety benefits of different transportation systems. Contributing objectives include

Developing a convention for formatting accident reports that allows data regarding secondary accidents to be extracted.
Developing a methodology for processing intranet traffic reports that tracks the progression of individual incidents.
Developing an appropriate a methodology to model incomplete incident reports

Project Abstract

Secondary accidents are accidents which result from an existing primary incident. Many times these accidents occur at the end of queues that resulted from the primary incident. Quickly opening the highway after an incident reduces the potential for secondary accidents. It is easy then to see the value of analyzing secondary accidents when considering traffic incident management strategies such as Intelligent Transportation Systems and the MA program. On the other hand, the effects of such systems on primary accidents would be much less, because many of these accidents are caused by driver error such as fatigue, intoxication, or aggressive driving. Therefore traditional analysis of primary accidents and accident rates will not reveal the full potential of such systems.

In order to use secondary accidents as a performance measure, it is necessary to first separate such accidents from the rest of the database. Since the effect of primary accidents can persist long after a roadway has been cleared, it is difficult to determine at the scene of an accident if it is due to recurrent or non-recurrent congestion. By analyzing individual traffic reports in detail, the reporting times of the incident and the dynamic locations of the back of the queue can be found. However, the intranet reports currently need to be processed significantly to accomplish this. This project proposes the use of data fusion of intranet traffic reports with the accident database, and will result in a near-term technology for analyzing the safety impacts of transportation assets.

Task Descriptions, Milestone, and Dates

Accident Database Processing (January 2006)
Traffic Report Processing (January 2006)
Modeling Incident Progression Curves (February 2006)
Secondary Accident Methodology (February 2006)
Generate Final Report (February 2006)

Student Involvement (e.g., Thesis, Assistantships, Paid Employment)

None

Relationship to Other Projects

None

Technology Transfer Activities

The project will result in a final report that contains the accident data format and required accident information, the Intranet traffic report processing methodology, the incomplete incident data modeling methodology and final model specification, and secondary accident extraction methodology. This report will be distributed or made accessible to other in and out of state agencies as appropriate.

Potential Benefits of the Project

There is a great potential for the immediate technology transfer and the implementation of the results of this research in Missouri and Kansas and possibly in other neighboring states. This implementation is in the form of a standard method for extracting secondary accidents from the primary accident database maintained by the police. This standardization would guarantee that the evaluation of safety and other asset management systems will consider secondary accident data in a consistent manner.

Budget

$39,217 MTC/ $39,261 Cost Share (12-month project)

TRB Keywords

Secondary Accident, Primary Accident, Non-Recurrent Congestion, Queue, Incident Progression, Accident Report, Intranet Traffic Report, Data Extraction, Data Fusion, Model, Intelligent Transportation System