Home About ATSIP Program Awards Venue Sponsors/Exhibitors
Roadway Safety Management System in CT: Data Integration, Analysis, Validation and Implementation

Abstract:  The State of Connecticut is in the era of developing a distinguished roadway safety management system, including the process of network screening, diagnosis, countermeasure selection, economic appraisal, project prioritization, safety effectiveness evaluation as well as systemic safety project selection.

The newly designed system is expected to provide both state and local agencies with effective solutions in identifying the roadway locations with the highest potential for safety improvement, recommend the most cost-effective countermeasures, and evaluate the implemented safety projects for better project planning in the future. In this project, the State roadway network was divided into roadway segments and intersections.

 Each component was further separated into diverse facilities based on the land use type, roadway functionality, traffic control, lane count etc. Furthermore, in order to evaluate the roadway safety, crash prediction models were estimated for each facility by crash type and crash severity respectively. The crash prediction models were combined in use with the three most commonly used approaches to identify and rank locations with over-represented crash occurrence including simple ranking method, sliding window method and peak searching method. To help engineers implement such a network screening process, a customized and user-friendly web application was developed.

The application allows the users to select focus area, crash type, crash injury, facility type, area of interests, statistical performance measures, and appropriate screening methods. The tool has interactive maps that show the location of the selections and the results of the top ranked sites. Given the identified locations, a variety of diagnosis options are proposed to investigate the crash contributing factors, including crash descriptive analysis, statistical test, data mining technology and crash/roadway geometric diagram etc.

The next step involves the recommendation and selection of relevant countermeasures related to the aforementioned issues. Engineering economic analyses such as life cycle cost analysis, net present analysis and benefit-cost analysis methods etc. will be provided to compare the cost-effectiveness of candidate countermeasures and determine the most reliable ones. All proposed projects will be further ranked based on the State budget availability, and the implemented safety projects will be further investigated for the effectiveness evaluation as well as for the development of state specific crash modification factors. In this presentation, we will introduce our experience of developing state-specific safety performance functions by facility type and by crash type/injury, present the customized network screening tool, discuss experience and lessons learned, and introduce the other modules to be developed in this roadway safety manage system.

Kai Wang - received his Ph.D. in Civil & Environmental Engineering with a concentration in Transportation & Urban Engineering at the University of Connecticut, as well as a Master of Transportation Engineering at the South Dakota State University. His research interest includes State-of-the-art statistical models in crash count and crash severity prediction, crash data analysis, data mining technologies, transportation economics, risk analysis and simulations in transportation decision making, driver errors and human behavior analysis as well as GIS and GPS applications in transportation. Dr. Wang joined the Connecticut Transportation Safety Research Center (CTSRC) in August, 2016, and is currently working as a transportation safety engineer to conduct traffic safety related research, develop Safety Performance Functions (SPFs) in the context of Highway Safety Manual (HSM), and offer cost-effective countermeasures to improve transportation safety for the State of Connecticut.



Session Material
Quick Links


Chris Osbourn


fb_logo   linkedin