Paper Details

TRB 90th Annual Meeting (January 23-27, 2011)

Event Number:	326
Event Title:	Advancements in Quantification of Highway Safety Performance
Event Date:	Jan 24 2011 2:30PM- 5:00PM
Event Location:	Marriott, Salon 2
Event Description:
Event Agenda:

Getting Started with SafetyAnalyst: Georgia’s Triumph (11-3691)

The identification and prioritization of sites with potential for safety improvement (PSI) is an initial step in the roadway safety management process. This step could be performed by traditional methods (crash frequency, crash rate, safety indices etc) and advanced methods (using Empirical Bayes or Full Bayes approach). Understanding that the traditional methods are fraught with limitations, biases and false assumptions, recent research has opened doors to new and advanced methodologies in the form of SafetyAnalyst and Highway Safety Manual. Unlike basic methods, Empirical Bayes approach used in these tools involves rigorous calculations and requires safety performance functions (SPFs), a relation between crashes and exposure, in addition to roadway characteristics, traffic volume and crash information. The notion of automation of such rigorous tools led to the creation of “SafetyAnalyst”. SafetyAnalyst is a set of software tools used for highway safety management that integrates all parts of the Safety Management System. Using data from Georgia DOT, import files for SafetyAnalyst were generated for roadway segments and network screening was performed to prioritize sites with potential for safety improvement. The intent of this paper is to document various problems that arose in the process of getting started with SafetyAnalyst and present some recommendations for reducing the learning curve for states wishing to adopt the software. Additionally, the network screening tool was used to compare top ranked sites identified by SafetyAnalyst with those identified by traditional methods. These results portray the major differences in rankings by method and further support need for using more advanced methods.

Authors
     Ogle, Jennifer Harper, Clemson University
     Alluri, Priyanka , Florida International University