Applicability of enhanced Interchange Safety Analysis Tool (ISATe) for local applications: A Florida case study
Article
Saha, D, Alluri, P, Gan, A. (2019). Applicability of enhanced Interchange Safety Analysis Tool (ISATe) for local applications: A Florida case study
. 49 129-144. 10.4399/978882552809110
Saha, D, Alluri, P, Gan, A. (2019). Applicability of enhanced Interchange Safety Analysis Tool (ISATe) for local applications: A Florida case study
. 49 129-144. 10.4399/978882552809110
The Enhanced Interchange Safety Analysis Tool (ISATe) automates the safety evaluation procedures of freeway facilities that are discussed in the Highway Safety Manual. ISATe provides simple worksheets for data entry and facilitates the calculation of expected and/or predicted average crash frequencies for different components of a freeway facility. The ISATe models are designed to evaluate the safety performance of an existing freeway facility, or proposed design alternatives for a new interchange construction and prior to reconstruction of an existing interchange. However, these models are based on national data that may not accurately represent the local conditions. This paper describes a study to evaluate the applicability of ISATe for safety performance evaluation of interchanges in the State of Florida, United States. A before-and-after safety analysis was conducted using ISATe for five interchanges in Florida. The sites were selected based on two criteria: (i) whether crash data were available, and (ii) whether the site characteristics conform to the ISATe analysis criteria. The required roadway characteristics data were collected from different sources, including roadway characteristics inventory, aerial images, and Geographic Information Systems (GIS) maps. Based on the collected roadway and traffic features data, the observed, predicted, and expected crash frequencies by severity (i.e., total, fatal and injury (FI), and property damage only (PDO)) for both before and after periods were calculated and compared for each study interchange. The study results show that the variations between the observed and predicted FI crashes were lower compared to the variations between the observed and predicted PDO crashes. In particular, the ISATe models were found to overpredict the PDO crashes on the study interchanges. Crashes on crossroad ramp terminals were found to be highly overpredicted compared to crashes on freeways and ramps. The paper concludes by providing specific recommendations related to ISATe application in Florida.
