GIS-Based Instructional Tool for Crash Prediction Methods
UTC Project Information
Problem: The first version of the Highway Safety Manual (HSM) was released in 2010 and is currently being deployed by several states as the primary methodology for performing predictive analysis to identify critical segments of the network and to evaluate the benefits of countermeasures. In this context, it is critical to train the current and future professionals on the underlying theory behind these methods and the effective application of the same. Although the HSM methods rely on vast amounts of spatial data (roadway network and geometry, geo-coded crashes etc.) the training materials rely mostly on spreadsheet-based tools for application of the methods and the HSM software are also non-spatial and do not directly integrate with Geographic Information Systems (GIS). Objective: The intent of this study is to develop a GIS-based instructional tool which can be used by both graduate students and current professionals to learn about the HSM-based predictive methods. The GIS platform of the tool will be immensely beneficial so that the students can appreciate (visually) the context in which these methods are being applied. As such, this study will contribute to both the educational and technology transfer goals of STRIDE.
Methodology: The overall project methodology comprises three steps. First, the HSM crash-prediction methods will be coded into the Signal Four Software for selected facility types. This will involve coding in the appropriate Safety Performance Functions and Crash Modification Factors. Next, Instructional Modules will be developed that will provide overviews of both the software and the analytical methods in addition to providing step-by-step guidance for segment- and intersection- level analyses. Finally, the software and instructional modules will be tested using students from various transportation engineering and urban planning classes and be presented to practitioners via a webinar.
Anticipated Results: This project will result in the development of an interactive GIS web-based instructional tool for Crash Prediction Models. Several self-instructing tutorials will be developed which can be used by students either independently or in the context of a course. These tutorials will use real world data from Florida and the GIS-environment will facilitate the students appreciating the context in which the data are obtained and methods applied and thereby leading to improved understanding of the methods. A webinar will also be developed aimed at practitioners. This will cover data issues in greater detail in addition to instructing the audience on application of predictive methods. Since the software is web-based, these can be accessed and used easily by anyone within the region.
Contribution: The project directly contributes to enhancing the goals of transportation safety within the region. The instructional module will facilitate improved understanding of the HSM-based predictive methods and the appropriate application of the same. In the longer term, we envision that the consistency checks and comparative analysis capabilities supported by the software will also lead to improvements in data and methods, which in turn, would translate into better predictive capabilities. The instructional module will be designed to allow future scalability into a full crash prediction feature of the Signal Four Analytical system in order to support the needs of researchers and practitioners in the traffic safety improvements efforts.
Return to Current Projects