01733251

Component

http://www.trb.org/Publications/Blurbs/179567.aspx

Crash location information is a necessary component of safety analysis and decision-making. Unfortunately, its quality is frequently limited, which poses a major problem in road safety management, project development, and safety research. Without reliable location information, crash data may lead to incorrect safety management decisions and to biased research results. According to the 2012 Indiana Five Percent Report, 74% of reported crashes were assigned to the state road network and, consequently, nearly 50,000 crashes could not be used to analyze safety at individual road locations. It is also important to mention that the rates of unassigned crashes are different for different types of crash and roads. This situation may lead to incorrect prioritization and selection of safety improvements. Many crashes on Indiana local rural roads included in the state crash database have missing or incorrect geo-coordinates due to the incorrect or incomplete location information. The most frequent cases of faulty location information are: 1. An incomplete description of the crash location (missing name of road, distance to the reference intersection, etc.); 2. Missing geographical coordinates; 3. Insufficient precision of reported coordinates; 4. Complete but incorrect crash location description; and 5. Complete but incorrect geographical coordinates. Cases 1 through 3 are easily detected. Cases 4 and 5 are detected if the associated road information collected by police officers at the crash scene is inconsistent with the road information found in a roadway database for the location with the recorded geo-coordinates. Currently, a manual checking of geo-coordinates is performed to correct the crash geo-coordinates or to add missing ones. This procedure is time- and labor-consuming and needs to be automated. To help mitigate the problem, Steiner et al. created a custom interface to enter and geocode crash location by extending the capabilities of Environmental Systems Research Institute (ESRI) geocoders. Bigham et al. used a multistep process that included preprocessing the street name information, geocoding using ArcGIS plus StreetMap Pro 2003 and Google Earth Pro, and Visual Basic for Applications (VBA) to incorporate the offset and the direction. Unfortunately, these tools are not utilized in Indiana, and probably in many other states, because local agencies and other users often do not have knowledge or access to GIS tools. This project aims to develop a convenient method that addresses the needs of local agencies. The developed convenient self-contained computer application supports both fully automated and supplemental semimanual checking and correcting of the crash coordinates without reaching for GIS tools. It uses the police crash reports, available online geocoders, interactive maps:, and a database supplemented with street names, addresses, and coordinates. The developed tool allows agencies to correct location information of most of crashes in the database and to use them in safety analysis. Once implemented, the developed method dramatically reduces the amount of time and effort needed to correct crash coordinates and to assemble data for analysis.

01730920

English

pp 489-493

2019-9

Transportation Research Circular

12th International Conference on Low-Volume Roads

Digital/other

Figures; References; Tables

Automation; Coordinates; Crash locations; Errors; Geographic information systems; Low volume roads; Rural areas

Data errors correction

Indiana

Data and Information Technology; Highways; Safety and Human Factors

TRIS, TRB, ATRI

Mar 9 2020 2:21PM