<abstract xmlns="http://www.w3.org/1999/xhtml">

<p>Predicting the number of crashes that may occur as a result of specific highway features is critical in evaluating different treatment or design alternatives. Since different highway geometric characteristics can influence crash distribution datasets, Highway Safety Manual’s (HSM’s) predictive method encourages users to predict crashes based on their severity and collision type proportions. This study used crash data from rural two-way two-lane road segments in the Oromia region over seven years to develop Oromia’s fixed crash distribution dataset on Interactive Highway Safety Design Model (IHSDM) software. The crash distribution dataset has two parts; the crash severity proportions and the collision type percentages. The developed Oromia’s fixed crash distribution dataset was compared and validated against the default HSM crash configuration. As a result, the Crash Prediction Model (CPM) evaluation results confirmed that the developed crash severity proportion (the first part of the crash distribution dataset) estimates are more accurate and closer to the observed values. Furthermore, the findings show that crashes in the Oromia region are severer than in the states where the HSM crash configuration was developed. According to the second part of the crash distribution dataset evaluation (collision type percentage), the developed fixed crash distribution dataset outperforms the default HSM configuration in most collision type proportions, but not in all. For instance, from the ten collision type proportions developed, Right-Angle and sides-wipe collision proportions are predicted more precisely by the default HSM configuration. This points to the need for developing collision type proportion (the second part of the crash distribution dataset) as a function rather than a fixed configuration for a better result, based on the availability of complete crash data (i.e. crash location). In general, the study revealed that in order to exploit the full potential of HSM’s predictive approach, researchers must develop a jurisdiction crash distribution dataset using local crash data. The methodology demonstrated in this study to develop the jurisdiction’s crash distribution dataset has been validated as true thus, safety practitioners are encouraged to adopt it.</p>
</abstract>

Predicting the number of crashes that may occur as a result of specific highway features is critical in evaluating different treatment or design alternatives. Since different highway geometric characteristics can influence crash distribution datasets, Highway Safety Manual’s (HSM’s) predictive method encourages users to predict crashes based on their severity and collision type proportions. This study used crash data from rural two-way two-lane road segments in the Oromia region over seven years to develop Oromia’s fixed crash distribution dataset on Interactive Highway Safety Design Model (IHSDM) software. The crash distribution dataset has two parts; the crash severity proportions and the collision type percentages. The developed Oromia’s fixed crash distribution dataset was compared and validated against the default HSM crash configuration. As a result, the Crash Prediction Model (CPM) evaluation results confirmed that the developed crash severity proportion (the first part of the crash distribution dataset) estimates are more accurate and closer to the observed values. Furthermore, the findings show that crashes in the Oromia region are severer than in the states where the HSM crash configuration was developed. According to the second part of the crash distribution dataset evaluation (collision type percentage), the developed fixed crash distribution dataset outperforms the default HSM configuration in most collision type proportions, but not in all. For instance, from the ten collision type proportions developed, Right-Angle and sides-wipe collision proportions are predicted more precisely by the default HSM configuration. This points to the need for developing collision type proportion (the second part of the crash distribution dataset) as a function rather than a fixed configuration for a better result, based on the availability of complete crash data (i.e. crash location). In general, the study revealed that in order to exploit the full potential of HSM’s predictive approach, researchers must develop a jurisdiction crash distribution dataset using local crash data. The methodology demonstrated in this study to develop the jurisdiction’s crash distribution dataset has been validated as true thus, safety practitioners are encouraged to adopt it.

Crash Distribution Dataset: Development and Validation for the Undivided Rural Roads in Oromia, Ethiopia

Rostock University Justus-von-Liebig-Weg 6

Jimma Institute of Technology, Jimma University

Transport and Telecommunication Journal

This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 3.0 License.

Predicting the number of crashes that may occur as a result of specific highway features is critical in evaluating different treatment or design alternatives....

Crash Distribution Dataset: Development and Validation for the Undivided Rural Roads in Oromia, Ethiopia

Online veröffentlicht: 18. Feb. 2022

Seitenbereich: 11 - 24

DOI: https://doi.org/10.2478/ttj-2022-0002

Schlüsselwörter
crash distribution dataset, crash prediction model, HSM, IHSDM, Oromia, rural roads

© 2022 Alamirew Mulugeta Tola et al., published by Sciendo

This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 3.0 License.

Crash Distribution Dataset: Development and Validation for the Undivided Rural Roads in Oromia, Ethiopia

Online veröffentlicht: 18. Feb. 2022

Seitenbereich: 11 - 24

DOI: https://doi.org/10.2478/ttj-2022-0002

Schlüsselwörtercrash distribution dataset, crash prediction model, HSM, IHSDM, Oromia, rural roads

© 2022 Alamirew Mulugeta Tola et al., published by Sciendo

This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 3.0 License.

Schlüsselwörter
crash distribution dataset, crash prediction model, HSM, IHSDM, Oromia, rural roads