Segmentation Effect on the Transferability of International Safety Performance Functions for Rural Roads in Egypt

Abstract

This paper examines the transferability of the Safety Performance Function (SPF) of the Highway Safety Manual (HSM) and other 10 international SPFs for total crashes on rural multi-lane divided roads in Egypt. Four segmentation approaches are assessed in the transferability of the international SPFs, namely: (1) one-kilometer segments (S1); (2) homogenous sections (S2); (3) variable segments with respect to the presence of curvatures (S3); and (4) variable segments with respect to the presence of both curvatures and U-turns (S4). The Mean Absolute Deviation (MAD), Mean Prediction Bias (MPB), Mean Absolute Percentage Error (MAPE), Pearson χ2 statistic, and Z-score parameters are used to evaluate the performance of the transferred models. The overdispersion parameter (k) for each transferred model and each segmentation approach is recalibrated using the local data by the maximum likelihood method. Before estimating the transferability calibration factor (Cr), three methods were used to adjust the local crash prediction of the transferred models, namely: (1) the HSM default crash modification factors (CMFs); (2) local CMFs; and (3) recalibrating the constant term of the transferred model. The latter method is found to outperform the first two methods. Besides, the results show that the segmentation method would affect the performance of the transferability process. Moreover, the Italian SPFs based on the S1 segmentation method outperforms the HSM and all of the investigated international SPFs for transferring their models to the Egyptian rural roads.

Systematic review, evaluation and comparison of different approaches for the implementation of road network safety analysis

“…Finally, an overview of the criteria adopted for the segmentation are provided. Indeed, segmentation is always a crucial and not trivial issue in order to define a road network safety analysis [ 99 , 100 ].…”

“…As highlighted by several authors [ 99 , 100 ], segmentation is a crucial step in a network safety analysis. Segmentation allows to divide the network in shorter road elements which can be easily analyzed, and which generally have homogeneous characteristics (e.g., the same lane width, the same traffic, etc.).…”

Assessing the transferability of the Highway Safety Manual crash prediction model for divided rural multilane highways in Brazil

“…Besides, as opposed to the reduction in the total number of people who died on federal highways between 2009 and 2019 (Figure 1), the proportion of fatal crashes in Brazil has increased (Figure 2). Although the increase might be related to the recent Elagamy et al, 2020;Farid et al, 2016Farid et al, , 2019Feng et al, 2020;Silva, 2017). Hence, this manuscript aims to assess the transferability of HSM predictive models when applied to an international context, such as on Brazilian highways.…”

“…The calculated calibration factors were 3.67, 3.77, and 2.60, respectively. The conclusions elucidate the lack of parameters to consider a model transferable and a gap in the knowledge for models in other facility types Elagamy et al (2020). used four types of segmentation procedures in Egypt, which affected the performance of the transferred SPFs.…”

“…Percentage Error (MAPE) measures the deviation between observed and predicted values, and it was calculated based onElagamy et al (2020) to avoid dividing predicted values by a possible zero observed number of crashes. 𝑌 ̂𝑖 corresponds to the predicted data values, 𝑌 𝑖 is the observed data values, and n is the sample size.…”