Citation
Mohd Razelan, Intan Suhana
(2015)
Development of composite risk index for federal roads in Malaysia.
Doctoral thesis, Universiti Putra Malaysia.
Abstract
According to WHO (2013), middle-income countries like Malaysia, Indonesia and several other ASEAN countries suffer the highest traffic fatality rates compared to most developed countries where crash statistics are used to evaluate the safety status of these countries. Crash data has been acknowledged as the most popular and acceptable road safety indicator in recognizing road section’s safety status. However, the reliability of crash data in correctly identifying the road section’s safety status has been widely argued by road safety experts. In light of that, a new method called composite risk index that would act as a proactive measure in evaluating road section’s safety status has been introduced and tested in this research. This research attempts to fill in the missing links on the role of different road environment factors in producing risk towards road users. Other than that, a significant contribution to the knowledge in the theory of road safety index is made by developing a risk index in evaluating road section’s safety status. Identifications of the road environment factors of the existing road networks were done by adopting naturalistic driving method in recording different road environment conditions for 315.5 km length of federal road. The road environment factors for the whole study area were identified by clustering fourteen original attributes into several groups having similar characteristics. In the development of the composite road environment risk index, several procedures were involved in defining the underlying structures of the original indicators, weighting, normalizing and aggregating the indicators before the composite indexes were developed. A statistical method of principal component analysis was adopted in defining the underlying structure of the original indicators, clustering the original indicators according to current road conditions and assigning statistical weight to each indicator to avoid the possibility of biased results. After that, z-score method was used to normalize the indicators so that the indicators could be added up and finally, the weighted sumscore method was employed to combine the original indicators and formed a composite index. The validation procedure was carried out by utilizing the Spearman’s rho correlation coefficient procedures so that the ability of the composite index to be used in the real world is verified. The research outcomes revealed four main road environment risk factors for federal roads, namely road operational environment, roadway environment, roadside environment and road infrastructure environment. Following that, three composite road environment risk indexes were successfully developed for three different road environments. Results from the composite index demonstrated that specific attentions should be given on the combinations of motorcycles and heavy vehicles, especially at locations where human activities on the roadside areas are high and signalized intersections are mutually existed. Also, highly developed roadside areas have been recognized to contribute higher risk, especially on the aspect number of accesses and an existence of median. The outcomes from this research provide useful preliminary inputs in highlighting the role of road environment risks in defining crash factors especially in developing countries. The development of the composite risk index in proactively evaluating road section’s safety status is definitely a state-of-the-art method that can be used in other developing countries to evaluate their road section’s safety status when the crash data does not exist or in poor quality.
Download File
Additional Metadata
Actions (login required)
|
View Item |