Development of a modified road safety risk assessment model (iRAP@Highlands) for motorcyclists along the mountainous road network of Cameron Highlands

Motorcycle crashes in mountainous regions remain a persistent road safety challenge due to complex terrain, variable weather conditions, and demanding road geometry. In Malaysia, motorcyclists constitute a significant proportion of road fatalities, yet existing studies have rarely examined their vulnerability within mountainous environments. Previous research often isolates roadway design from broader safety dimensions, overlooking the combined influence of road engineering, rider behaviour, and environmental factors. This study addresses this gap by developing a modified road safety risk assessment model, iRAP@Highlands, specifically tailored for motorcyclists navigating the mountainous road network of Jalan Simpang Pulai to Blue Valley in the Cameron Highlands, Malaysia. This study are structured into four objectives: (1) to identify the thematic structure and key factors contributing to road crashes in mountainous road networks; (2) to develop a road crash data profile for the Cameron Highlands mountainous road network; (3) to establish road condition reports for the main road that captured the highest fatal crash frequency in Cameron Highlands; and (4) to propose a modified road safety risk assessment model for motorcyclists along Jalan Simpang Pulai to Blue Valley, Cameron Highlands. A Systematic Literature Review (SLR) synthesized 34 sub-factors related to road engineering, 18 to rider behaviour, and 7 to environmental factors. The secondary crash data (2015 to 2018) from RMP were analysed using Tableau 10.4, while road survey data were collected by the researcher via a RADIS-equipped vehicle and assessed using MiREV and ViDA software. The star rating risk map was produced to visualise high-risk zones. The modified Multiple Linear Regression (MLR) model showed an increase in adjusted R² from 0.248 (before the intervention) to 0.433 (after the intervention), indicating that the modified model explains a substantially greater proportion of variance in the Star Rating Score (SRS). This study contributes theoretically by localising global iRAP principles through statistical modelling, and practically by enabling quick-win countermeasure planning. The iRAP@Highlands model provides a scalable, data-driven tool for enhancing motorcyclist safety in high-risk mountainous regions, with potential applications in similar geographies worldwide.