Speed Reduction Model for Assessing Geometric Consistency of Two-Lane Highway in Central India

Abstract

Speed reduction (SR) on horizontal curves is a critical measure influencing safety on two-lane rural highways. The SR measures based on fixed-point speed observations (ΔV<inf>85</inf>) underestimate the actual SR compared with those derived from continuous speed profiles (Δ<inf>85</inf>V). This study addresses this gap in the literature by developing a Δ<inf>85</inf>V model using high-resolution continuous speed profile data on different horizontal curves across a two-lane rural highway in Central India. The model ensures the influence of horizontal curves within the curve discovery region. It integrates rarely studied critical design parameters, such as transition curve length, superelevation, and gradient, alongside commonly used parameters within a single comprehensive study. The model was validated using out-of-sample data to demonstrate transferability across similar roadway classifications. In addition, this study illustrates the application of the Δ<inf>85</inf>V model in the geometric design and consistency evaluation. The findings could be useful in a design feedback mechanism to guide geometric improvements and classify existing tangent-to-curve transitions based on their expected level of consistency. The insights contribute to improved geometric design practices tailored to India and lay the foundation for future research incorporating deceleration length, crash data, and driver characteristics. Overall, the developed model advances speed consistency evaluation and supports safer highway designs under heterogeneous traffic and driving behavior conditions. © The Author(s) 2026