COMPARATIVE ANALYSIS OF THE MULTI-STAGE EXPONENTIAL MARKOV HAZARD MODEL (MUSTEM) AND HIGHWAY DEVELOPMENT MODEL (HDM-4) FOR ROAD MANAGEMENT IN LAOS

Abstract

 Enhancing road network conditions and reducing maintenance costs are vital for social and economic growth, particularly in developing countries like Laos, which face financial constraints and diverse environmental challenges. This study aims to compare two robust modeling approaches for road network infrastructure management i.e., probabilistic, represented by the Multi-Stage Exponential Markov Hazard Model (MUSTEM), and deterministic model, exemplified by Highway Development Management Model Four (HDM-4). The MUSTEM model offers a superior stochastic framework for assessing asset condition transition over time as it suitably models uncertain pavement degradation. Whereas HDM-4 provides precise performance metrics and is acknowledged for its cost-effectiveness, it requires a lot of data and its deterministic deterioration model does not capture uncertainty in pavement deterioration. The comparison analysis utilized International Roughness Index (IRI) survey data from the Laos Road Management System (RMS) database from 2014–2015. The findings of the study showed that the MUSTEM model is more effective in developing maintenance schedules and budgeting based on the probabilistic transition of road condition state considering uncertainties using limited data. On the other hand, despite HDM-4 offering cost-effectiveness that incorporates user and society benefits, it has a huge data requirement and fails to incorporate uncertainty that is typical of pavement degradation. This study investigated the effectiveness and superiorities of the two models, specifically in road network performance prediction, maintenance strategy and life-cycle costs optimization. The study informs decision-making for road infrastructure management systems, particularly the Laos RMS and similar systems.