Probabilistic pavement performance modeling using hybrid Markov Chain: A case study in Afghanistan

• Published in: Case Studies in Construction Materials 2024-07, Vol.20, p.e03023, Article e03023
• Main Authors: Wasiq, Samiulhaq, Golroo, Amir
• Format: Article
• Language: English
• Subjects: Markov Chain; Pavement performance models; Probabilistic modelling; Smartphones
• ISSN: 2214-5095; 2214-5095
• DOI: 10.1016/j.cscm.2024.e03023

Road networks play a crucial role in the economic fabric of every country, particularly in nations like Afghanistan, strategically positioned on the international transit route from Europe to East Asia. In such contexts, the development of pavement performance models holds paramount importance for effective pavement maintenance planning, ensuring the provision of high-quality infrastructure to facilitate the transportation of goods and passengers. Despite Afghanistan's strategic location, the absence of budget allocations for pavement monitoring and maintenance within its transportation networks has led to limited acquisition of pavement condition data, hindering the development of pavement performance models. The primary objective of this study is to pioneer the creation of a pavement performance model for Afghanistan, utilizing a cost-effective and reasonably accurate approach based on data collected through smartphones. The study focuses on approximately 7000 km of Afghanistan's highways. It initiates by examining the current state of Afghanistan's road network through smartphone data collection. Subsequently, the gathered data is meticulously prepared and analyzed to derive the Pavement Condition Index (PCI). Finally, a pavement performance model for PCI is formulated, employing a combination of homogenous and non-homogenous Markov Chain processes. The model's efficacy is successfully validated with real-world data, leading to the conclusion that the proposed approach proves efficient and effective in developing performance models for other developing countries grappling with similar data and budget constraints.