Main article

Ayesha Bhatti*
National Institute of Transportation, School of Civil and Environmental Engineering, National University of Sciences and Technology (NUST), Islamabad 44000, Pakistan
ayesha.bhatti@nust.edu.pk
Muhammad Usman Khan
Department of Transportation Engineering and Management, University of Engineering and Technology (UET) Lahore, Lahore 54890, Pakistan
Tariq Mehmood
Pakistan Highway Research and Training Centre (PHRTC), National Highway Authority, Islamabad 44000, Pakistan
Sana Farooq
Department of Civil Engineering, NED University of Engineering and Technology, Karachi 75270, Pakistan

Abstract

Pavement distress on motorways that traverse arid transportation corridors accumulates under a distinctive combination of stressors—strong solar radiation, wide diurnal temperature swings, aeolian abrasion, limited moisture for self-healing, and concentrated heavy-vehicle traffic—that together shorten service life and complicate inspection. Conventional highway-agency practice addresses this problem in a fragmented manner: condition surveys, automated defect detection, deterioration forecasting, and maintenance scheduling are typically operated as independent processes by different organisational units, and the outputs of each process rarely flow back to recalibrate the others. This paper argues that the productivity gains of recent advances in automated distress detection—attention-augmented one-stage object detectors, dynamic-focusing regression losses, edge-device inference—can be realised at asset-management scale only if detection is embedded within a life-cycle management framework that binds detection, deterioration modelling, and maintenance optimisation into a single closed loop. We propose such a framework for arid-corridor motorways and ground it in a case-study analysis of the Pakistani motorway network, whose M-1, M-2, M-8, and M-9 segments present representative climatic and traffic challenges. The framework is organised around four stages: condition-inventory assessment, AI-based automated distress detection, deterioration modelling through a combination of Markov chain and mechanistic-empirical approaches, and maintenance optimisation through life-cycle-cost analysis subject to reliability and budget constraints. An illustrative cross-strategy analysis indicates that an AI-assisted life-cycle strategy reduces thirty-year normalised cost by approximately sixty per cent relative to a reactive baseline while maintaining target reliability of 0.80 at the segment level. The paper closes with specific recommendations for policy instruments, procurement reform, and institutional capacity building that would support adoption by highway agencies operating in arid regions across South Asia, the Middle East, and North Africa.

Article details

How to Cite

Bhatti, A., Khan, M. U., Mehmood, T., & Farooq, S. (2025). A LIFE-CYCLE MANAGEMENT FRAMEWORK FOR HIGHWAY PAVEMENT DISTRESS IN ARID TRANSPORTATION CORRIDORS: INTEGRATING AI-BASED DETECTION WITH MAINTENANCE DECISION MODELS. International Journal of Infrastructure Research and Management, 13(1), 29-55. https://doi.org/10.63646/j.ijirm.2025.130104