DOI: 10.11648/j.jccee.20261103.17 ISSN: 2637-3890

A Novel Shannon Entropy Approach and Interface Computer Calculators for the Diagnosis of Highway Pavement Performance (Pavement Entropy Index — PEI)

Radu Andrei
Show PDF Cite
Road pavement condition assessment is a fundamental component of highway asset management, underpinning decisions related to maintenance scheduling, resource allocation, and infrastructure investment. Existing methods — including the Pavement Condition Index (PCI), the International Roughness Index (IRI), rutting measurement, deflection testing, and automated distress detection — each provide valuable but inherently partial perspectives on pavement performance. A persistent limitation of these methods is their reliance on scalar metrics, empirical thresholds, and — in the case of visual survey techniques — subjective human judgement, none of which offer a theoretically grounded framework for characterizing the complexity or disorder of pavement deterioration. This paper proposes an original approach for the individual and integrated quantitative evaluation of road pavement condition based on Shannon information entropy. Rooted in information theory, entropy provides a mathematically rigorous measure of disorder and uncertainty that is directly applicable to the multi-dimensional, stochastic nature of pavement degradation. The proposed framework introduces a suite of entropy-based indices covering distress diversity, roughness profile complexity, rut pattern irregularity, crack network structure, and structural non-uniformity, which interface systematically with each of the established assessment methods. These component indices are synthesised into the Pavement Entropy Index (PEI) through a hierarchical weighted model. The framework is applied to three categories of case studies: flexible (asphalt) pavements, rigid (concrete) pavements, and transport infrastructure earthworks. Specific interactive digital calculators implementing the framework have also been developed with the aid of AI.

More from our Archive