Evaluation of Road Condition Indices Methods and Applicability for Use in Machine Learning - read online paper summary

Aug 4, 2026

Upload by Levi

Road maintenance under a Pavement Management System (PMS) relies on accurate pavement condition data to support cost-effective decisions on maintenance and rehabilitation. The study focuses on South African lower road authorities facing suboptimal management and poor road condition. It compares visual and deduct-point condition indices for flexible pavements using descriptive and inferential statistics on low- and high-volume case studies. Results show inconsistencies in VCI versus CISURF and CIPAVE; four machine-learning models are evaluated, with Gradient Boosting offering the strongest deployment potential and best accuracy for CISURF/CIPAVE.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages17

Reading Time43 min

AccessFree to download

Abstract

Introduction

Pavement Management System and decision-making

Road condition indices used in South Africa

Motivation for continuous evaluation with ML

EVALUATION OF ROAD CONDITION INDICES METHODS AND APPLICABILITY FOR USE IN MACHINE LEARNING M SIMELANE1* and A RAMPERSAD1** 1Smart Mobility, Council for Scientific and Industrial Research, Pretoria, South Africa; *Tel: 012 841 2924; [Email: ](Email: msimelane@csir.co.za)[msimelane@csir.co.za](Email: msimelane@csir.co.za) **Tel: 012 841 3831; Email: [arampersad@csir.co.za](arampersad@csir.co.za) ABSTRACT Road maintenance is a crucial process for pavement management systems. South African local roads managed by lower road authorities (municipality, etc) are in critical condition, and their management is not at optimum level which is evident from their poor condition. The aim of this paper is to provide a Machine-Learning algorithm to assist road authorities to provide optimal maintenance strategies. The objective of the study was to determine the most effective condition index for management of flexible pavements. This is achieved by conducting descriptive and inferential statistical analysis of two case studies (Low volume roads and High-volume roads) . Statistical analysis indicated that the visual condition index (VCI) has inconsistencies compared to the deduct point surface condition index (CISURF ) and deduct point pavement condition index (CIPAVE ) found in TMH 22. Four machine learning models were created which included the Gradient Boosting Classifier, Random Forest Classifier, Support Vector Machine Classifier, and Decision Tree Classifier. Of the four models explored, the model with the greatest potential for deployment was the Gradient Boosting Classifier (GBC) model. The GBC model had an accuracy of 74 %, 85 % and 93 % in relation to the VCI, CISURF & CIPAVE respectively. The CISURF and CIPAVE was identified as the most effective index for use in flexible pavements. 1. INTRODUCTION A Pavement Management System (PMS) is a set of defined procedures for collecting, analysing, maintaining, and reporting pavement data, to assist the decision makers in finding optimum strategies for maintaining pavements in serviceable condition over a given period of time for the least cost. Key components of a PMS include, but are not limited to (Sabita, 2020) : • Road inventory. • Pavement condition surveys. • Database for information recording. • Analysis schemes. • Decision criteria. • Implementation procedures. The decision criteria and implementation procedures are dependent on the information collected as part of the road inventory and pavement condition surveys. This paper addresses the Machine Learning (ML) component of the analysis schemes within PMS. Having effective ML tools to analyse road data can advance the sector in making critical decision on the maintenance and rehabilitation (M & R) . Balaram (2022) exposed the shortcomings of manual visual assessment methods which has been used in industry which slowed production and performance of various organisations. The study further highlighted the evolution to the existing Pavement Management System through technology. The contribution technology can make creating an effective and efficient Pavement Management System whereby synergising the components of the PMS life cycle creating an integrated solution for South Africa (SA) . This includes Designs, Construction, and a fixed Asset evaluation of the road network for forecasting and budget allocation at government level which in the future can dissolve the maintenance backlog that is experienced. As the road indices contribute to decision making for M & R, they must be accurate and a true reflection of road conditions, since delayed M & R can be detrimental financially. It is important to note that M & R activities are not solely based on road visual condition indices, but also other factors such as traffic, funding, type of road class, etc. Nonetheless, visual condition indices contribute significantly to M & R. A critical evaluation of indices is necessary to determine if the visual condition i

FAQs

What problem does the paper address in pavement management?

It addresses that road maintenance by lower road authorities is not at an optimal level, reflected in poor road conditions, and that better analytical tools are needed for effective maintenance and rehabilitation decisions.

Which road condition indices are compared for flexible pavements?

The study compares the Visual Condition Index (VCI) with deduct-point indices including CISURF and CIPAVE (from TMH 22), which differ in how distress information is aggregated.

Which machine learning model showed the greatest deployment potential?

Among four tested models (Gradient Boosting, Random Forest, SVM, and Decision Tree), Gradient Boosting Classifier (GBC) showed the greatest potential for deployment, with reported accuracies of 74% (VCI), 85% (CISURF), and 93% (CIPAVE).

What conclusion does the study draw about the most effective condition index?

CISURF and CIPAVE are identified as the most effective indices for use in flexible pavements, while VCI was found to have inconsistencies compared with the deduct-point indices.

Aug 4, 2026

Upload by Levi