Exploring the potential of machine learning adoption to improve construction project management practices: practitioners’ perspectives
Nurhayatul Khursniah Hasim, Nurshuhada Zainon, Hafez SallehPurpose
The purpose of this study is to explore practitioner perceptions, readiness and barriers regarding machine learning (ML) adoption in construction project management. It examines ML’s potential to enhance decision-making and investigates the structural and behavioural barriers to digital adoption.
Design/methodology/approach
A quantitative descriptive approach was conducted, surveying 262 construction professionals across Malaysia. The data collected via structured questionnaires evaluated practitioner readiness, decision-making inefficiencies and perceptions of ML constraints. Descriptive and non-parametric statistical analyses were used for data interpretation.
Findings
Significant inefficiencies exist in manual practices like scheduling, cost estimation and risk management. While professionals recognise ML’s potential to improve these areas via predictive analytics, actual adoption is heavily constrained by limited technical expertise, high upfront costs, data quality issues and cultural resistance.
Research limitations/implications
The study is limited by self-reported perception data capturing industry readiness rather than direct algorithmic performance. Future research should investigate empirical case studies of actual model deployments and broader executive-level stakeholder representation.
Practical implications
The study highlights the need for targeted data literacy upskilling, standardised data governance frameworks and scalable implementation strategies to bridge the gap between awareness and practical application.
Social implications
A successful ML adoption relies entirely on bridging the internal digital literacy divide. Cultivating data literacy minimises human behavioural resistance, enabling construction organisations to transition smoothly into a modern data-driven workflow.
Originality/value
This study offers context-specific empirical insights into human and organisational prerequisites for digital readiness, bridging the gap between technical ML and real-world digital adoption constraints.