Advancing human-centric sustainable construction: barriers and adoption of neuro-safety wearable technologies

Purpose This paper investigates the barriers, adoption mechanisms, ethics and governance requirements of neuroimaging and bio-signal wearable technologies in sustainable construction under the Industry 5.0 paradigm. The objective is to understand human-centred technology adoption while accounting for privacy, surveillance and responsible-use conditions that shape ethical, resilient and sustainable construction practices. Design/methodology/approach A sequenced hybrid analytical framework combining partial least squares structural equation modelling (PLS-SEM), relative importance index (RII), machine learning and SHAP-based explainable AI was applied to survey data from 119 construction professionals. PLS-SEM first validated the theoretical adoption framework and examined structural relationships; RII then translated validated item-level perceptions into weighted adoption-readiness measures; ensemble machine learning provided an exploratory predictive extension; and SHAP analysis identified the most influential barriers shaping adoption-readiness predictions. Findings The results demonstrate substantial explanatory capability (R2 = 0.638, Q2 = 0.622 for PLS-SEM) and indicative predictive capability within the survey dataset (R2 = 0.7744 for the AI ensemble). Technological factors, particularly maintenance complexity and data quality, emerged as the most critical barriers, followed by organizational and ethical considerations. Research limitations/implications The paper is based on 119 responses within an Australian context, which may limit generalizability across regions and construction sub-sectors. The cross-sectional design captures perceptions at a single point in time and does not reflect longitudinal adoption dynamics. Additionally, the adoption index relies on self-reported assessments rather than real-time physiological deployment data. Future research should incorporate larger, cross-cultural samples and longitudinal case studies, integrating live neurophysiological datasets to enhance robustness. The findings nonetheless provide a validated empirical framework to guide policymakers, engineers, and industry leaders in advancing human-centric, sustainable neuro-safety implementation. Practical implications The findings provide actionable guidance for industry stakeholders seeking to implement neuro-safety wearables in construction. Prioritizing improvements in device reliability, signal quality, interoperability with existing digital systems (e.g. BIM/IoT), and maintenance support will significantly enhance adoption readiness. Organizations should complement technical upgrades with targeted workforce training, transparent data governance policies, and clear ethical guidelines to build trust and acceptance. The hybrid predictive framework can support decision-makers in benchmarking adoption readiness, allocating resources strategically, and designing phased implementation roadmaps aligned with safety performance, sustainability goals, and Industry 5.0 principles. Social implications The adoption of neuro-safety wearable technologies has the potential to significantly enhance worker well-being by proactively detecting cognitive fatigue, stress, and psychological risk factors before incidents occur. By promoting human-centric safety systems, the approach supports safer workplaces, reduces injury rates, and strengthens long-term occupational health outcomes. However, responsible implementation is essential to safeguard privacy, autonomy, and informed consent. Transparent governance frameworks and ethical data management practices are critical to maintaining worker trust. Ultimately, neuro-safety integration can contribute to more inclusive, resilient, and socially sustainable construction environments aligned with broader societal well-being objectives. Originality/value This paper integrates theory-driven causal modelling with data-driven predictive analytics to develop a context-specific framework for understanding and supporting the adoption of neuroimaging and bio-signal wearable technologies in sustainable construction. The contribution lies in applying and sequencing established adoption and analytical perspectives within the emerging neuro-safety construction context, while explicitly linking adoption readiness to human-centred Industry 5.0 principles, responsible governance, worker privacy and the Sustainable Development Goals.