From immersion to intention: a structural model of virtual reality-based human–robot collaboration training in construction

Adetayo Onososen, Innocent Musonda

Abstract The integration of robotic systems into construction workflows poses a workforce challenge with two intertwined dimensions, physical robots remain too costly to allocate to repetitive hands-on training, while workers commonly exhibit distrust and resistance toward autonomous systems, making safe and productive human–robot collaboration (HRC) difficult to achieve without preparation. On-site training is neither scalable nor safe at the proximities that construction HRC demands. Virtual reality (VR) offers a hazard-free, repeatable, and cost-efficient alternative, yet no prior investigation has modelled the cognitive and technological mechanisms by which VR-based training translates into a willingness to collaborate with construction robots. The present study addresses this gap by developing and empirically testing a structural model that traces the pathway from trainees’ pre-training state of mind, through VR system features and usability, to the immersive learning experience and behavioural intention to collaborate. A purpose-built multi-robot VR environment incorporating drone, crane, forklift and loader scenarios was deployed with 42 construction-aligned postgraduate students and equipment operators; data were analysed using partial least squares structural equation modelling (PLS-SEM). The learning experience emerged as the dominant predictor of collaboration intention (β = 0.846, p < 0.001), and VR features influenced learning entirely through usability rather than directly reframing training design priorities from fidelity-first toward usability-first. This paper contributes the first unified structural model of VR-based HRC training to integrate technology-mediated learning theory, the Technology Acceptance Model, the Theory of Planned Behaviour, and human–robot interaction theory, offering an empirically validated framework for construction workforce preparation in human–robot collaborative environments.