Trust dimensions and operational performance in VR-based drone tasks in construction

Purpose The successful deployment of drones in construction increasingly depends on human operators' willingness to work collaboratively with robotic systems. However, limited understanding of how trust develops during human–drone interaction and its relationship to performance outcomes hampers the design of effective human–drone interaction frameworks and adoption strategies. Design/methodology/approach We conducted a within-subjects experiment with 41 participants who completed collaborative drone flight operations in a virtual reality environment. Trust was measured pre- and post-interaction across four dimensions: perceived risk, benevolence, competence and reciprocity. Objective performance was quantified via task completion time and collision count. Findings One-sample t-tests revealed significant increases in trust for benevolence (d = 0.35), competence (d = 0.44) and reciprocity (d = 0.43), with overall trust improving substantially (d = 0.48, p = 0.004). Larger increases in trust were associated with operational performance, inversely correlating with both completion time (r = −0.60, p < 0.001) and collisions (r = −0.71, p < 0.001). Among trust dimensions, perceived reciprocity was the most associated with objective performance in this task, showing significant correlations with reduced collisions (r = −0.32, p = 0.041). Originality/value Collaborative interaction fosters measurable trust gains in humans toward drones, with reciprocity, the perceived sense of mutual responsiveness, showing a significant association with performance in the VR-based inspection task. These findings offer preliminary insights that may help inform the design of collaborative drone systems that consider mutual responsiveness, the development of operator training programs focused on reciprocal interaction patterns and organizational strategies to accelerate autonomous technology adoption in construction.