Human responses to augmented virtual scaffolding models

Abstract

This study investigated the effect of adding real planks, in virtual scaffolding models of elevation, on human performance in a surround-screen virtual reality (SSVR) system. Twenty-four construction workers and 24 inexperienced controls performed walking tasks on real and virtual planks at three virtual heights (0, 6 m, 12 m) and two scaffolding-platform-width conditions (30, 60 cm). Gait patterns, walking instability measurements and cardiovascular reactivity were assessed. The results showed differences in human responses to real vs. virtual planks in walking patterns, instability score and heart-rate inter-beat intervals; it appeared that adding real planks in the SSVR virtual scaffolding model enhanced the quality of SSVR as a human – environment interface research tool. In addition, there were significant differences in performance between construction workers and the control group. The inexperienced participants were more unstable as compared to construction workers. Both groups increased their stride length with repetitions of the task, indicating a possibly confidence- or habit-related learning effect. The practical implications of this study are in the adoption of augmented virtual models of elevated construction environments for injury prevention research, and the development of programme for balance-control training to reduce the risk of falls at elevation before workers enter a construction job.

Keywords:

• Virtual reality
• Tactile augmentation
• Height
• Construction
• Plank

Acknowledgments

The authors gratefully acknowledge the support of the entire research team, including Doug Cantis, Rotha Hall, Darlene Weaver, Joyce Zwiener, John Powers, Paul Keane and Scott Hendricks. A thank you is also extended to Drs. Mark Redfern (University of Pittsburgh), Amit Battachaya (University of Cincinnat), Stephen Ellis (NASA), Larry Hodges (Georgia Institute of Technology) and Ronald Mourant (Northeastern University) for providing review comments on our research protocol.