Comparing Paper- and AR-Based Assembly Manuals in Task Performance,dlPFC Hemodynamic Responses,and Perceived Workload

Abstract

Objective

This study comparatively evaluated three assembly manual types (traditional paper-based, user-fixed AR, and world-fixed AR) in terms of task performance, dlPFC hemodynamic responses, and perceived workload.

Background

Prior studies of AR and traditional assembly manuals have relied mainly on subjective workload ratings, providing only an indirect view of cognitive effort. Neurophysiological assessment can quantify mental workload more directly. Measuring dlPFC activation may offer a process-specific index of cognitive control demands.

Method

Twenty-four participants performed a 7-min modified Purdue Pegboard Test with each manual type in a within-subjects design. Task performance, dlPFC hemodynamic responses, and perceived workload were assessed. fNIRS was used to measure hemodynamic responses.

Results

The world-fixed AR manual yielded significantly better task performance than both paper-based and user-fixed AR manuals, particularly in reducing location errors. Both AR manuals elicited significantly higher dlPFC activation than the paper-based manual, indicating elevated cognitive control demands. Perceived workload did not differ significantly across manual types.

Conclusion

The world-fixed AR manual improved task performance by reducing information access cost, yet both AR manuals increased cognitive control demands relative to the paper-based manual, likely due to perceptual conflicts inherent in AR. The user-fixed AR manual did not outperform the paper-based version, indicating that AR effectiveness depends on design specifics.

Application

The study recommends the world-fixed AR manual among the three types examined while emphasizing the need for innovations that reduce perceptual conflicts. It also shows that neurophysiological assessment provides insights beyond task performance and subjective ratings, supporting multidimensional evaluation to optimize interface technologies.