Impact of different feedback conditions on worker's performance in carbon-fiber reinforced plastic manufacturing

Submitted by philipp on Thu, 08/20/2020 - 12:57
Abstract
Carbon fiber reinforced polymers (CFRP) are materials with excellent mechanical properties. However, slight variations in the manufacturing process lead to substantially decreasing mechanical stability and additional production costs for assuring the product’s quality. For large lot sizes a high process quality can be achieved through automation. However, for small lot sizes manual processes are required, which introduce variances in process quality. To support workers in the manual production however, there is little support to guide them and to reduce the variance in product quality. Augmented Reality (AR) applications are successfully used to offer guidance in other manufacturing processes. Yet, no empirical studies of the applicability of AR to support CFRP processing have been carried out. Therefore we present an AR based prototypic worker support system for CFRP manufacturing. Empirically, the impact of different feedback modalities (none, audio, visual, auditory+visual) as well as individual factors on effectivity, efficiency and user satisfaction was investigated. Key findings show that combined feedback worked best for accuracy, however no feedback was fastest. Regarding user preferences, auditory followed by auditory+visual feedback was rated best. Taking user preferences and performance into account will provide useful guidelines for the development of a support system.
Brauner, P., Bremen, L., Ziefle, M., Altorf, L., Rast, M., Rossmann, J.: Impact of different feedback conditions on worker's performance in carbon-fiber reinforced plastic manufacturing. In: Ahram, T., Karwowski, W., and Marek, T. (eds.) Proceedings of the 15th International Conference on The Human Aspects of Advanced Manufacturing (HAAMAHA): Manufacturing Enterprises in a Digital World. pp. 5087–5097. CRC Press, Boca Raton (2014).

Tags

Figures and Tables (selection):
Figure 1: Illustration of different variables and orientations. It is ?!=±90°, γ=±90°, and ?! =-45°, so ?!=45°. With ψ'(t)=120°, it follows that φ(t)=30° and hence ?!(t)=-15°. Values for ?! and γ are ambiguous due to symmetry.
Figure 2: Starting position of a task (base orientation ?=-45°) (left). Two additional carbon fiber cloths plates have been placed during a task (?!=90° and ?!=45° relative to base orientation) (right).
Figure 3: A user interacting with the prototype (visual feedback) (left). User placing a carbon fiber cloth layer on bass (right)
Figure 4: Mean completion time [s] of feedback type (left). Mean accuracy [°] of different feedback types (right)
Figure 5: User experience ratings of the three feedback modes (6 = completely agree).
Figure 7: Mean completion time [s] and mean User experience score of feedback types (left). Mean accuracy [in °] mean User experience score of feedback types of different feedback types (right).