TY - JOUR
T1 - Evaluation of Augmented Reality instructions based on initial and dynamic assembly tolerance allocation schemes in precise manual assembly
AU - Zhang, Xiaotian
AU - He, Weiping
AU - Bai, Jilong
AU - Billinghurst, Mark
AU - Qin, Yunfei
AU - Dong, Jiepeng
AU - Liu, Tianyu
N1 - Publisher Copyright:
© 2024 Elsevier Ltd
PY - 2025/1
Y1 - 2025/1
N2 - Augmented Reality (AR) assembly instructions can be used to guide precise manual assembly, providing detailed real-time information to assist in executing operations. Most AR instructions for precise manual assembly are based on initial assembly tolerance allocation schemes (ATASs), which provide a precision constrained zone and require assembly according to the initial zone. Dynamic ATASs can adjust the subsequent tolerances according to the completed assembly errors, thereby reducing the assembly difficulty. However, ATASs do not affect the AR cues for assembly guidance, and their effects on the usability of AR instructions are subtle and unclear. This study focuses on evaluating the effects of initial and dynamic ATASs on the usability of AR instructions. An AR instruction based on an initial ATAS and an AR instruction based on a dynamic ATAS were developed for a precise manual assembly task. A user study was conducted with 16 participants to compare the two AR instructions. The results showed significant differences in assembly speed, assembly quality, perceived ease-of-use, and perceived workload. We found that the AR instruction based on the dynamic ATAS can expedite assembly with better perceived ease-of-use and lower perceived workload, but may reduce assembly quality. Based on the results, several recommendations were provided to help researchers and developers design better AR instructions for precise manual assembly.
AB - Augmented Reality (AR) assembly instructions can be used to guide precise manual assembly, providing detailed real-time information to assist in executing operations. Most AR instructions for precise manual assembly are based on initial assembly tolerance allocation schemes (ATASs), which provide a precision constrained zone and require assembly according to the initial zone. Dynamic ATASs can adjust the subsequent tolerances according to the completed assembly errors, thereby reducing the assembly difficulty. However, ATASs do not affect the AR cues for assembly guidance, and their effects on the usability of AR instructions are subtle and unclear. This study focuses on evaluating the effects of initial and dynamic ATASs on the usability of AR instructions. An AR instruction based on an initial ATAS and an AR instruction based on a dynamic ATAS were developed for a precise manual assembly task. A user study was conducted with 16 participants to compare the two AR instructions. The results showed significant differences in assembly speed, assembly quality, perceived ease-of-use, and perceived workload. We found that the AR instruction based on the dynamic ATAS can expedite assembly with better perceived ease-of-use and lower perceived workload, but may reduce assembly quality. Based on the results, several recommendations were provided to help researchers and developers design better AR instructions for precise manual assembly.
KW - Assembly tolerance allocation
KW - Augmented Reality
KW - Manual assembly
KW - Precise assembly
UR - http://www.scopus.com/inward/record.url?scp=85210037914&partnerID=8YFLogxK
U2 - 10.1016/j.aei.2024.102954
DO - 10.1016/j.aei.2024.102954
M3 - 文章
AN - SCOPUS:85210037914
SN - 1474-0346
VL - 63
JO - Advanced Engineering Informatics
JF - Advanced Engineering Informatics
M1 - 102954
ER -