Evaluating Multilingual User Interfaces for AR+BIM Construction Training

Abstract

Augmented Reality (AR) interfaces integrated with Building Information Modeling (BIM) are increasingly used to support construction training and on-site tasks. Existing AR+BIM research has largely emphasized visual augmentation and spatial guidance, with limited attention to linguistic accessibility as a user interface design factor. This gap is significant in the U.S. construction industry, where many workers are native Spanish speakers who routinely interact with technical systems in a second language. This paper investigates how user interface language influences performance and cognitive workload in AR+BIM-assisted construction training. We conducted a controlled study in which Spanish-speaking participants completed a multi-step electrical installation and modification task using identical AR+BIM interfaces presented in either Spanish (L1) or English (L2). In addition to completion time and procedural accuracy, we evaluated subjective workload, usability, and performance variability while accounting for individual bilingual language differences. By framing language as an interaction variable in XR interface design, we introduce a linguistically grounded perspective for evaluating AR instructional systems. We show that first- and second-language interfaces influence performance consistency and cognitive strategy use during AR-assisted construction tasks. This highlights the importance of considering language when designing and evaluating multilingual AR+BIM interfaces.