Longitudinal Usability and UX Analysis of a Multiplatform House Design Pipeline: Insights from Extended Use Across Web, VR, and Mobile AR

Computer-Aided Design (CAD) software has long served as a foundation for planning and modeling in Architecture, Engineering, and Construction (AEC). In recent years, the introduction of Augmented Reality (AR) and Virtual Reality (VR) has significantly reshaped the CAD landscape, offering novel interaction paradigms that bridge the gap between digital prototypes and real-world spatial understanding. These technologies have enabled users to engage with 3D architectural content in more immersive and intuitive ways, facilitating improved decision-making and communication throughout design workflows. As digital design services grow more complex and span multiple media platforms—from desktop-based modeling to immersive AR/VR environments—evaluating usability and User Experience (UX) becomes increasingly challenging. This paper presents a longitudinal usability and UX study of a multiplatform house design pipeline comprising a web-based application for initial house creation, a mobile AR tool for contextual exterior visualization, and VR applications that allow full-scale interior exploration and configuration. Together, these components form a unified yet heterogeneous service experience across different devices and modalities. We describe the iterative design and development of this system over three distinct phases (lasting two years), each followed by user studies targeting different participant profiles and design maturity levels. The paper outlines our approach to cross-platform UX evaluation, including methods such as the Think-Aloud Protocol (TAP), standardized usability metrics, and structured interviews. Results from the studies provide insight into user preferences, interaction patterns, and system coherence across platforms. From both participant and evaluator perspectives, the iterative methodology contributed to significant improvements in system usability, reduced cognitive load, and a clearer mental model of the design process. Our findings highlight important considerations for future research and practice in the design of integrated, multiplatform XR services for AEC, but other domains as well.