Users’ evaluation of a virtual reality architectural model compared with the experience of the completed building Bo ¨rje Westerdahl * ,1 , Kaj Suneson 1 , Claes Wernemyr 1 , Mattias Roupe ´ 1 Mikael Johansson 1 , Carl Martin Allwood 1,2 3 The Visualisation Studio, at the Department of Civil and Environmental Engineering, Chalmers University of Technology, Go ¨teborg, Sweden Accepted 4 February 2005 Abstract It is important that the properties of a Virtual Reality (VR) model support its purpose in the specific context in which it is used. This study investigated how employees of a company experienced a VR model of their yet-to-be-built office building. We also compared the VR experience with the employees’ experience of the completed building. The results showed that the employees felt that the VR model was a useful aid in the decision-making process concerning their future workplace. In addition, The Semantic Environment Description Scale was used to compare the experience of the VR model with the experience of the real building. The results suggest that the VR model gave a fairly accurate representation of the real building. D 2005 Elsevier B.V. All rights reserved. Keywords: Virtual Reality (VR); Usability; Architectural model; Realism; Experience 1. Introduction The Virtual Reality (VR) technique, i.e. digital visual- isation of 3D content in real time, has evolved from being a very expensive technique implemented on supercomputers handled by experts to a technique implemented on non- expensive PCs affordable even for small firms. However, the development of the technique has not been matched by research aimed at adapting the features of the VR technique to the different demands of the various situations in which it is used. The usability issue (taken in a broad sense) is thus of great importance in the context of VR. In the present study we investigate the usability of a VR model of a building and its environment at a late stage in the building process. In general, research in Human Computer Interaction (HCI), including the concept of usability, has undergone a shift from being computer- and program-oriented to being more focused on the actual use context (e.g. [1,7,8]). However, as noted above, the concept of usability, although central in HCI research in general, has not been explored very much in the context of VR models. An exception is Kalawski [9], who developed general criteria and a general instrument for measuring the usability of VR models. It is important that the properties of the VR presenta- tion are adapted to the specific use context, e.g. the specific stage in the building process in which it is used, including the different categories of users using the VR model. For example, perfectly realistic lighting in a VR model would presumably be an excellent experience, but may not be worth the effort. Questions arise concerning how and why VR is used in the building production system. Even today most of the design work in urban planning and architectural projects is done in 2D. The main tools used for communication between co-workers in the different 0926-5805/$ - see front matter D 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.autcon.2005.02.010 * Corresponding author. Box 8873, SE-402 72 Go ¨ teborg, Sweden. E-mail address: borge@chl.chalmers.se (B. Westerdahl). 1 Correspondence can be directed to any of the authors. 2 Equal responsibility assumed by all the authors. 3 Allwood is also at the Department of Psychology, Lund University, Sweden. Automation in Construction 15 (2006) 150 – 165 www.elsevier.com/locate/autcon