|
|
© 1997 Copyright on this material is held by the authors.
Supermarket design requires careful prototyping of many aspects like layouts and routings, visual break throughs, and commercially attractive overviews of product groups. Traditionally the managers responsible for the implementation of new and renewed stores make their decisions by evaluating the conceptual design of a store based on drawings made and explained by architects during extensive sessions. An important question is if the design process of stores may be accelerated or improved when one can prototype the conceptual design of the stores using virtual reality.
Store Designer Pro is a tool that supports prototyping stores in VR. The user can create and walk through various designs of the store. The research question is whether this prototyping in VR is a plus over design using traditional media or not, and on which attributes and constraints possible improvements may depend.
An experiment is reported concerning the behavior of domain experts using a VR tool to evaluate and redesign a given store design made by architects. The experiment is a pilot study with real domain experts from several levels in the organization of a large supermarket company.
Especially for the creative processes of design and evaluation - both essential within prototyping - as realistic as possible feelings of telepresence in VR are necessary [1, 2]. These principles are implemented in the tool Store Designer Pro (SDP). With SDP domain experts are able to design their own store or supermarket in virtual reality without programming. SDP runs on the standard Pentium PC platform with many VR devices.
The user can walk through various designs of the store. This enables the user to evaluate the designs, layouts and routings and determine which is the most suitable.
SDP is built around a virtual reality engine. A user-interface is realized which enables the user to perform all actions, see figure 1.
Figure 1. VR design and user interface in SDP.
The system includes an object library. In the library items like fixtures (shelves), cash registers, refrigerators, check-ins, etc. are available. A tailor-made library with objects can be created per customer. In addition to the normal objects there are also special objects in the library such as cameras and mirrors. These can be placed anywhere in the store. The picture which is captured by the cameras and mirrors is displayed on a virtual monitor in real-time.
Two methods of navigation are possible: via plan-view or via 3D-perspective-view. While in plan-view it is possible to zoom in and out into specific areas of the store. In 3D-perspective-view it is possible to walk through the store as a human being. The human has normal gravity and collision detection, so he cannot walk through objects.
The goal of this inductive case study is to compare the process and the opinions - not the final store designs - of expert store designers and managers using different media. The subjects had to evaluate two different stores using traditional drawings, using a VR-representation, and finally by visiting the real existing stores, subsequently. The stores were totally unknown to the subjects.
The eight subjects were all professionals with different roles in the supermarket company with experience in evaluating store designs using traditional drawings.
Two supermarkets each about one year old were selected. The detailed architectural drawings were updated to the current state of the supermarkets. The same two stores were represented as exactly as possible in VR using the SDP tool. The tasks of the subjects were to discuss and to evaluate the characteristics of the designs represented in each medium during a session of about 30 minutes. An architect participated in the sessions to explain the context of the city, the store, and to answer technical questions during the sessions. The VR presentation was projected on a spheric video screen of about 1.5 square meters. The navigation within the store was done on demand of the subjects, by a technical assistant. After these in-house sessions the subjects moved to the two real existing stores. In real stores again a session of 30 minutes took place by walking through the supermarket. The stores were open to the public but only a few regular consumers were present in the stores during these sessions.
The experiment was explained to the subjects verbally immediately before the first session. All discussions were recorded on tape. After each session the subjects filled in the standard questionnaire. After these four (2 stores x 2 media) in-house sessions the group moved to the two real existing stores.
The questionnaire consists of 30 questions for subjective answers on a 5-point scale. Some questions concerned commercial and spatial viewpoints; others asked for opinions on the possibilities of VR for more extensive prototyping.
| medium | SM | CM | staff | |||
|---|---|---|---|---|---|---|
| s | c | s | c | s | c | |
| drawings | 3.0 | 3.6 | 2.8 | 2.8 | 3.6 | 3.1 |
| VR | 1.9 | 3.5 | 2.2 | 3.2 | 1.9 | 3.2 |
The three categories of subjects store managers (responsible for one supermarket), cluster managers (responsible for about 50 supermarkets) and staff from the central department of store technology are presented in table 1. More details of this pilot study are reported in [3] and [4].
The results indicate a not significant positive outcome for the VR representation over the drawings. Comparing the spatial versus the commercial questions it is clear from the data of this pilot study that VR supports the evaluation of the spatial features very good. This may be due to the fact that with traditional methods (drawings and mock-ups) the design is not experienced from a human eye-position as it is with VR. The smaller improvement of evaluation of the commercial features may be caused by the relatively poor rendering of textures in this first pilot, see figure 1. The stores are not yet "glossy". Improvement will be necessary.
From the recordings of the discussions can be concluded that the VR representation using SPD causes strong involvement by the participants. Especially the interaction and walking are appreciated. The moving through narrow areas, the influence of columns for the effective use of space is told to be evaluated better using VR.
The authors wish to thank the staff and the subjects from Albert Heijn Supermarkets in Zaandam for the possibility to carry out this experiment.
|
|