QUALITATIVE COMPARISON OF VIRTUAL AND AUGMENTED PROTOTYPING OF HANDHELD PRODUCTS

In recent years, powerful computer aided design engineering tools have been introduced to realize so-called Virtual Prototypes: digital displays of the product in a simulated environment, offering various evaluation and modification means. However, virtual prototyping of physical products will not always produce a representative experience. For example, Kuuti et al. [2001] describe the limitations of concept testing with a screen-based 3D browser: 1) product is not shown in its intended context but floating in the air, 2) users easily get lost in 3D space (navigation), 3) the sense of scale and the ability to test physical ergonomics is difficult yet impossible. Similar shortcomings can be found in Bochenek et al. [2001], in which the performance of four different VR displays – head mounted displays, Fake Space Boom, stereoscopic screen, and monoscopic 2D screen based VR - is compared in a design review setting. Often, the monoscopic monitor-based systems performed best and were preferred by the subjects. Furthermore, it was observed that the overwhelming technology introduced a positive bias towards judging the actual design The concept of Augmented Prototyping combines Rapid Prototyping techniques to obtain 3D physical objects (e.g. Stereolithography, CNC milling), with Augmented Reality systems [Verlinden et al., 2003a]. The aim is to establish a high sense of engagement in the design process, supporting both exploration and presentation. We have documented some successful explorations with the shader lamps technique [Raskar and Low, 2001], spanning from the design of cars [Verlinden et al., 2003a] to interiors [Verlinden et al., 2004] and handheld devices[Verlinden et al., 2003b]. With this technique, 3D perspective images are projected directly on the object. The mapping of virtual 3D object and projected image depends on the viewer’s location, shape of the physical object/screen, and the projector parameters. If there is a direct correspondence between virtual object and physical surface, the projector can be treated as the inverse of a pinhole camera. This reduces the complexity of the pre-distortion to a simple projection matrix. In that case, standard 3D rendering systems can be employed. A software architecture was devised labelled WARP (Workbench for Augmented Rapid Prototyping), see Figure 1. Previous systems based on this architecture allows the projection of several details, materials, and colours on physical models; different 2D and 3D locatio5n tracking devices were employed to interact with the prototype. In considering the case studies and prior research systems, four different design support types for augmented prototyping have been identified: 1) component layout, 2) material and colour selection, 3) interaction prototyping, and 4) engineering simulation. Although we hypothesize that augmented prototyping has a large potential within the field of industrial design, a number of issues remain uncovered. Main concern is determining the added value of such prototyping means.