Conclusions

Having a GIS available on a large-scale, wall-mounted device enables users to interact with geographic information in completely new ways. This paper explored user-interface considerations for the WallBoard, an organizing metaphor for wall-size GIS devices. Basing the design on the metaphor of a whiteboard, affords most users an immediate idea of how the WallBoard can be used. Multi-modal interactions with the WallBoard occur primarily through a combination of gestures, eye-movements, and voice. Different modalities are used depending on how close or remote a user is from the WallBoard. The availability of a certain interaction mode leads to a categorization of interaction spaces, in which users perform different types of tasks. Most challenging is the use of hand gestures to perform some of the most common GIS operations, such as selecting, panning, zooming, but also for such innovative interactions as navigating through an animated 3-dimensional scene. Such interactions are dramatically different from those used for mouse-based panning and zooming on current desktop GISs (Jackson 1990). When users stay further away from the WallBoard, they lack immediate contact with the WallBoard, but still can select objects through gestures and voice. The farther away from the WallBoard, however, the lower the accuracy of their empty-handed gestures. To compensate for distance, remote users may use hand-held pointing devices such as a laser pointer for selection. Instead of contact, the selection of individual objects is made through pointing at the same location for a certain time interval.

We have simulated interactions with a WallBoard through an animated movie on a small-scale display, using MacroMind Director on a Macintosh. Gestures were derived from interactions with a whiteboard. The study was invaluable to identifying the nature of interactions possible with a GIS operating on a wall device, and the enhanced possibilities for collaborative spatial decision-making.

The WallBoard is a spatial technology that provides a framework for studies of innovative GIS interactions, allowing for entirely new approaches to problems. The mere existence of this concept will advance our knowledge of interaction methods with GISs through comparisons with the often-so-limiting constraints of today's GIS desktop environments. At the same time, the WallBoard concept is expected to serve as a framework for specific questions related to multi-modal interactions. The design of the WallBoard is certainly only the starting point and there is a long way to go before prototyping and user testing with a comprehensive scenario. Significant theoretical advancements will be necessary to enable smooth group work with the WallBoard.

Some of the open questions include:

• What hand gestures can be identified sufficiently precise in real-time?
• When do users combine voice and gesture, and how?
• In an experiential space such as the WallBoard, is it desirable to provide simultaneously multiple views, and if so what alternatives to windows integrate well with the WallBoard?
• Which new tools beyond lenses and measuring devices will enable innovative analysis methods on the WallBoard?
• What are the requirements for a spatial database management system to support real-time manipulations on the WallBoard, including choice of data models and data structures?

Last updated on January 31, 1996.