Project website
On the advent of large displays equipped with multi-touch technology we are evolving to new, more natural ways of interaction with computers. Direct manipulation of content becomes the main interaction technique and is often more intuitive than traditional interaction techniques. Until now, most multi-touch approaches were too expensive or did not offer the required comfort or performance. Recently, more affordable tables have been constructed in research settings and there are companies that started building similar technologies for the consumer market. There are still some aspects that need to be researched to enable a breakthrough of this technology; the usability and usefulness in various settings, the software architectures that support application developers, appropriate interaction and visualisation techniques,...
The realization of the software architectur will involve building a collection of “multi-touch widgets”, in the form of user interface building blocks. In the remainder of the proposal we will refer to these widgets as MTWs. A MTW is a graphical interactive presentation that supports interaction using multiple touch points at once. Examples of such widgets are photo viewing widgets, video manipulation widgets, text entry widgets,... Each MTW can be used on its own, but can also be composed with other MTWs to create a custom front-end, or workspace, for a domain. The MTWs are inspired by the Web 2.0 widgets approach (e.g. Yahoo widgets, Google widgets, ...) that allow web developers (professional as well as non-professional) to compose a new application reusing these widgets that are (freely) available for that purpose. This is also known as the creation of a “mashup” that reuses and parametrizes existing components. With the widget set application developers can create user interfaces that allow to focus on a part of the data without losing the overall context (similar functionality can be found in zoomable UI toolkits), drill down on information items and get more detail, and allow to visualize relations between information items on display (Bederson, Grosjean, & Meyer, 2004), (Bederson, Meyer, & Good, 2000).
We envision these multi-touch displays being used in public areas, such as libraries, museum
exhibitions, arts centers or even train stations. In the demonstrator cases of this project we will focus on typical public spaces such as libraries, exhibitions... and investigate the ease of use and usefulness of these setups. Our expectation is that most interaction techniques that can be used in public spaces are also applicable in other contexts such as business meetings. A specific stadium that often lacks support nowadays is the “post-usage” stadium. Given the fact the user is interacting with a device installed for use by many different users, the results of a usage session should be made available afterwards to the user. There are several approaches to accomplish this. Nevertheless, the continuation of a session on later moment in time (persistence) remains to be a problem for a display that can be used by multiple users simultaneously and is part of our research.
In contrast with more traditional mouse-keyboard interaction, MTW are also more suitable to support simultaneous collaborative interaction. Multiple users can interact with the content on the same display. In general there are two approaches: the first one limits the amount of touch points that can be tracked but knows the users apart. This is the approach that is supported by the MERL DiamondTouch Table (Dietz & Leigh, 2001). The second one does not limit the number of touch points but can not tell what touch point belongs to which user. This is the approach used by Han's Interactive Table (Jefferson, 2005).
With the multi-touch display setup, users are allowed to explore information in a highly interactive way. To obtain fluid interaction real-time rendering of the graphical representations of large sets of data is required. A more natural look and feel also needs to be supported by smart physically based enhancements: interaction with virtual objects displayed on the screens simulates the behavior of their physical counterparts. A rendering engine will be created to encompass this functionality, and that can be easily reused for different domains and new custom workspaces.