Video Wall Displays

Introduction

In Mark Weiser's vision of ubiquitous computing, he mentioned three different classes of devices - tabs, pads and boards, with different sizes and capabilities. The video wall display is an example of a board, maybe even more. In this paper, I will first describe what a video wall display, its classes of users, and potential applications. Then I will present the technical challenges in the development of video wall displays, and their commercial availability. Finally, I will conclude with a brief discussion of the possible impact that video wall displays might have on our lives.
 

What are video wall displays?

This is not a dumb question. Large displays have been around for a very long time. The most common examples are movie projectors and the large video advertisement billboards you see on tall buildings. More recent examples of large displays come in the large, flat screen, high-definition TV screens that may one day be in every modern home.

Obviously, video wall displays must be much more than just a display technology. Perhaps a better description would be "interactive wall displays". There are some properties (not exhaustive) that video wall displays share, and they are:

• Scalability - Many video wall displays are composed of an array of screens, partly to save on cost, but also so that they are scalable, Video wall displays can be assembled in various configurations to fit the needs of any room, and can be reconfigured when the users' requirements change.
• Interactive - Without the ability to interact with the screen, video wall displays would just be yet another big TV. Large interactive displays are precisely what makes this a new and exciting technology that opens up a wide range of potential killer-apps.
• Dynamic configuration - To be truly a ubiquitous computing technology, video wall displays must work in conjunction with other input and computing devices. Without this ability to connect and communicate with other machines, the video wall display will just simply be a workstation with a really huge screen.

Users

Having defined what video wall displays are, we come to the question of who are the potential users? It depends on what the application is. Here are some examples of possible user groups:

• Group or collaborative work in an office or conference environment - The large screen allows the ability to present a lot of data, and the ability to connect your laptop to use the display or part of the display, makes this a wonderful tool for sharing data and for collaborative work. Group members can see and manipulate the data on the same screen during the meeting and they can take home the contents of the meeting when they are done.
• Scientific, medical or mathematical visualization - Many scientific visualization applications require a lot of screen real-estate. Scientists and other technical professions can also take part in collaborative work with video wall displays.
• Architectural design and building walkthrough - Architects and their clients can work together in a more immersive environment than a small PC display. The walkthrough can be rendered life-sized and there are advantages over VR headsets in that both parties have the same view of the screen.
• Military applications - Commanders can see the entire battlefield situation, communicate with their troops in the field, and plan their next step.
• As a dynamic display system, for entertainment, learning or information - Applications include interactive bulletin boards, interactive information displays for visitors, or simply for entertainment. The user group in these cases is then the general public. Another possible application is in video games. Hmm, life-size Quake anyone?
• E-commerce applications? - One might envision huge interactive advertisements (you need the size to attract attention) that people can walk up to and interact with to find out more about a product, and maybe even buy directly from the display.

Video wall displays would not be suitable for traditional PC tasks such as word processing and file management. The huge display would be an overkill for simple text input, and might even be a distraction as the field of view is focused on only a small area of the screen for word processing applications. Switching betwwen the keyboard and pointing device may also be a pain. In general, single-user tasks that work well on a desktop will probably not benefit much, if any, from video wall displays. In some cases, they may even degrade user performance.
 

New or improved affordances

Most of the above classes of users have one common need, and that is the need for multiple persons to work collaboratively within the same physical space. The large interaction area provides the screen real-estate often necessary for such tasks. All users can see all the various data and views at the same time on the same screen. Networking capabilities also allow users to present their own data on the screen to share with members of the group. Working within the same physical space also affords a more directly manipulative interface, in contrast with current distributed collaborative software on multiple desktops. Users can manipulate data directly on the screen. The large screen allows for life-sized rendering of virtual environments, thus providing a more immersive experience without the need for expensive VR immersive systems.
 

Interaction styles and metaphors

It is generally agreed that traditional input devices used on the desktop like the mouse and keyboard are not suitable on a large wall display. This is due to the fact that the users are usually standing and walking around rather than sitting at some terminal. Furthermore, the mouse is designed to move over a short distance and does not lend itself well to controlling a pointer over a huge display that can span the entire room.

For many collaborative applications, a more suitable style of interaction would be by touching the screen directly, either with the hand or with a stylus. The metaphor here is that of the whiteboard. Many techniques exist to enable this mode of interaction. The ability to manipulate the data directly rather than through a physical device is a major contributing factor to the feel of direct manipulation on a video wall display.

One of the goals of ubiquitous computing is the ability to dynamically support different devices, so other interaction styles must be supported. The user should be able to use the pointing device and keyboard on his laptop to interact with the screen. Therefore, interaction on a video wall display will most likely be multi-modal. A lot of research is being carried out to discover other modes of interaction (for example, the Princeton "magic wand") and in supporting and enhancing the use of traditional input devices on video wall displays (for example, the "Interactive Workspaces" project at Stanford).

A desktop-style GUI might not work as well on a large wall display. For example, menus do not work as well because they are usually at the top of the application. On a wall display, they might not be reachable by shorter people. The same goes for many desktop GUI widgets. Depending on the input technology, dragging things around on the screen, and fine selection and control of GUI objects can be a frustrating experience.

Voice recognition is being looked into as a form of hands-free interaction, and as a solution to the problem of the traditional GUI. Gestures are also being researched as an input method.
 

Technical challenges

There are many technical hurdles to overcome in the development of video wall displays:

• Screen resolution - Constructing a large one-piece display is hard and expensive, so most video walls are made up of an array of screens lit from the back by projectors.
• Seamless images - The projectors have to be aligned and calibrated to provide a seamless image over the entire display. Variations in image quality and color is also a problem, so these properties must also be calibrated.
• Cost - Reducing cost without compromising image quality. Research is being done at Princeton on using ordinary projectors to build a cost-effective video wall.
• Parallel/Distributed architecture and load balancing - Due to the large display area, rendering images with a single workstation would be very inefficient and slow, so a network of workstations have to be used to distribute the workload. The same argument applies to the running of applications, handling of input devices, etc.
• Sound - Cheap, immersive spatialized 3D sound technology capable of handling multiple input streams.
• Input devices - Problems include the design and construction of specialized devices and supporting traditional input devices, design of gestures, camera-tracking technology, voice recognition technology, etc.
• Networking - Protocols and mechanisms for the above, dynamic configuration capability, communication with laptops and PDAs, etc. Other problems involving networking include collaboration over tele-connected spaces, where remote users can interact with the video wall to communicate ideas to the people in the room.
• Software design - Support for legacy applications, and the design of new applications that take advantage of the new architecture. Development of UI toolkits. Usability issues.

Current commercial availability and research

All the basic hardware, from servers to projectors to screens, are readily available. What is lacking is the software and the need to construct suitable rooms. Commercial solutions do exist however, but they don't come cheap. Some companies that offer video wall solutions are Fakespace, Trimension, SGI and MechDyne. From what I can tell, their main focus is on entertainment and scientific visualization applications.

Princeton is doing research on making video walls more affordable, and in new applications and input devices. Stanford's Interactive Workspaces project is doing a wide range of research on video walls, focusing more on the architectural and user-interface issues. There are also many joint projects with other departments to develop video wall applications.
 

Social and economic implications

Reducing the costs of building and deploying a video wall display is the key to their widespread use in offices and conference rooms. Video walls will allow groups to work and collaborate more efficiently. Video conferencing will be greatly enhanced. As an entertainment and learning medium, video walls can provide a much more immersive experience.

However, the ability to dynamically attach laptops and PDAs to a video wall for sharing information means that privacy and security will be important issues. The video wall is like a mini-network and can be connected to even larger, even global networks through these connected devices. It might be possible pilfer company secrets through a video wall display. We may find one day that video walls will be subject to hacker attacks. With everything interconnected, it might even be possible to track a person as he make his way from meeting room to meeting room as he log into the video wall at each location. The security and privacy issues pertaining to video walls are the same as those for ubiquitous computing devices in general. Research will have to be done to make video walls more secure and to protect the users' privacy.
 

Online resources

I obtained the following excellent list of sites from Princeton's Scalable Display Wall web site (http://www.cs.princeton.edu/omnimedia) and I think it is a good list to start with so I'll just reproduce it here.

Related Research Projects

• Interactive Workspace, Stanford Interactive Workspaces Project
• Telepresence, UNC
• Workspace of the future, IPSI - Darmstadt
• Active Mural, Argonne National Labs.
• Ubiquitous Computing, Xerox PARC
• Digital Office Project, Xerox PARC
• PowerWall, University of Minnesota
• Electronic Visualization Lab, University of Illinois at Chicago
• Interactive DataWall, ROME labs
• DataWall, MIT Media Lab
• Virtual and Collaborative Environments, ANL

Commercial Immersive Systems

• HP's Immersive Workspace
• SGI's Reality Center
• Fakespace
• Trimension
• SEOS Display Limited
• Panoram
• Pyramid
• MechDyne
• Alternate Realities
• Concurrent Technologies