Videoconferencing 1990s Style:
Sharing Faces, Places
ONTARIO TELEPRESENCE PROJECT
List of figures ............................................................................................. ii
Preamble .................................................................................................... 1
Part I: An Overview of the Issues in Videoconferencing Today
Introduction .................................................................................. 2
The development of videoconferencing technologies ......... 6
Why were forecasts in the 1960s and 1970s so inaccurate? ... 8
Why should we expect things to be different today? ............. 9
What does video really add? ....................................................... 12
Part II: Media Space - a New Approach
Media space ................................................................................... 15
Summary: Part I and II ............................................................................ 17
A Comparison of Two Systems
Description of the field study ..................................................... 18
Description of VISIT .................................................................... 20
Description of TMS ...................................................................... 22
Use of the two systems ................................................................ 29
Comparison of the two systems ................................................ 34
Issues Emerging from the Study ........................................................... 41
Conclusion ................................................................................................ 44
Bibliography .............................................................................................. 46
List of Figures
Figure 1 A Conference Room Facility ......................................... 3
Figure 2 Roll-Around Units .......................................................... 4
Figure 3 A Desktop Unit ................................................................ 5
Figure 4 A Picturephone ................................................................ 7
Figure 5 Portholes .......................................................................... 13
Figure 6 VISIT ................................................................................ 21
Figure 7 TMS: Telepresence Media Space ................................. 24
Figure 8 CSG-TMS System Block Diagram ............................... 25
Figure 9 Doors on the TMS Interface ......................................... 26
Figure 10 Software Doors ................................................................ 26
Figure 11 Use of Multiple Cameras/Monitors ........................... 27
In the fall of 1992 Provitel, Inc., a telecommunications organisation, established a working relationship with the Ontario Telepresence Project (OTP). One of the outcomes of this collaboration was the installation of a Telepresence Media Space (TMS) system in May 1993 between the offices of two members of CSG, a client support group in the organisation. At the same time VISIT systems were being deployed to a number of members of CSG including the participants with the media space. This provided a unique opportunity to evaluate these two systems - one a commercial product, i.e.,VISIT, the other, a futuristic prototype system, i.e., TMS.
Part I of the report provides an overview of videoconferencing in the 1990s and a discussion of the factors that distinguish the 1990s from the 1960s and 1970s when the first wave of videoconferencing technologies failed to have the predicted impact.
The concept of a media space is introduced in Part II and the Ontario Telepresence Project model of research and development is described.
Part III reports on the experiences of the early adopters of videoconferencing in CSG and provides a comparison of the two systems used. The focus is not technical, but rather on gaining early insights into the impact of and opportunities created by this new class of technology. The results demonstrate the need for:
* an ecological approach to understanding videoconferencing technologies. Successful deployment, adoption, continued use and growth of the technology require a 'goodness of fit' between the organisation's culture(s) and the work practices of employees.
* a better understanding of the role of video. Image quality and size, and image location are all important factors affecting use. In addition, the ability to make a distinction between background awareness and foreground interaction is highly valued. The primary point to be made is that bandwidth on demand - the ability to scale video images dynamically - coupled with a price structure that is affordable - (and unlikely to be a linear relationship between bandwidth and cost) - presents both a major challenge and a major market opportunity for the network provider positioned to understand the potential of video.
Videoconferencing 1990s Style: Sharing Faces, Places and Spaces
Gale Moore - Ontario Telepresence Project
Ken Schuyler - Provitel, Inc.
This document was written by Ken Schuyler and Gale Moore using the Telepresence Media Space (TMS) system. The authors were physically separated, but shared a virtual office - that is, they could see and talk to each other and share the document on which they were working. The technologies used were TMS and a collaborative writing tool - SASSE - developed in the University of Toronto's Department of Computer Science.
The paper is divided into three parts:
* Part I: An overview of the issues in videoconferencing today
* Part II: Media space - a new approach
* Part III: Impressions of an early adopter: a comparison of two systems
Part I An Overview of the Issues in Videoconferencing Today
Today we classify commercially available videoconferencing systems into three distinct types: conference rooms, roll-around units and desktop units.
1. Conference Rooms
There are two types of conference room facilities:
i. Commercial teleconferencing facilities
Teleconferencing facilities are available for lease from a third party, typically a telephone company, and provide the user with a meeting room equipped with high end commercial equipment and the technical support and services required to operate it. In general, these facilities offer high quality video and audio, are suitable for a specific clientele, and for specific types of meetings. Inherent disadvantages lie in the fact that the room design cannot be readily modified to accommodate a variety of meeting types or presentation styles. In addition, use may be discouraged by the fact that meetings must be planned in advance, bookings made and honoured (charges may be incurred on default), and local travel which may be both inconvenient and costly is frequently involved.ii. Corporate teleconferencing rooms
In recognition of the advantages of teleconferencing, the inconvenience of commercial sites, and the declining cost of teleconferencing systems, many corporations have built in-house facilities. While these conference rooms overcome some of the problems encountered in the commercial sites, they also tend to score low on flexibility, and make many of the same assumptions about use as the commercial systems. For example, formally planned meetings of a particular type may be well-supported, but informal meetings or meetings of short duration are not easily accommodated. In addition, travel, while more limited, is often still involved.
Fig. 1 A Conference Room Facility
2. Roll-Around Units
Roll-around units such as those available from PictureTel and CLI, are designed to be moveable, and provide a degree of flexibility in terms of where they can be used within a building. However, while moveable, the units are large and what is possible may not be practical. In some cases roll-around units have been built into corporate conference rooms - a design which restricts usage to a single location!Fig. 2 Roll-Around Units
3. Desktop Units
A more recent development has been desktop videoconferencing. An example of this approach is VISIT developed by Northern Telecom. This type of system provides the advantage of desk or office access, but at the present time the video is low quality which severely limits its usefulness as a videoconferencing system. In the VISIT system, for example, the screen sharing and file transfer features are far in advance of the videoconferencing features. While the desktop units suggest flexibility, the locations in which they can actually be deployed today are limited by telephone network capabilities - i.e., it requires 56 Kbps, and it is not possible to use this technology in all organisations or in the home offices of many teleworkers.
Fig. 3 A Desktop Unit (CLI-Cameo)
The Development of Videoconferencing Technologies
The concept of videoconferencing is not new. Picturephone, a type of desktop videoconferencing system, has been under discussion at Bell Labs since the 1920s and videoconferencing rooms have been in existence at AT&T since the 1960s where they were used to support large corporate meetings such as the annual shareholders meeting (Egido,1988,14).
At the 1964 World's Fair, the picturephone was introduced to the public with the prediction that it would replace the voice telephone by the early 1970s (Egido,1988, 14). In spite of the failure of such optimistic forecasts, enthusiasm about this technology continued. The economic recession of the 1970s and the oil crisis were seen by the champions as an impetus for corporations to reduce spending on travel and choose videoconferencing as an alternative. But the technology was not widely adopted and there is consensus that the first wave of videoconferencing was not a success (Egido, 1988, 16).
Fig. 4 A Picturephone (Oki)
As we move toward the second wave of videoconferencing technology, three key questions must be asked at the outset.
* Why were the forecasts in the 1960s and 1970s so inaccurate?
* Why should we expect things to be different today?
* What does video really add?
Why were the forecasts in the 1960s and 1970s so inaccurate?
As Egido convincingly argues, the needs assessment methodologies used to make these predictions were woefully inadequate (Egido, 1988, 16). Even more significant is the fact that the failure of early videoconferencing systems was not primarily technological or economic in nature, but a failure to understand and to take into account the sociological and psychological factors involved in the deployment and diffusion of this technology. The pioneers in the field had conceptualized videoconferencing as a direct replacement for face-to-face meetings or encounters (Egido, 1988,16). In terms of the corporate bottom line it was anticipated that videoconferencing could replace to some extent both long distance travel and commuter travel. The result would be a direct saving of time, money and energy. Reality, however, was more complex than this simple equation could explain. The recognition of the importance of factors such as corporate culture, the nature of the work group, and work practices was in its infancy. In the 1990s the incorporation of sociological and psychological perspectives in the development and deployment of new technologies is still limited, but there are examples of research and product development sites where these concepts are taken seriously, e.g., Xerox Corporation, Ontario Telepresence Project.
Why should we expect things to be different today?
There have been major changes in two areas since the first wave:
1. Technical change
In the past twenty years there have been major developments in the computer and electronic technologies central to telecommunications. The results have led to substantially more cost-effective videoconferencing, with enormous improvements in quality. The technology changes have been in all facets of the application:
1) huge increases in computing power at very low cost and physical size have made it practical and inexpensive to implement the coding/decoding/compression algorithms needed for digital video transmission,
2) major deployment of digital transmission equipment, with higher and higher data rates, and reduced error rates have made it possible to transmit compressed digital video reliably over long distances, and
3) standardisation of digital transmission services and video coder/decoder (codec) interoperability have allowed the shipment of digital video over telecommunications systems already in existence and between brands of equipment.
Taken together these factors have lead to a level of quality of video and pricing of services that was not possible during the first wave of videoconferencing.
2. Large Scale Social Change:
On the social side, two broad areas of social change, reflected in organisations, are particularly significant to the success of videoconferencing technologies in the 1990s.
i. Organisational redesign and re-engineering
Until the 1970s there was widely accepted assumption in the business literature that there was an optimal form or structure for organisations. To be successful an organisation had to adopt the structure 'known' to be optimal for its size, and production type. Similarly in the development and deployment of computer and communication systems a 'one size fits all' model of design and production was used - with efficiencies coming from standardization, centralization, and mass production. It is now clear that it is crucial to find the 'fit' between an organisation's culture, work practices and technology. In addition, it is crucial to follow up with appropriate implementation strategies. As Bikson has demonstrated in the case of the deployment of a variety of office technologies, there are 3 classes of hypothesized influences on the successful implementation, namely:
* the characteristics of the organisational setting
* the features of the technology, and
* the implementation strategies employed (Bikson, 1987,156).
ii. Changes in the labour force:
Information and knowledge workers now dominate the labour force. There are no precise statistics to support this observation because the official Canadian statistics do not classify the labour force in this way. We do know, however, that by the mid 1960s the number of white collar workers had surpassed the number of blue collar workers in Canada, the United States, and many western European nations (Moore, 1990, 60). This shift from an industrial to a service-based global economy has major implications both for individuals in the work force and for employers. The average white collar employee today, whether a clerical, professional or technical worker, is increasingly familiar with computers and a variety of communication technologies such as voice mail and email. This suggests that the potential users of videoconferencing systems will now be more technically literate than in the 1960s and early 1970s.
What does video really add?
This has been considerable debate in the research literature on this question. While this remains an important question, some generalizations can be made.
1. Increased social presence
Video-enhanced communications are an information 'rich' media and a valuable resource in mediating interpersonal interaction (Tang and Isaacs, 1992, 24). Video provides eye contact, the ability to exchange gestures and other visual cues which provide feedback on the communication - e.g., whether it has been understood, whether there is agreement, etc. It is more like 'being there' than other communication media such as the telephone or email.
2. Support of formal communication
The availability of a visual channel facilitates some types of collaborative work. For example, it is possible for employees who are away from the office to attend meetings, or to handle tasks that require a degree of physical presence such as negotiation.
3. Support for informal communication
While the development of new relationships is best done in person, face to face, videoconferencing helps maintain and build relationships. This is particularly important for informal communication. Users report, for example, that video capability makes their interactions more satisfying (Tang, 1992, 24). In addition, informal communication can be enhanced through background awareness services such as Portholes and Cruiser[13.] These services provides an awareness of colleagues or co-workers that is similar to that experienced by those working in close physical proximity - in offices along the same corridor, for example. This awareness encourages the spontaneity of interaction and lowers the opportunity cost for brief interactions.
Fig. 5 Portholes (Background Awareness)
It is too early to determine whether video has an effect on productivity 14 but as John Seely Brown, chief scientist and director of Xerox Palo Alto Research Center (PARC) has commented: "the inability of economists to prove the value of information systems may have more to do with our present understanding of productivity than with any real offset between the cost of information systems and their inherent value to an organization" (Brown, 1994,17).
Part II Media Space - A New Approach
The media space concept provides a new way of thinking and moving about in time and space. A media space is essentially an environment in which audio, video and computer technologies are integrated in novel ways to provide new ways of interacting and working together. Participants are able to actively share their presence and a common work space, as well as having a background awareness of others that parallels the awareness of colleagues experienced in the traditional office.
The Ontario Telepresence Project (OTP) is one of about four sites in the world where research on media space is being carried out. The OTP is innovative in several ways:
* it is the only project in which prototypes are deployed and evaluated in workplaces outside of the developer's environment. The design philosophy acknowledges that designs must be tailormade to accommodate organisational culture and specific work practices if they are to be successful - in other words 'one size does not fit all'. This approach to design differs dramatically from traditional models of development and deployment or what have been called the Field of Dreams Model of Development - 'if you build it they will come' and the Velcro Model of Deployment - 'toss in the technology and see where it sticks' (Moore, 1994).
* the research team is cross-disciplinary in nature and includes sociologists, psychologists and information scientists in addition to the user interface designers, computer scientists and engineers traditionally involved in the research and development of new technologies.
* the researchers live in an environment that necessitates the use of their technology - the team is split between Ottawa and Toronto, split within the cities and within buildings in the same site. Thus the research teams lives and works in the type of environment they are creating.
The Ontario Telepresence model of research and development can therefore be distinguished from other media space projects in that it is used and studied by a cross-disciplinary team of researchers and it is deployed to actual workplaces with a variety of users with a diverse set of needs. This provides the development team with the opportunity to systematically observe the use of the system and to gain early insights into precisely those areas that were considered problematic in the first wave of videoconferencing.
Summary - Part I and II
The first wave of videoconferencing technologies in the 1960s and 1970s did not penetrate the marketplace to any degree. In the 1990s we are experiencing the second wave of interest in video communication technologies. We have suggested that the reasons for failure in the 1960s and 1970s were a combination of social, economic, psychological and technological factors. Overly simplistic assumptions about the cost saving advantages of the technology and the focus on video as a means to replace face-to-face communication led to disappointment among early adopters. In the 1990s, the field has changed radically. Technical capabilities and capacities, in terms of hardware, software and communication networks far surpass the first wave systems. In addition, the economic realities of the 1990s - such as the extended recession and the globalization of the economy - are driving a trend to organisational redesign and re-engineering. While these changes are not fully understood, there is growing recognition that there is a strong relationship between successful deployment of technology and a knowledge of the context into which it is being deployed. It is not longer enough to work harder, we must also work smarter. Designers of videoconferencing systems today are advised to learn from the experiences of the first wave. Success in the 1990s is unlikely to reside in solutions that are purely technical - a more complete understanding of the process of human interaction, a respect for the cognitive and social skills of the users, an awareness of the dynamics of organisations, and the characteristics of the work force must all be accommodated if we are to ride the second wave of innovation to successful adoption in the 1990s.
Part III Impressions of an Early Adopter: A Comparison of Two Systems
In the fall of 1992 Provitel, Inc. established a working relationship with members of the Telepresence team. One of the outcomes of this collaboration was the installation in May 1993 of a Telepresence Media Space (TMS) system between the offices of two of the members of CSG, a client support group within the corporation. It is important to note that at this time Provitel, Inc. was encouraging the use of VISIT desktop videoconferencing within the corporation. VISIT systems were being deployed to a number of members of CSG including those who had the media space. This provided a unique opportunity to evaluate these two systems - one a commercial product, i.e.,VISIT, the other, a futuristic prototype system, i.e., TMS.
Description of the site and its environment, the participants and their tasks:
The site and its environment:
The participants were located in a downtown office building in a major city. The offices of the particpants were approximately twenty feet apart on an upper floor of the building. The offices were open concept in plan (with six foot baffles separating work spaces) and arranged along an external wall that faced directly onto the office building next door. Each office was equipped with a Macintosh II computer, a Meridian phone set and Octel voice messaging. One of the participants regularly used a cellular phone.
Participants and tasks:
At the time of the study one participant was a Senior Consultant in visual networking, and the other, on loan to CSG from NSG - a network services group - worked as a Consultant. The responsibility of CSG is to develop visual network planning solutions for clients. They need to be aware of the variety of communication products available and how these work on company networks in order to be able to showcase a variety of visual networking solutions. A telecommuting project in this organisation is a good example of the type of work CSG does. CSG put together a software and hardware environment and made certain that the system was workable before it was deployed into the homes of the teleworkers.
Description of the two videoconferencing systems:
Description of the system:
VISIT is a desktop multimedia system which integrates desktop videoconferencing, screen sharing, high speed data transfer, electronic voice mail access and voice call management on a desktop computer. When augmented by Aspects software, VISIT also allows application sharing for word processing, drawing and spreadsheet operations.
VISIT consisted of:
- a plug-in video board, which occupies one NU-bus slot and
provides video digitization, compression and decompression,
- a black & white, fixed focus, CCD camera with electronic auto
- application software
In this installation there were network connections to both voice and data lines - the voice line added control of phone and voice mail and provided access to dialing directories. The system was run on a Macintosh computer.
Fig. 6 VISIT (Northern Telecom)
Personal desktop videoconferencing provides desktop to desktop connectivity through a video window on the computer monitor or workstation. A 56 Kbps link allows participants to send and receive voice, video and data. Standard telephony features are integrated with VISIT.
The video window presents a real-time black and white image (now available in colour) of both the local and remote participants. Screen sharing allows participants to capture and share information and to make simultaneous annotations to the information displayed. Application sharing allows participants to work on the same file simultaneously and then to save a local copy of the updated file.
This was one of the earliest deployments of VISIT and sourcing, installation and training were not yet routine matters for the third party authorized distributor. In particular, a lack of availability of components of the system hampered installation and affected timetables.
2. Telepresence Media Space System (TMS):
Description of the system:
TMS is best conceptualised in terms of media ecology - that is a system that strives to seamlessly integrate a variety of media technologies in ways that are appropriate to the users and the organisation. More than simple videoconferencing (whether desktop or conference room) TMS is designed to respect the existing physical, cognitive and social skills of the users, supporting subtle social protocols, confidentiality, intimacy and trust relationships that develop among people in everyday face-to-face situations. In addition, TMS is designed to support cooperative work practices in a variety of ways, e.g., real time document sharing and editing and other forms of CSCW applications as well as innovative video services such as video mail, video receptionist, and video recording of meetings (including videoconferences). TMS has the potential to transform the workplace in ways that are profound.
Telepresence Media Space (TMS) consisted of:
- a Sun workstation working as the central server,
- a computer controlled audio-video switch,
- a PictureTel codec,
- a Sony VCR, and
- a camera mounted on the roof with a view of a busy city intersection.
Each office cubicle was equipped with:
- a colour camera, external speaker and pzm microphone to provide desktop audio-video connections,
- low resolution, black & white 'far' camera equipment (i.e., a small monitor/camera combination),
- a 14" Macintosh color monitor connected to a Rastor Ops Media-time board to display video signals on the desktop monitor,
- MediaGrabber software to capture analog video inputs and display them on the Macintosh monitor,
- a hypercard-based user interface which permitted connections to be made to the codec, VCR, roof camera and the other office with a simple point-and-click operation.
Audio-video signals were transmitted over dedicated customer provided cable while control signals were transmitted over ethernet. The commands are received by the server via the local office ethernet. The server interprets the commands and in turn issues commands to the audio/video switch to complete the requested connection.
Fig. 7 TMS: Telepresence Media Space
The configuration of TMS is illustrated below:
Fig. 8 CSG-TMS System Block Diagram
TMS is an environment rather than a product or appliance. The goal is to provide users with a space that provides a variety of technical capabilities without distorting well understood social protocols and physical and cognitive behaviours, or demanding radical changes in behaviour.
One of the distinguishing feature of TMS is the ability of users to control their accessibility. Software 'doors' on the user interface and in Portholes allow the user to signal their availability. Figure 9 below illustrates how the door states are selected and displayed on the interface and Figure 10 maps the door state to the actions permitted. The Portholes figure included earlier (see Figure 5, p.13) also indicates the door state.
Figure 9 Doors on the TMS Interface
Figure 10 Software Doors
Open * others may contact your office without having to ask for permission. * others may glance into your office. Ajar * others may contact you but you must explicitly accept the call. * others may glance into your office Closed * others may contact you but you must explicitly accept the call * others may not glance at you at any time Locked * no-one may contact you or glance at you while your door state is locked
The use of multiple cameras and monitors offers another form of access control. For example, callers can glance into each other's offices from the 'doorway' and avoid the potential problem of interrupting ongoing work - a problem that can occur when a colleague appears full size on a desktop monitor. In this way callers can 'approach' in a way that parallels traditional office practice - i.e., they appear at the doorway and then can be 'invited' in - see Fig. 11 below. TMS also supports the virtual office.
Figure 11 Use of Multiple Cameras/Monitors
The provision of a 'window' - live video feed, in this installation from a rooftop camera, provides a view of the city when the system is not in use.
In the configuration used in CSG, the second monitor provides additional 'desktop' workspace for regular Macintosh operations.
A number of difficulties were encountered in establishing the TMS at the Provitel, Inc. location. While the staff from the Ontario Telepresence Project were available to assist in the installation and provide support, the amount of effort required was more than anticipated on both sides - a not unexpected result given the difficulties encountered with deploying VISIT, a commercial product with an authorized distributor.
Use of the Two Systems
The use of both VISIT and the TMS was hindered by technical and social constraints.
During the period in which VISIT and TMS were being deployed, one of the participant's work required him to be out of the office almost constantly. While the participants were in regular contact the majority of this was by telephone and voice mail. In fact, they estimate that they were only in the office at the same time perhaps one day every two weeks. This was a major factor affecting use of both systems.
From the outset of this experimental deployment it was understood that use would be constrained by the small user communities - a limited number of VISIT systems had been deployed in Provitel, Inc. at the time, and in the case of the TMS only 2 nodes were set up. It is important to note that it was not expected that much practical use would be made of the TMS system until links to other members of the media space community could be made, but this was to provide the CSG staff associated with the Ontario Telepresence Project first hand experience of a media space. TMS and VISIT were both used by CSG for customer demonstrations and for experiments in virtual collaboration. Using VISIT for collaborative drawing worked especially well as it allowed the two participants to work on the same 8.5" x 11" drawing simultaneously.
Prior to installing Telepresence and VISIT, the internal electronic mail system was accessed by connecting the Mac serial port to the MCA data port on the back of the Meridian phone set. When VISIT was originally installed it was discovered that it was incompatible with the existing electronic mail system as both systems were trying to access the network through the same data port on the Meridian phone set. As access to electronic mail was considered essential VISIT was deactivated unless a demo was scheduled. This problem was eliminated by gaining electronic mail access through the ethernet.
When TMS was installed it was discovered that the Meridian Teladaptor software was not compatible with the MediaGrabber software used to display TMS video signals on the second Macintosh monitor. This problem was not solved. As TMS was considered the more novel of the two systems, VISIT was de-activated unless a demo was scheduled or a VISIT call necessary. Spontaneity was lost as VISIT calls could not be received unless arrangements were made in advance to ensure that system was active.
In spite of the social and technical constraints involved in the deployment of both VISIT and the TMS the participants felt they had gained a number of early insights into factors that shape the use of these technologies, and more specifically into factors that should play a role in the future development of these systems. These include a number of observations around the role of video, and on the interrelationship of these technologies to the overall communication infrastructure of the organisational setting in which they are being deployed.
1. Technology fails as often for social as for technical reasons
The limited use of both these installations validated a premise of the Telepresence research group that the deployment of new technologies must be planned and that part of the planning process involves an assessment of the organisational culture and the work practices of the potential users. The traditional Velcro Model of Deployment - deploy it and see where it sticks is costly both in terms of capital and in terms of getting employees to 'invest' in adoption. In this study, an analysis of organisational culture would have revealed the CSG to be a highly voice driven as opposed to text driven culture - a culture better supported by telephony rather than an office based system. For example, the participants reported that they received and responded to dozens of voice mail messages a day, but that email, while used, was not reliable in terms of fast response time. Use of both VISIT and TMS required users to be in their offices, and messages could not be accessed remotely. However, the work practices of the participants required that they spend considerable time out of the office, visiting client sites, etc. The mobile rather than stationary nature of their work meant that an office based system was highly restrictive. In addition, both participants reported that a written record was not highly valued. This contrasts sharply with the culture and work practices within the NSG where one of the participants normally works. In this group the norm is for employees to be in their offices and the type of work carried out requires a permanent record of transactions as these play a role in later activities. VISIT or TMS would be more appropriate in this environment. The failure to take organisational factors into account in a large scale deployment would result in costly errors in terms of staff morale and capital. One of the participants indicated that he now has used this insight in his work with external clients.
2. Technology is an enabler, not a driver
Technology is most effectively deployed to support work practices - not to drive them. Individuals and work groups do not communicate because they have a specific technology, but rather they use technology because they need to communicate. An analysis of work practices which permits us to discover the nature of the communication network and communication patterns, both formal and informal, is thus an important part of the process of deploying appropriate tools to the appropriate personnel to support and enhance their communication needs. As already noted above, members of CSG were away from their offices on a regular basis - at client sites, in other parts of the organisation, at meetings, etc. They relied heavily on voice communication technologies and reported low use of electronic mail. Attention to these aspects of the group culture from the outset would have suggested that this was not an ideal work group in which to test office-based videoconferencing services. In NSG, on the other hand, where the staff are primarily office based and there is need for a written record, the use of email is high. It is expected that adoption and use of videoconferencing would also be higher in this group. A number of members of this group are now teleworking. As the participant in this study - who has now returned to NSG and is teleworking confirms it is the 'goodness of fit' between his work practices and the technology that make this an appropriate alternative for him.
3. Rationale for deployment should be linked to user's work practices
The maximum benefit of new technologies often comes from the discovery of innovative uses that emerge with experience. It may seem obvious to suggest that these innovative uses are most likely to be discovered when the use of the technology is high. In the January 1993 deployment of VISIT systems in Provitel, Inc. the rationale for the delivery of VISIT to personnel in contact with customers was clear - sales groups would benefit by gaining experience with multimedia technology and could both demonstrate the system to their clients and uncover networked multimedia applications. However, members of this group seldom needed to contact each other, and as the VISIT community was extremely small at the time there was little opportunity to use the system except for demonstrations. Without day to day familiarity with the system, the opportunity for discovery was minimized.
Comparison of the Two Systems:
1. The problem of definition and the role of demonstrations
The participants found that TMS was hard to describe to others, the power and the potential of the system being intangible and the fact that it is an environment means that it is best grasped by demonstration and first-hand experience. VISIT, on the other hand, could be easily described - people could see it and understand it as a communication device. It was also easier to give demonstrations of VISIT as it was a commercial product at this time. However, it was the live video window on the city street that appeared on the TMS monitor when it was not in use for videoconferencing that elicited most questions from colleagues passing by the open office. As one participant reported: "People would come by and say - how did you do that? (i.e., looking at traffic or into the parking lot). This was the thing that people noticed - no one ever commented on a multipoint VISIT session, but the live video feed drew attention. It blew them away and they'd want a full explanation. "
2. Role of video
TMS and VISIT have taken distinctly different approaches to the use of video and having the two systems side by side highlighted their differences. These will be discussed under image quality and size, image location, and awareness/communication.
Image Quality and Size:
TMS currently uses analog video and the image occupies the full screen of the monitor on which it is displayed. VISIT, on the other hand, is a digital image which appears in a small window on the computer or workstation monitor. The quality of the analog signal was highly superior to the digital signal. One of the participants indicated that this experience taught him that there is a definite threshold for video quality and size if the video is actually to be used- He commented: "VISIT was below it and TMS and PictureTel are above it. In these two systems you really use the video - VISIT users never look at the video. In fact, they are always moving around and are often off camera - but nobody ever complains. "
"The use of dual or multiple monitors is one of the good ideas to come out of Telepresence" reported one of the participants. The commercial desktop units, including VISIT, incorporate the video image on the computer screen. In the TMS the number of monitors is optional and it is possible to move images from one place (monitor) to another as the nature of the activity changes - e.g., you stop by the door (a small image on a distant monitor) and then move to the visitor's chair (a larger monitor appropriately placed) or to the desktop (a monitor beside the workstation) for collaborative activity. With the image on the workstation, as with VISIT, you must choose between the video image and the document. He continued: "The video has to go to the back when you are working on a document. People rationalise that the video is not important, but how do they know as they are not using it. The choice is built into the design."
One of the themes of the Ontario Telepresence Project has been to distinguish foreground communication from background awareness, to recognize the unique importance of each and to design a system that provides a seamless transition between them. Background awareness can be compared to the type of awareness one has of colleagues in an open office, while foreground awareness results from an intentional act such as placing a telephone call. The use of both VISIT and the TMS revealed that VISIT video is very useful for background awareness, but of limited use for foreground communication. To reconceptualise and reposition VISIT video as a background awareness server turns what has been perceived by some users as a design flaw into a feature. For example, in a multipoint call VISIT provides a small portrait of each of the participants in the meeting. While these images appear to be too small for effective foreground communication they do provide a way for members of the meeting to 'keep track' of who is still in the meeting, who has left, etc. This is one of the major problems experienced in audioconferences, especially if a participant remains silent for a period of time. This use of video is inexpensive within a local calling area, but given the cost of long distance this may not be an acceptable market position without innovative pricing structures.
3. Integration of desktop and conference room videoconferencing
VISIT is exclusively a desktop system. One of the advantages of TMS is that a desk or office can be linked easily and seamlessly to any conference room or directly to an external codec. This provides some advantages which were considered by the participants to be of high value:
* maximizing the use of expensive resources, e.g., a high-end
codec which still costs in the range of $40,000 -$50,000.
In TMS, the capital cost of the codec is distributed over all the participants in the media space - each member of the group has access to a codec which delivers high quality video. Desktop systems, such as VISIT, take a dedicated approach to codec resources - each unit has its own codec but the codec is of lower quality (a cost factor) and subsequently lower quality video.
* the ability to attend meetings remotely, both locally and those
at a distance
With TMS it is possible to attend a meeting from your office or a local conference room. Also there is flexibility in terms of the role of the office attendee. At times this may be a substitute for travel to a meeting, or make possible attendance at a meeting that tight scheduling would otherwise prevent. Alternately, the meeting attendees may need access to documents, files or reference materials in the office that cannot readily be transported to the meeting. The flexibility to attend a relevant portion of a meeting, or for the attendees to be able to call a co-worker into a meeting for a brief period is a further attraction of TMS. Activities such as those above refer primarily to foreground communication, but the linkage between the office or desk and conference room also can be used to provide background awareness. It is possible to monitor a meeting in progress from the office - the attendees 'see' the remote participant who can continue working on other tasks, but is able to intervene when appropriate and relevant. By keeping the audio off in the remote office - except for interventions - the meeting is not disrupted. Requests for intervention can be handled in the same way as for those physically present, e.g., the raised hand of the remote participant is seen and can be handled as it is for any other attendee. This feature was not available during this short study, but the participants consider this to be an important feature, and one that would have facilitated and enhanced their ability to be present with colleagues at the Ontario Telepresence Project.
4. Integrating telephony
The VISIT desktop system is integrated with the telephone, and the participants liked this feature. As one reported: "I don't use it all the time, but when I do it's really neat and I realise I should use it more often." All calls to a VISIT user are recorded on the VISIT call log. When the user returns he has a visual record that can be checked instantly - i.e., there is far less overhead for the user than checking a large number of voice mail messages - and by double clicking on the caller's name he can call back directly. It is also possible for the caller to leave a written message, very effective for notifications that do not require a call back - e.g., a change in meeting time. Another feature appreciated and used by these participants was the ability to leave yourself a memo in the VISIT directory when you have left voice mail. When the call back is received, the memo appears thereby serving as a reminder of the reason for the initial call. However, there is a drawback to the integration in that the computer or workstation is 'frozen' for a 'couple of minutes' while the video is being set up. This can be extremely annoying as this following comment from one of the participants captures: "I'll be furiously working away on a document that I have to get done, I have 5 minutes left to hit the send key and someone calls me and the computer is tied up!" This is a problem that will become increasing noticeable and annoying as the number of VISIT users increase. There is an off switch to put VISIT over to the telephone, but this involves overhead for the user. As one participant reported: "it's not intuitive, nor is it easy to use and you forget to turn it back on".
5. Virtual office or the office share
The primary use of the TMS by the participants was as a virtual office. A continuous link between the two participants' offices was maintained and proved very useful. Although working on separate tasks, questions that arose could be discussed without interrupting the task in progress at the other end. One of the participants has also used VISIT in this way. Working on a Sunday with a colleague in Ottawa to prepare a presentation they left up a VISIT link during the time they were working. The primary use was for background awareness. He commented: "I felt that I wasn't the only one left on the face of the earth, especially working on Sunday. With VISIT it's not exactly working together, but at least you're working. Separate monitors and a high quality image as there is with TMS, would have really enhanced the feeling of working together." At present, however, the virtual office is an expensive option if long distance charges apply.
1. The need for an ecological approach
We are dealing with a communication ecology that involves understanding the relationship between organisations, work practices and technologies. To conceptualise the environment in this way brings new insights to all aspects of innovation from design, and development to deployment and use.
Some specific examples of how getting away from an appliance model of design helps move us forward.
* bandwidth on demand
In an intelligent communication environment we want to be able to 'move' about comfortably. If we focus on what people do and how they do it rather than giving them another appliance that will do some of what they do in yet another way, we can support work practices and grow use. For example, the user should not have to decide in advance the amount of bandwidth they need, but it should adjust automatically with their activities. 'Living ' in a media space made one of the participants in the study sensitive to the fact that nobody is asking about the user and how they might want to vary bandwidth. For example, the user might place a phone call, and once connected move to a full scale video call. Shortly after they might drop to background awareness mode using lower video quality or audio only. The network needs to be able to scale dynamically. It also needs to be made affordable, and the relationship between bandwidth and cost is unlikely to be a linear one. This presents new challenges to the bandwith vendors.
* services and technologies must be integrated
The current communication environment places considerable overhead on the user. Rather than trying to stay under the threshold of tolerance we need to understand the ways in which overhead is experienced by a variety of users and work at designing this out of the environment.
2. Bringing the user and the organisation back in
Innovation needs to be planned and managed to maximize effectiveness and efficiency. Technology fails as often for social as for technical reasons. Time spent at the outset in understanding the organisation and the work practices of the employees will guide the development of a deployment plan . Planned deployment both helps the organisation identify the 'sweet spots' for the use of specific technologies and identifies the work groups that have the potential to be successful early adopters thereby creating a positive impression about the technology within the organisation.
3. Productivity issues
The traditional productivity equation has been unable to demonstrate a positive relationship between use of various information technologies and increased productivity. Our experience with both VISIT and TMS has provided many examples of ways in which productivity has been increased, but these observations are not captured by the traditional productivity metric. Until such time as the metric is improved, it is the responsibility of those involved in the diffusion of these innovations to record and report on demonstrated productivity increases. This is the new productivity.22
This paper has presented an overview of the ways in which videoconferencing is being carried out today, and a comparison of two different approaches to the use of video in the workplace - a commercial desktop product, VISIT and an experimental media space technology -Telepresence Media Space (TMS). Our focus has not been technical, but on gaining early insights into the impact of and opportunities created by this new class of technology. Our results demonstrate the need for an ecological approach to understanding videoconferencing technologies. That is, successful deployment and adoption and continued use and growth of the technology require a 'goodness of fit' between the organisation's culture(s) and the work practices of employees. The integration of telephony with the videoconferencing system is a strong feature of the VISIT desktop system, but video quality, location and size are a serious problem. However, if VISIT is reconceptualised as a system to support background awareness rather than for supporting dynamic interaction it may be more highly evaluated. The Telepresence Media Space forces us to pay attention to the distinction between background awareness and foreground communication. The primary point to be made is that bandwidth on demand - the ability to scale video images dynamically - coupled with a price structure that is affordable - and unlikely to be a linear relationship between bandwidth and cost - presents both a major challenge and a major market opportunity for the network provider positioned to understand the potential of video.
Special thanks to the members of CSG, Provitel, Inc. for the time they have spent participating in this study.
Bikson, Tora K. 1987. 'Understanding the implementation of office technology'. Pp. 155-176 in Robert E. Kraut, ed., Technology and the Transformation of White-Collar Work. Hillsdale, N.J.: Lawrence Erlbaum.
Brown, John Seely. Informart Magazine 2: 1994, 17. [Interview]
Egido, Carmen. 1988. 'Videoconferencing as a technology to support group work: a review of its failure.' Pp. 13-24 in CSCW 88: Proceedings of the Conference on Computer-Supported Cooperative Work , Portland, Oregon,Sept. 26-29, 1988. New York: ACM.
Fish, Robert S. et al. 1993. 'Video as a technology for informal communication.' Communications of the ACM 36 (1 ): 48-61.
Moore, Gale. 1994. 'Field Studies - Social Sciences': Presentation to the Bell-Ontario Telepresence Project Meeting, Jan. 19, 1994. Toronto: Ontario Telepresence Project.
Moore, Gale. 1990. The Information Industry in Ontario: A Test of the Post-Industrial Thesis. Toronto: University of Toronto. [Phd thesis]
Tang, John C. and Ellen A. Isaacs. 1992. Why Do Users Like Video? Studies of Multimedia-Supported Collaboration. MountainView, CA: Sun Microsystems Laboratories, Inc. [Technical report, Dec. 1992]
Note: A video of Ken Schuyler and Gale Moore carrying out this collaborative work is available from the Ontario Telepresence Project.