Transcending Time and Space

The sole superpower standing at the end of the Cold War found itself with interests at stake in more areas of an increasingly volatile and complex world, harassed by smaller but dangerously agile potential adversaries.

With fewer budgetary, personnel, and organizational resources at its disposal, how would this superpower’s government, and particularly its intelligence community, secure its citizens?

Specifically, how would it shorten its timelines for decision making, satisfy a heightened demand for sharing assets, and facilitate coordination through team building and the flexible use of skilled but dispersed personnel?

Collaborative Virtual Workspace (CVW) is part of The MITRE Corporation’s response to these post-Cold War contingencies. It is both concept and prototype whose many applications include organizational interactions, operations management, distance learning, and crisis management. It introduces a new way of thinking about collaboration together with a technology that allows an organization to overcome many of the limitations ordinarily imposed by time and space. CVW also reflects MITRE’s freedom to experiment with leading-edge technology in the context of systems engineering. David Lehman, MITRE’s chief technology officer, notes that this freedom allows MITRE to "stretch the government, be out ahead of them and change the concept of operations."

Development

In the mid-1990s, software tools that enabled virtual meetings began appearing in the marketplace. Around the same time, virtual organizations–teams of people from disparate locations and organizations working together–began to pop up, together with limited technology to support them. Text-based, virtual-reality adventure games on the Web became popular as well–games populated with people, places, and things, open and free to anyone who wanted to play, develop, and extend them.

What if these existing technologies could be integrated in support of virtual teams that would enjoy a persistent, place-based environment for collaboration?

With seed money from an Air Force research project in collaborative computing and from corporate research and development funds, a team of MITRE employees created a proof-of-concept prototype. That prototype, in turn, justified a MITRE-sponsored research project and led to a breakthrough in virtual collaboration that is changing in basic ways how organizations organize themselves to do business.

From the beginning, CVW exemplified collaboration, leveraging incremental technologies in the development of more comprehensive technologies. Xerox PARC’s LambdaMOO, the base upon which the virtual-reality games were built, was available on the Internet for release under Open Source license. Team members used the extensible, object-oriented scripting language that MOO technology provided to develop the software that generated their prototype’s collaborative environment.

In time, the virtual building and client-user interface that are characteristic of CVW’s object-oriented environment took shape. The team added audio and video conferencing capabilities, using software developed by Lawrence Berkeley Laboratories that was also available on the Internet for release under Open Source license. Finally, they developed an HTTP-based document server for importing and exporting all kinds of files, from simple texts to satellite images.

Functionality

From a technical point of view, Collaborative Virtual Workspace is a framework for seamlessly integrating diverse tools that enable collaborative capabilities.

From a user’s point of view, CVW is an environment for collaboration modeled on reality. It is a virtual building divided into floors and rooms. Each room incorporates the people, information, and tools appropriate to a task, operation, or service and provides a context for communication and document sharing.

People gather in rooms to talk via text chat and audio/video conferencing, to share information from virtual file cabinets and URLs, and to mark up shared whiteboards. Users can move from room to room, just as they would in a building. They can meet team members to converse, collaborate, and interact, prepare documents, generate ideas, solve problems, and perform other functions as if they were actually collocated and present at the same time, rather than geographically and temporally dispersed.

"The system actually allows for chance encounters when moving from room to room," observes Esther Rhode, business lead in collaborative solutions. "Just as in a real building, you can bump into someone in the hallway and get to talking; as with a face-to-face meeting, this sometimes leads nowhere, but other times provides the key ingredient in solving a problem."

With early products, intended to solve discrete collaborative problems, all the work, all the resources, and all the links and addresses simply disappeared once a collaborative session was over.

But with CVW, because the rooms persist even if no one is in them, and because the documents remain there until some authorized user moves or deletes them, collaboration can occur continuously from one session to another. The user, who has access from anywhere in the network, does not need to set up a session or know the location of a collaborator; the user needs only to enter a room.

Put another way, CVW is a robust and stable prototype that supports integrated synchronous and asynchronous interaction, with the potential for saving time by transforming serial processes into parallel ones and for increasing the quality of distributed training, analysis, and decision making by integrating geographically distributed groups into cohesive teams.

CVW and Collaborative Culture

With its potential for saving time and increasing the quality of distributed training, analysis, and decision making, the Collaborative Virtual Workspace (CVW) can significantly advance the ways in which MITRE’s customers and other organizations conduct business.

But only by developing CVW in concert with MITRE’s customers could the CVW team understand their requirements for collaboration and refine the use of spatial metaphors. Only with input from MITRE’s customers could the team learn how to insert collaborative technologies into an organization in a way that could itself help ensure the organization’s commitment to exploring and enhancing collaboration’s potential. So the team introduced CVW to many different government agencies and refined the prototype through an iterative process of pilot and assessment until it could facilitate collaboration scaled to hundreds of users.

In February 1997, the challenge of integrating CVW into a major intelligence organization’s one-week Intelligence Crisis Exercise (ICE) provided the MITRE team with its most productive opportunity for testing and developing their model. Hoping to use CVW to define a vision of the future of collaboration, and wanting their personnel to be so proficient in the technology that they truly could conduct the exercise virtually, the customer began preparing to use CVW about half a year before the actual start of the exercise. The team helped define the concept of operations for how to use the collaborative environment, assisted with systems and network engineering, and trained users and systems administrators.

Collaboration, it became clear, was not technology driven but need driven— shaped by people, process, and culture. The technology itself, by preserving the human element critical to collaboration, supports and encourages the development of a collaborative culture. As human factors engineer Tari Fanderclai points out, "Since the objects in the CVW environment are persistent—until deleted by their owners, objects remain in the environment even when the original creators are not connected—users can furnish their virtual workrooms in the same way we furnish physical workspaces: with the tools and objects that the group needs, with items we want non-members to notice if they enter our spaces, and sometimes with items that simply please or amuse the members of the group."

CVW mimics the semantics of the physical world as closely as possible, giving users the illusion of physical space that permits natural social behavior.

The ease with which team members can share resources, the socially-oriented locus of control, and the ability to close geographic, temporal, and sometimes even hierarchical distances between users all contribute, like CVW’s persistent environment and place metaphor, to the development of virtual community and collaborative culture among CVW users.

But attention to the underlying processes of that intelligence organization and to the interactions of its members also revealed significant challenges to collaboration. For example:

Intelligence personnel are usually rewarded for what they know, but if they have to share, how do they get credit? How does an organization reward the team behavior it wants to encourage?

Security normally wants to close everything down; collaboration wants to open things up. Now that we have the technology to facilitate collaboration, how do we create appropriate security policies that can enable collaboration while protecting information?

When multiple internal elements or multiple agencies coordinate a project, who keeps the log and who "owns" the final report? Who, ultimately, is accountable?

As an organization learns to work virtually, how will members of the organization know when they've completed an action? How will managers manage them?

By acknowledging that issues such as these must be addressed as part of the organization’s comprehensive requirements, MITRE was able to configure the technology and identify the organizational and cultural processes by which it would be implemented.

The results were beyond expectations. What CVW was able to accomplish for the ICE–live briefings, a readily available broad range of expertise, real-time observation by decision makers of how intelligence products were being used, less time coordinating product reports, coordination without relocation or disruption–transformed that organization's ability to manage information flow during crisis operations.

In fact, so successful was CVW in "scratching an itch" that the organization decided to continue to employ CVW after the crisis exercise.

Today, commercial virtual workspace products are available or being developed. As we await the appearance of a product on the marketplace that fully meets our needs, the intelligence organization that experimented with CVW in its Intelligence Crisis Exercise continues to experiment with CVW to help it understand issues relevant to a number of important activities. These include crossing boundaries to collaborate with other organizations within their community of interest, readying their infrastructure–network, desktop, and security–for large-scale collaboration, and adapting collaborative technology to improve business operations.

The customer also continues to sort out which aspects of the old organizational culture can be remapped onto a virtual world in order to enhance collaboration and which cannot and should not.

CVW and the Academic Community

The Collaborative Virtual Workspace (CVW) Outreach server is a special project created by MITRE's CVW team to provide space for academic collaborations. The site provides another opportunity for the team to learn about how people use collaborative tools. In exchange for the use of CVW Outreach, users are asked to provide the team with feedback regarding their experience collaborating with CVW.

Eligible users of the server might include university classes whose students and faculty could benefit from collaborating on class work, groups of researchers who need an online place in which to collaborate, or cross-institutional academic organizations or projects. To date, the most enthusiastic response to CVW Outreach has come from students in writing cIasses, who use virtual collocation not merely to edit each other's work but to participate jointly in the creative process.

Students at the University of Puerto Rico

A second application for CVW in the academic community has arisen from MITRE's work as a member of the Industrial Affiliates Program (IAP). For a number of years, MITRE has been actively engaged with the University of Puerto Rico (UPR) in sponsoring research projects that allow participating undergraduate students in engineering to develop hands-on industrial experience above and beyond their regular curriculum. Not long ago, the university requested MITRE's support in developing a new network engineering curriculum. At the same time, a similar request came from North Carolina Agricultural and Technical State University (NCA&T), with whom MITRE is also actively engaged.

The confluence of these two requests created an opportunity for collaboration among MITRE and the two universities. CVW accounts are being established for both universities so that their faculty, primarily in the disciplines of electrical engineering and computer science, will be able to cooperate in developing parallel curricula in network engineering and distributed systems. The universities will have access to MITRE's expertise in curriculum development as well as to MITRE personnel who can lecture on campus.

At the same time, CVW will serve as a platform to enable project tasking of selected students by MITRE's sponsors. Students will perform their work virtually at MITRE under the guidance of UPR and NCA&T professors while located physically on campus. A strong and productive relationship has already been established between students at UPR and NCA&T and MITRE's site at Eatontown, New Jersey, promoting collaboration on tactical support of the Army in the area of command, control, and communications for battlefield systems.

Students supporting the different research programs can support each other; the universities can enrich their students with practical experience that satisfies the expectations of the Accreditation Board for Engineering and Technology; MITRE's sponsors can get effective talent applied to their projects; and MITRE can get the exposure and insight it values in recruiting qualified engineering candidates.

In the words of David Gentes of MITRE's Human Resources Department, "Effectively what we're doing is working a co-op program where nobody goes anywhere." And what difference, ultimately, will the program make in the future of collaborative research? As Angel Colon, MITRE's IAP liaison to UPR, puts it, "The moment the technology gets out there, folks start thinking all kinds of things."

A Future for Collaborative Research

To encourage the development of commercial products that can make Collaborative Virtual Workspace (CVW) available, affordable, and supportable, MITRE has converted the current CVW effort to an Open Source project modeled after the Linux, GNU EMACS, and Apache Open Source projects. With Open Source, the source code is posted on the Internet, allowing MITRE to make CVW available on a nonexclusive basis to anyone interested through a general public or other license. Moreover, Open Source allows code "check-ins" by anyone who wishes to contribute enhancements, with MITRE serving as the Open Source moderator–integrating code contributions, moderating conflicts, and republishing the source code as needed. (See the Open Source Collaborative Virtual Workspace Web Site for further information about this effort.)

Opening the CVW source code to the public, according to Thomas Gannon, Intellectual Property Management Director in MITRE’s Technology Transfer Office, also offers a way for MITRE to use the Web as a collaborative research and development tool. "Open Source is a mechanism to assist us in furthering our own research and understanding how collaborative virtual workspaces can be applied to different public domains." Using Open Source to make CVW available is itself an act of collaboration, supporting a new channel for the free exchange of ideas and promoting cooperation across many disparate entities.

Such collaboration will, MITRE hopes, fulfill the vision that Eric Raymond articulated in The Cathedral and the Bazaar: "I think that the cutting edge of open-source software will belong to people who start from individual vision and brilliance, then amplify it through the effective construction of voluntary communities of interest."

Will a voluntary community of interest actually evolve around CVW? In investigating the reasons for the success of Linux, Raymond focused on peer review. "When programmers on the Internet can read, redistribute, and modify the source for a piece of software, it evolves. People improve it, people adapt it, people fix bugs. And this can happen at a speed that, if one is used to the slow pace of conventional software development, seems astonishing . . . and produces better software." But unlike Linux or Apache, CVW is a unique code base with which, as a combined entity, very few people are familiar. Moreover, it demands social reengineering for its users to experience enhanced productivity. Even the intelligence organization that has adopted CVW for everyday work has never looked at the code; it is more focused on the process of how to work collaboratively, more concerned that the code works than with how it works.

As our customers gain more experience with CVW, the initial feedback from many of them is that the system can yield significant benefits for providing virtual collocation. It also stimulates their thinking about the potentially far-reaching effects of this kind of collaboration on the future workplace. At MITRE, the current focus of collaborative research is the integration of collaboration services into a Web-based infrastructure. MITRE is also researching multicast security, semantic collaboration, virtual environment visualization, and virtual community federations. In other words, MITRE, too, is thinking ahead to that future workplace.