5
Research Recommendations

Distributed work, including telecommuting, relies heavily on computing and communications technologies. Recent improvements in the capabilities and availability of tools in these interrelated fields have meant that distributed work can be undertaken more easily now than in the past. Likewise, future improvements will benefit those engaging in distributed work and provide both economic and social benefits to the nation. Thus the committee supports the continuation of broad national computing and communications research programs in academia, industry, and government.

However, in order to optimize the opportunities for and effectiveness of distributed work, the committee recommends that research also be conducted with distributed work as a specific focus. The research needed can be conceptualized as indicated in Figure 5.1: infrastructure research is most needed at the extremes of the range of computing and communications capabilities, and applications research is most needed in the middle of the spectrum.

INFRASTRUCTURE RESEARCH

The national initiative to develop the technologies for and establish the National Information Infrastructure (NII) will benefit distributed work. For example, the NII will require both richer and more robust security features than are now available on the Internet. These



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 56
Research Recommendations: To Facilitate Distributed Work 5 Research Recommendations Distributed work, including telecommuting, relies heavily on computing and communications technologies. Recent improvements in the capabilities and availability of tools in these interrelated fields have meant that distributed work can be undertaken more easily now than in the past. Likewise, future improvements will benefit those engaging in distributed work and provide both economic and social benefits to the nation. Thus the committee supports the continuation of broad national computing and communications research programs in academia, industry, and government. However, in order to optimize the opportunities for and effectiveness of distributed work, the committee recommends that research also be conducted with distributed work as a specific focus. The research needed can be conceptualized as indicated in Figure 5.1: infrastructure research is most needed at the extremes of the range of computing and communications capabilities, and applications research is most needed in the middle of the spectrum. INFRASTRUCTURE RESEARCH The national initiative to develop the technologies for and establish the National Information Infrastructure (NII) will benefit distributed work. For example, the NII will require both richer and more robust security features than are now available on the Internet. These

OCR for page 56
Research Recommendations: To Facilitate Distributed Work FIGURE 5.1 Types of research needed to support distributed work across a range of capabilities. same features will of course be useful to distributed workers to help ensure proper access, control, authentication, and tracking of both individual and group work. The committee recommends that specific technical infrastructure research to enhance and extend distributed work practices be undertaken in the following areas. Mobile Work Working while mobile, for example, while traveling in a automobile or standing in a customer's office, presents many challenges due to the limited communications bandwidth and computing power that are typically available compared to those offered in a stationary environment with wire or optical fiber network connections. While pocketsize cellular telephones have enabled voice communications for many users, reliable mobile data communications services are just now being deployed and are still typically limited to asynchronous, low-bandwidth applications like electronic mail or its equivalent. Similarly, computing power becomes limited as equipment size and weight are restricted. Laptop computers, weighing a few pounds and suitable for use while sitting down, are capable of running most programs from a desktop computer environment, but computing power

OCR for page 56
Research Recommendations: To Facilitate Distributed Work and thus the programs that can be used become extremely constrained on a ''palm-top" device suitable for use while standing or walking. To help increase the usefulness of mobile devices and services, the committee recommends a comprehensive research program with the following major elements (goals): Increasing the bandwidth available to mobile devices. This might include using multiple channels or transmission technologies to increase the effective delivered bandwidth. Developing network protocols that specifically facilitate mobile addresses or end-points. (The current protocols, TCP/IP and ISO TP4, were developed under the assumption of geographically fixed end-points.) Effectively and efficiently allocating among multiple media (e.g., image, text, audio, and signaling) the varying bandwidth that is available to a mobile user. Both short-term variances caused by circumstances such as changes in network load and long-term variances attributable to advances in transmission technology will need to be gracefully accommodated with minimal user attention. Balancing and accommodating the competing service demands and network traffic generated by multiple, simultaneous mobile users. Maintaining the essential functionality of communications that originate in a high-bandwidth environment but must be delivered in environments of varying bandwidth, including those of low-bandwidth mobile devices. Managing the real-time tracking and transfer of users moving between wireless cells, and ultimately the transfer of users moving from a wired to a wireless network environment in the course of a single, multimedia work or conference session.1 Developing computer processes and programs that can migrate across a network with the user as that individual moves from place to place. (Ubiquitous computing and a mobile computing environment are similar concepts.) Facilitating periods of "disconnected work" done when a network connection is not available. Currently, even the best-connected 1   Wireless systems already track users in order to provide uninterrupted service. However, safeguards must be established to prevent the system from being abused as a way to simply monitor the movements of citizens for other reasons. This is one concrete example of the need for implementing privacy safeguards as the nation's technology advances.

OCR for page 56
Research Recommendations: To Facilitate Distributed Work mobile workers have some periods when they must work with no network connection at all, such as during commercial air flights. This can present a problem when their work involves files that are shared with other workers: when a temporarily disconnected worker returns to connected operations, how can he or she know which version of each file is the latest, and how can the changes made by the temporarily disconnected worker be reconciled with a file that may also have been changed by workers within the group who remained connected? Complicating but also adding promise in low-bandwidth-environment research is the fact that computing power and communications bandwidth can often be substituted, one for the other. For example, if a suitable processor and algorithm are available, data compression can reduce the need for large bandwidths by effectively decreasing the quantity of information that must be transmitted. Multipoint, Multimedia Distributed Work A broad-ranging technical research program should be initiated to explore possibilities for distributed work practices using multipoint, multimedia communications at different, but generally abundant, bandwidths. 2 In addition to facilitating distributed work, a research program in this area would help ensure the timely implementation of the NII. It should include research aimed at achieving the following: Understanding and specifying minimum quality-of-service standards for a representative variety of application areas. Currently, the impacts of conditions such as latency, jitter, buffering, and synchronization of high-quality, interactive audio, image, and video are not fully understood. For example, the amount of video jitter that is acceptable for distance learning may be figuratively and literally deadly in a telemedicine application. The requirements for latency, jitter, and the like may also differ between live and delayed (recorded) viewing. 2   Related sociological, organizational, and human-factors research will need to be conducted contemporaneously with technical research. For example, videoconferencing is still a relatively new technique: we use it, but not necessarily well. How can it be made more effective? What formats are good for meetings? How can both previously prepared and spontaneously written materials best be integrated into a videoconference?

OCR for page 56
Research Recommendations: To Facilitate Distributed Work Predicting and controlling the various types of simultaneous communications traffic to ensure the required quality of services for both point-to-point and multipoint sessions. This may involve individuals using devices with varying bandwidth capabilities joining and leaving during an active session. Determining how to charge for distributed patterns of resource consumption on complex networks that may include sections of varying bandwidth. For example, the current conference-call model of charging a single party with the entire cost of a distributed work session is very likely to inhibit usage when newer, more complex networks and connections are available and are used by multiple organizations that have agreed to be responsible for their own costs while cooperating on a particular project. Specifically, a party participating via a limited-bandwidth connection in an otherwise high-bandwidth, multimedia, six-way conference session may not wish to pay one-sixth of the total cost of the session. Ideally, a variety of methods for billing according to bandwidth use or allocation should be supported. Establishing multimedia communication sessions among multiple, remote parties, to include notification of individuals and initiation of the actual session with minimal human attention. APPLICATIONS RESEARCH Domain-Specific Research Although various types of distributed work are currently carried out in the U.S. workplace, the implementation and effectiveness of important organizational group processes are often constrained by the lack of relevant computing and communications applications in a distributed work environment. For example, the informal methods that people in organizations have for sharing information, coordinating tasks, learning new skills, recruiting members, and sustaining group activities cannot be applied easily in distributed settings. The committee believes that many of these processes can be facilitated by a sustained research effort focusing on computing and communications applications for distributed work. Thus, the committee recommends research-oriented field trials of distributed work in areas of national or commercial interest, such as delivery of health care in rural areas, collaborative design of complex systems, or the maintenance and repair of deployed equipment. These trials should be conducted across large geographic areas and should utilize heterogeneous hardware, software, and communications services whenever possible. As appropriate, they could be combined

OCR for page 56
Research Recommendations: To Facilitate Distributed Work with other federal and state field trials of distributed education, health care, distributed science, or electronic commerce. The field trials should lead to the development of applications tailored to improve performance in each of these domains. Importantly, it appears that with minor modifications they will enhance distributed work in other domains as well. The committee believes that similar efforts in other important areas of distributed work would help solve the difficult problems of information sharing, group socialization, remote supervision, and related constraints in the context of employer-employee relationships. Properly designed field trials of distributed work would provide the opportunity to conduct extensive research in important substantive areas such as education and on tools such as group writing products that could enhance and extend future distributed work in many other fields or domains. Research should be conducted to discover approaches to the following: Improving the capability to search for and share unstructured information. Research aimed at improving the tools needed to identify potential group members and to identify and access personal stores of information in the same manner as information stored in libraries and other institutional archives is only starting, and more is clearly needed. Even if the technological problem of finding experts were solved, the social issues of ensuring privacy and controlling access to expert individuals across networks need extensive research. Numerous opportunities for research also exist in the area of search algorithms capable of handling nonstandardized information formats on heterogeneous hardware platforms. Developing techniques for forming and sustaining distributed groups, including methods for developing group cohesiveness, enabling remote supervision, providing individual and group motivation and rewards, and ensuring continued commitment to a project or organization. Behavioral and organizational research should be conducted concurrently to better identify exactly how physical proximity stimulates effective recruitment of group members, so that functionally analogous processes can be encouraged in distributed work settings. Scaling and extending information retrieval techniques so that users can search all relevant and available databases, not just previously known and selected databases. Query techniques that are independent of the underlying database structure will be increasingly valuable as increasing amounts of information become accessible via public networks such as the Internet. Related research is needed into highly reliable global file systems and environments and on improving

OCR for page 56
Research Recommendations: To Facilitate Distributed Work the performance of distributed databases, and their impacts on network traffic. More generally, research is needed to develop techniques for browsing and other ways of facilitating queries in which the user may not have a well-specified question. Developing portable, extensible tool kits for building distributed remote control systems and their user interfaces. For example, key control functions must be identified and made accessible to remote control, security and safety must be foolproof, and both novice and expert users must be accommodated. Additionally, researchers should investigate the potentials for controlling heterogeneous devices with the same end functions through a single remote control program. The field trials recommended by the committee should be interdisciplinary research efforts, with expert personnel drawn from the behavioral and social sciences as well as from the computing and telecommunications disciplines. Social and behavioral scientists have extensive understanding of the functions that groups need to fulfill to successfully accomplish their work and experience in evaluating the success with which new computer and communications-based tools have met these needs, as well as expertise in assessing the social impact of new tools and technologies. Technology researchers have expertise in inventing successful applications that push the limits of technology, that can be generalized across domains, and that raise fundamental research issues in the supporting technical disciplines. Together they can explore the interrelated technical, human, organizational, and social factors that can enhance and extend distributed work across many discipline-and situation-specific domains. Generic Research Computers have provided powerful new capabilities to the nation's workers. Word processors and desktop publishing programs make writing, editing, laying out, designing, and publishing sophisticated documents dependent on the skill of the individual rather than on the availability of expensive specialized equipment. Spreadsheet and graphics programs allow legions of workers to analyze scientific, financial, and other numerical data in sophisticated ways that prior to 1980 were possible only for those with access to large mainframe computers. Database programs allow workers with only minimal programming skills to record, store, retrieve, and analyze huge amounts of data. However, with these ever-increasing capabilities has come increasing

OCR for page 56
Research Recommendations: To Facilitate Distributed Work complexity. Workers needing to cooperate with others must now be conversant about multiple file formats and their compatibility (or general lack thereof), translation utilities, diskette formats, data densities, and even features supported between versions or releases of the same program. While these intricacies are reasonably manageable within a single, physically contiguous organization with ample technical support, they can quickly become very limiting within a geographically distributed work group using heterogeneous hardware, software, and communications media. Mobile workers may have an even more difficult time as they attempt to communicate and cooperate across time zones utilizing unfamiliar telephone and data connections with little or no technical support. If the capacity for distributed work is to expand to individuals and groups beyond those with considerable technical proficiency, attention must be paid to simplifying many of the required operations and user interfaces. The committee thus recommends a research program with the following goals: Developing distributed work tools, interfaces, and systems that actively encourage information sharing. This activity is critical to group accomplishments and can become difficult in dispersed groups. New systems should include features to anticipate and accommodate work conditions described in the social science literature—such as social dilemmas, social loafing, and altruism—to improve information sharing. Bridging the gap between synchronous and asynchronous communications to facilitate both prior input and follow-up to group work sessions. Currently, distributed group discussions and work sessions tend to take place synchronously, using only a few specifically preannounced reference or discussion items. Extensive pre and post-session discussions occurring by asynchronous means such as electronic mail are often ignored or imperfectly summarized and considered due to the difficulties of indexing, referencing, conveniently using, and generally relating them to a work session occurring on a synchronous medium. Technical and human-factors research may reveal methods to better link these two major categories of communication. Improving and extending the user interface to the telephone network and services. The nearly universal 12-key touch-tone key-pad has become a limiting factor in many systems. So-called voice response systems use audio only on one end of the transaction, and users must be led through a long menu tree in order to limit the number of choices at any one branch. Interfaces might be both richer

OCR for page 56
Research Recommendations: To Facilitate Distributed Work and easier to use with some combination of alphanumeric display, voice prompts, and audio input from the user. Even in an age of growing computer networks, "plain old telephone service" remains the most ubiquitous means of sharing information. Improving audio processing technology and equipment to facilitate working at a distance. This effort should include research aimed at better managing and using captured audio as a data type, and improving noise and echo cancellation in less-than-ideal audio environments. Reducing the human support currently needed for the use of widespread, complex networks by distributed workers. Currently, wide area network services, whether provided directly through Internet connections or indirectly by value-added network vendors, generally require large amounts of human-provided technical support. This is especially true when end users do not wish to learn about underlying network technologies. Research aimed at better automating long-and short-term network connections as well as ongoing network trouble-shooting and maintenance would be valuable to all individuals engaging in distributed work and to most other network users as well. Improving the quality and reducing the cost of powerful, high-quality computer input and output devices. While these improvements would aid all workers, distributed workers in particular must often process large amounts of information with little support from others. Improved scanning, optical character recognition, and computer display systems would be very helpful. Better interfaces are needed for browsing, annotating, retaining, and communicating the context of information. Improving the reliability of communications and computing systems to reduce the chance of telecommuters becoming isolated due to system failures. Networks and multiuser systems of all types will need to have the same or better reliability as today's stand-alone, single-user systems. CONCLUSION Many observers believe that the U.S. economy is in transition to a postindustrial model. While the exact outlines of this new economy are not yet completely known, it appears that the nation would be well served by enhancing workers' capabilities to engage in distributed work and by carefully considering its wider adoption in order to increase locational flexibility, provide expanded employment opportunities, and make better use of our physical transportation resources. The research recommendations of the Committee to Study

OCR for page 56
Research Recommendations: To Facilitate Distributed Work Technology and Telecommuting: Issues and Impacts will by no means be the final word on the topic. New opportunities for research will appear as the nation's work practices change and its computing and communications technology continues to advance. The committee is convinced, however, that the research it has recommended will provide major new technological capabilities to enhance and extend the use of distributed work in the nation.