THE COLLABORATORY APPROACH(1)

Arnold M. Peskin*
Brookhaven National Laboratory
Upton, NY 11973
phone: 516-344-416/fax: 516-344-3211/email: peskin@bnl.gov


ABSTRACT

A "collaboratory" has been defined as a center without walls, in which researchers can perform their work without regard to geographical location. To an increasing degree, engineering design and development is also taking the form of far-flung collaborations among divisions of a plant, subcontractors, university consultants and customers. It has long been recognized that quality engineering education presents the student with an environment that duplicates as much as possible that which the graduate will encounter in industry. To that end, it is important that engineering schools begin to introduce the collaboratory approach in its preparation, and even use it in delivery of subject matter to students.


WHAT IS A COLLABORATORY?

A "collaboratory" has been defined as a center without walls, in which researchers can perform their work without regard to geographical location. [1] It is a term representing a suite of technologies and techniques for the conduct of science that is gaining widespread acceptance among research universities and scientific institutions. Collaboratories are beginning to replace face-to-face interactions as collaborations become larger and more geographically dispersed. It can be reasonably expected that large engineering projects will not be immune from this trend.

The U. S. Department of Energy (DOE) national laboratories represent fertile ground for the growth of collaboratories in science and technology. They are the custodians of large, expensive and unique scientific instruments such as accelerators, reactors, and supercomputers, whose design and use call for large collaborations on an international scale. These institutions also have long-standing mandates to transfer technology to industry and to provide educational resources in science and technology. The DOE has recently established a project called DOE 2000 which is explicitly funding developments in collaboratory infrastructure and establishing pilot projects of this nature.

One such institution, Brookhaven National Laboratory, is in many ways typical of such institutions. Among its world-class scientific instruments are the National Synchrotron Light Source, the Alternating Gradient Synchrotron, the High Flux Beam Reactor, and the Relativistic Heavy Ion Collider currently in development. It is also the base for unique information repositories such as the National Neutron Data Center and the Protein Data Bank which are referenced daily by hundreds of researchers all over the world. Furthermore, it is the home for thousands of scientists, engineers, and technicians, many of whom use these facilities or similar ones at other locations. Brookhaven has had to develop an imposing infrastructure to support these activities and people, including advanced computational, communications and visualization facilities. And as is the case at other Laboratories, it has very active Technology Transfer and Education offices. Most recently, its Office of Educational Programs has placed particular emphasis on engineering education, including the often neglected education of technicians and technologists. Hence, it is a rather a natural extension to consider establishing collaboratory relationships for engineering education.

EARLY COLLABORATORY EXPERIENCES

The BNL Computing and Communications Division has already engaged in a number of activities that might qualify as precursors to a fully formed collaboratory. They have been conducted as Cooperative Research and Development Activities (CRADAs in DOE parlance) with industrial partners. Some of these have left as a "residue" important infrastructure improvements that further enable the collaboratory concept. Three such projects and one proposal for future work are briefly described below.

A CRADA with NYSERNet, Incorporated, an internet service provider, is investigating the provision of Asynchronous Transfer Mode (ATM) wide-area networking. NYNEX and Sprint, though not formal CRADA partners, have also been participants. When the project began in 1994, it was far from clear that ATM was the likely next-generation internet technology, but that has fortunately turned out to be the case. This project established wide-area testbeds, connected them to ATM local area networks, evaluated commercial hardware and software, investigated problems of interoperability, facilitated applications and demonstrations, and stimulated the creation of high-speed commercial network offerings. This work seems particularly valuable in its relevance to the Internet II project, a nationwide development proceeding under the direction of the National Science Foundation, which uses much the same technology. [2] Figure 1 shows the ATM Testbed developed and used for this work.

AMTEX is the name of another CRADA, this one involving a number of DOE Laboratories and four Universities; the industrial partner in this case was no less than the entire American textile industry. Within this framework, several projects were implemented, among them are Demand Activated Manufacturing Architecture (DAMA), Computer-Aided Fabric Evaluation, Textile Resource Conservation, and Cotton Biotechnology. The DAMA project was the first and largest project. Its mission was to define, develop, integrate and deliver a system and structure to be used by all elements of the fibers, textiles, fabricated products production, and retailing chain to enhance productivity and competitiveness. One aspect of this project was an industry-wide information infrastructure, including networking, data bases and browsers. While AMTEX was a particularly large collaboration, Brookhaven's computing staff worked particularly closely with one of the University partners, Auburn, in the information infrastructure area. [3]

One potentially large collaboration, currently in the proposal stages, emphasizes another aspect of collaboratory structure, that of documentation and publication. The "industrial" partner in this case would be the American Physical Society, which publishes a number of prestigious publications for the scientific community. This effort, called E-PUBS 2000, envisions a unified electronic document system "from author to archive", handling production, editing, peer review, publication, distribution and archiving of scientific documents. As this paper is written, the proposal is under review.

Perhaps the best recent example of a prototypical collaboratory at Brookhaven has been a project sponsored under the Advanced Computational Technology Initiative (ACTI) program of the National Gas and Oil Technology Partnership. It is entitled "Synchrotron Computed Microtomography: Fluid-Rock Studies", but at Brookhaven it is referred to as the ACTI project for brevity. [4, 5]

The initial objective of this project was to create a facility for synchrotron computed microtomography to better understand and improve petroleum reservoir management. The industrial partners included Mobil Oil and GTE. Reservoir and seal rock core samples provided by Mobil were analyzed at the Brookhaven National Synchrotron Light Source and high-quality microtomographs were produced. A conference room sized stereo display facility has been installed for this project, where the core samples can be viewed and navigated. The entire analytic pipeline, from the NSLS instruments to the imaging computers, was put on an experimental high-speed network testbed to vastly reduce the time from data taking to analytic results. The system is depicted in Figure 2.

There are three enabling technologies that made this activity possible. The first is the Light Source itself, a major accelerator which is used by thousands of researchers each year. A second is the ATM networking that was pioneered in the previously mentioned CRADA with NYSERNet. Finally, there is the 3D Visualization theatre, which produces high-quality stereoscopic images that allow researchers to view the sample with a remarkable degree of realism. In fact, it is possible to replicate the theatre at the collaborator facilities in Boston and Dallas, and all be viewing the same object as well as each other.

The visualization theatre in particular has spawned other applications in diverse scientific disciplines that had not been intended. Its use has been demonstrated in protein chemistry collaborations by the Protein Data Bank staff. It has also found natural application in viewing large, complex mechanical computer-aided designs. Perhaps the most useful purposes to which it might be put are those in the field of medical imaging, where it may have an important role in comparing and co-registering two complementary 3D images (for example anatomy and dosage distribution). Again, the vision is of action at a distance, with the patient at one location, and an expert diagnostician perhaps thousands of miles away.

RELEVANCE TO ENGINEERING EDUCATION

It is widely recognized that there are systemic deficiencies in technician education [6,7]. Industry complains that new technician graduates are often ill-prepared for their careers. One reason is that many of the educational institutions lack the resources or state-of-the-art knowledge to provide truly adequate preparation. Improved linkage to industry/government research programs can go far toward bridging that gap.

It has also long been recognized that quality engineering education should present the student with an environment that duplicates as much as possible that which the graduate will encounter in industry. Engineering design and development also often takes the form of far-flung collaborations among divisions of a plant, subcontractors, university consultants and customers, making the collaboratory approach applicable to performing that work. To that end, it is important that engineering schools begin to introduce the collaboratory approach in its preparation, and even use it in delivery of subject matter to students. Such tools as are needed for this kind of interaction are not necessarily beyond the grasp of educational institutions, even small, remotely located ones. Indeed, these are the very institutions that would stand to benefit most from enhanced off-campus interaction. In any event, these are not developments that any educational institution can long afford to ignore.

Recently, discussions between Brookhaven and staff at a number of prominent Engineering Technology institutions has emphasized enriching a vital, yet historically neglected segment of the science and technology education spectrum, the Engineering Technician and Technologist. This is to be accomplished through a program of focused technology transfer, visiting professor appointments, student internships at Brookhaven along with network-based interaction with their home institution. The initial discipline addressed will be computing and telecommunications, with applications in distance learning, multimedia delivery, and medical imaging.

The overall objectives are to enhance both the curricular content and the computing infrastructure of the co-operating institutions and bring them together into a network connected "virtual campus". In this context, virtual campus refers to a collaborative environment wherein each participating institution can share in lectures, library holdings, coursework and even laboratory projects. Just as visiting students and faculty become full participants in facility design or conduct of experiments while on site, much of this cooperative work can also be performed at a distance in specially designed courses back at the home campus. Once success is demonstrated through this model, it allows for follow-on expansion of the program to include other disciplines of common interest, such as environmental science, and expansion of the number of participating institutions. The program can also serve to enhance the professionalization of technician careers by enhancing both the capabilities and the self-esteem of the participants.

The virtual campus is facilitated by robust network connections and realistic audio-visual presentation capabilities. ATM technology will provide the high-speed networking element. A 3D conferencing facility of the type developed in the ACTI project provides the presentation aspect. Notwithstanding the advanced nature of this technology, such a setup can be readily replicated at other locations by trained individuals. One of the outcomes is to provide that training and help effect the replication. The virtual campus can then take its place as a powerful tool at each institutional campus, and become an entity that will also serve as a model for expanded collaborations with other institutions.

Figure 1. NYSERnet ATM
Test Configuration up

Figure 2.
The ACTI Microtomography System up

The primary focus of such a project is the student. Every effort would be made to ensure that the student's experience is a rewarding one. Great care will be taken to make each assignment achievable, yet a valuable experience in both the use of technology and in scientific problem solving. The living arrangements and social aspects will not be neglected, with participants having opportunities to share campus life with each other as well as the many BNL visitors, including other students, faculty and some of the world's most distinguished scientists. Students who do not directly partake of the on-site portion of the experience are nevertheless an important part of the program; the participating team can carry concepts and techniques to their home institutions that will manifest themselves in significant curricular and infrastructure improvements and thus have its effect on all ET students at the participating schools.

The role of the faculty participants is particularly crucial. They are primarily responsible for dissemination of the knowledge gained back to their home institutions and more broadly to their professional community. It is they who will integrate the newly acquired knowledge at their institution to document and build upon the Brookhaven experience. New coursework in advanced computer networking, computer graphics, and even generic problem solving are among the anticipated outcomes. And whereas the student participants will likely graduate and leave their campuses within a year, the faculty participant will carry the legacy of the experience for many years to come.

OUTLOOK

The program outlined above anticipates that the experiment would be replicated by many research institutions and post-secondary schools. It has been designed for ready replication, anticipating the widespread recognition that are many mutual benefits to be derived. The schools will have at their disposal new methods of course delivery, new partners in education, new courseware and new infrastructure. The research institutions will have fulfilled their educational mandate more effectively , helped insure a more useful workforce, and more effectively reaped the many benefits of closer ties to academia. The vision, then, is for a future time when collaboratories are no longer considered experimental, and their widespread use in industry is accompanied by their infusion into the educational experience.

In order to realize this vision, many things must be done. It will be necessary to have a much greater level of cooperation among educational institutions, and between the educational sector and industry, than is now the case. All relevant institutions, the colleges, research laboratories, and accrediting agencies must adopt a forward looking perspective. It will also require the encouragement of governments around the world in order to achieve an international scope. Notwithstanding these challenges, the force of inevitability and the promise of mutual benefit should be welcome allies in this quest. (8)

BIBLIOGRAPHY

1. R. Kouzes, J. Myers, and W. Wulf, Collaboratories: Doing Science on the Internet, IEEE Computer, Vol. 29, No. 8, August 1996

2. W. Graves, Why Education Needs an Advanced Internet, IEEE Computer, Vol. 29, No. 11, November 1996

3. L. Anderson, R. L. Cheatham, A. Peskin, AMTEX-A University, Government, Industry Partnership, Proceedings of ASEE 1994 Annual Conference, Edmonton, June 26-29, 1994

4. A. Peskin, B. Andrews, B. Dowd, K. Jones, P. Siddons, Microtomography with 3-D Visualization, Proceedings of SPIE Annual Meeting, Denver, June 1996

5. M. E. Coles, R. D. Hazlett, E. L. Muegge, K. W. Jones, B. Andrews, B. Dowd, P. Siddons, A. Peskin, P. Spanne, W. E. Soll, Developments if Synchrotron X-Ray Microtomography with Applications to Flow in Porous Media, Society for Petroleum Engineers 1996 AnnualTechnical Conference, Denver CO, October 6-9, 1996

6. N. Lane, Putting the Pieces Together, Proceedings of the National Science Foundation Conference BUILDING THE SYSTEM: MAKING SCIENCE EDUCATION WORK, Washington DC, February 24-26, 1994

7. Workshop to Define a National Agenda for the Future of Engineering Education, sponsored by the National Science Foundation, Sinclair Community College, Dayton, October 26-28, 1995.

8. V. Hendley, The Basics of Successful Joint Ventures, ASEE Prism, January 1997

1. 1 Work performed under contract number DE-AC02-76CH00016 with the U.S. Department of Energy


Back to Table of Contents