David J. Coleman and Rupert Brooks
Department of Geodesy and Geomatics Engineering
University of New Brunswick
Fredericton, N.B., CANADA
Digital mapping and charting processes have traditionally been sequential in nature and were originally designed in an environment where work was completed in a single location. As contract production became more prevalent, this system was extended to permit shipping of the source materials between locations via courier. Until recently, wide-area telecommunication services were regarded as being too slow, complicated and expensive to be considered seriously in support of inter-office workflow and production processes.
Recent developments in spatial data processing and broadband data communications may provide users hundreds of miles apart with access to the same equipment and data at -- in some cases -- comparable levels of performance. When considered in combination with emerging groupware products and new approaches to computer-supported cooperative work (CSCW), these developments may offer a new approach to the collaborative production of digital maps and nautical charts -- enabling some processes to be conducted concurrently rather than sequentially.
This paper introduces the rationale and activities involved in research efforts now underway in applying collaborative production technologies and groupware tools to selected problems in digital data production, quality control and distribution. After describing some of the problems and concepts under examination, the authors conclude with a brief description of the research currently underway.
Digital mapping and charting production cycles contain many potential bottlenecks which may delay the final delivery of data to customers. From field data collection and original production through inspection, correction, initial distribution and recurrent updating, the process is largely sequential in nature [Coleman and McLaughlin, 1988]. While intended to produce a reliable product as quickly as possible, the data handling processes involved were originally designed in an environment where work was completed in a single location and complete sets of source materials were shipped in bulk from one unit to another -- usually in the same building.
As contract production has become more prevalent in government digital mapping and charting programs in Canada, this "process" has been extended to permit shipping of the source materials between locations -- either across town or even across the country. Until recently, the materials were typically shipped by courier, since the bulk transfer of digital spatial data files via telecommunication networks has been regarded as being either too slow, too expensive, too complicated or too untrustworthy for routine use ([Craig et al., 1991,] [Newton et al., 1992]).
Situations are increasingly arising where the data is collected by a supplier in one location, checked by staff in a different city, returned to the supplier (or perhaps even the field crew) for correction or verification, returned to the central office and then distributed to suppliers in various centres ( from [Coleman, 1994b]. Depending on the situation, source materials associated with each chart may travel tens, hundreds or even thousands of kilometres during the various production, distribution and updating processes. This is happening at a time when digital map and chart providers are under increasing pressure to bring their original and updated products to market in shorter time frames than ever before.
Recent developments in computer hardware technology and broadband data communications promise to change this situation, with new higher-speed communication services providing users hundreds of miles apart with access to the same equipment and data at -- in some cases -- comparable levels of performance [Coleman, 1994a]. These new services offer promise to individual users who may wish to use wide-area networks to display, manipulate and transfer large data files stored on remote systems. As well, these services may offer a new approach to the collaborative production of digital maps and charts -- enabling some processes to be conducted concurrently rather than sequentially.
Hardware, software and procedures to support collaborative production -- or, more specifically, computer-supported cooperative work (CSCW) -- have been discussed for more than thirty years (e.g., [Englebart, 1963], [Chapanis, 1975]). Particularly since the advent of computer networks, CSCW research has accelerated and a number of researchers have attempted to place these developments in some kind of framework (e.g., [Licklider et al., 1978]; [Grief et al., 1985]). CSCW has now reached the level of notoriety where collections of proceedings are available on the subject (e.g., [Baecker, 1993]; [Coleman, 1993]). In the corporate world, shared access to corporate resources and innovative new approaches to collaborative production using "groupware" tools like Lotus Notes(TM) are now being investigated [Marshak, 1993]. Significantly, behavioral research into the social and organizational interactions between members of a workgroup (e.g., [McGrath, 1984]) is now being applied in a corporate network setting ([Grudin, 1990], [Sproull et al., 1991]).
Whether we are dealing with operations in a single location (i.e., on a local area network) or multiple locations (i.e., across a WAN or LAN interconnect service), improving production throughput times will depend in large part on three things:
(1) shortening the "production float" -- the transportation, storage, handling and "sitting" time(s) consumed when files are moved from one location (or one stage of production) to another;
(2) streamlining existing methods or adopting alternative approaches to product development;
(3) improving communication to ensure: (a) a common understanding of the product requirements between the producer and the inspector; and (b) that each product ultimately meets stringent product specifications.
Collaborative production supported through broadband networks will help support (1) and (3) immediately and -- over the longer term -- may support changes to the production processes as well.
While authorities may disagree on the strict inclusion of electronic mail as being a form of "groupware", it is clear that e-mail has already made a major contribution to facilitating production by improving communications within and between "wired" workgroups in many organizations. However, collaborative production supported through broadband networks depends on more than just electronic mail. Specifically:
* Such a concept implies the use of a shared "database" or collection of files, and should permit the definition of group members' roles, task status reporting & tracking, and gateways to electronic mail and other sources of data.
* Such systems should permit the organization of correspondence, comments, reports, etc. associated with a project or product and should support the management of multiple versions of objects (e.g., images, vector-based charts, video and sound).
* Finally, preliminary research has already indicated that such systems should give two or more remote users the capability to simultaneously view the same file, modify or add comments to specific entities where necessary, and communicate via voice, video and/or e-mail while making these changes [Coleman, 1994b].
The development and implementation of such capabilities will be predicated on an integrated collection of tools and functions which might include the following:
* Mail, audio and perhaps even desktop video communication between Inspector, Production Supervisor(s) and technician(s);
* High-speed file transfer between production, inspection and distribution facilities;
* Network management tools which permit the access and viewing of files stored on remote servers/networks (mostly available already through tools like NFS and AFS);
* Simultaneous viewing and manipulation of the same file by two or more different users in different locations;
* Electronic "Markup" of entities requiring further attention either during individual or "shared" sessions (i.e., analogous to attaching "post-it" notes or comments to a hardcopy map sheet);
* Software which enables two users to share control of the same workstation; i.e., controlling the workstation of a remote user and seeing what that user sees.
Many of these capabilities already exist in some form from a variety of hardware and software vendors. However: (1) the level of integration is low -- individual capabilities are scattered among a wide variety of packages; (2) they have not been customized to meet the specific needs of users in digital mapping and charting environments; and (3) the performance and limitations of these tools in geomatics applications across wide-area networks is still to be determined.
Research now underway by the Geographical Engineering Group at the University of New Brunswick is re-examining the requirements of selected digital map and nautical chart production, quality control and updating processes as mapping organizations and their suppliers take fuller advantage of forthcoming broadband communications technology. The overall research program includes the following stages:
(1) Preliminary research (now underway) which: (a) examines the characteristics, strengths and weaknesses of existing sequential production models; (b) develops a prototype collaborative production model (or models) for digital map and chart production in self-contained and distributed operational environments; and (c) identifies hardware-, software- and operational constraints to collaborative production;
(2) Specification and development of prototype software which enables collaborative production, inspection and correction of digitized chart files in a wide-area network environment;
(3) Performance testing of these software packages across a broadband, wide-area network service (in comparison with stand-alone and LAN-based systems) to begin identifying optimal approaches to collaborative production and delivery;
(4) Identification and classification of collaborative production tasks which: (a) absolutely require broadband connection to be carried out; (b) may be acceptably completed across lower-speed services now enjoying widespread usage; and/or (c) those which may be temporarily redesigned to be handled on lower-speed links; and
Stages 1 and 2 are now underway. Individual components of this research are currently being supported by the Natural Sciences and Engineering Research Council of Canada (NSERC), the Atlantic Canada Opportunities Agency (ACOA), The Champlain Institute, as well as a variety of private firms.
Early research in two projects (still underway as of August 1995) has provided valuable input to date. These two projects are described briefly below.
The ChartNet Project is a multi-million dollar effort entailing the development and integration of a suite of software packages to permit the collaborative production, inspection, management and distribution of electronic nautical charts across broadband communication networks [Coleman, 1994c]. Led by Nautical Data International of St. John's, Newfoundland, the team involved in the development of ChartNet Stage 1 included Compusult Ltd., Universal Systems Ltd., ORACLE Canada, IDON Corporation, the Canadian Hydrographic Service (CHS), Environment Canada, the Canada Centre for Marine Communications and the University of New Brunswick.
The first stage of the ChartNet project focused largely on developing and implementing a more data-driven approach to electronic chart compilation and management. However, as part of this stage, Universal Systems Ltd. did develop a special electronic whiteboard module for its CARIS GIS software. This module provides separate users on the same LAN or WAN the ability to view the same GIS graphics file simultaneously for purposes of joint markup and comment. Rather than simply viewing the same static image, the users can share (or, rather, trade) control of the display and tracking functions in order to permit zooming and panning to specific portions of the vector digital map image. While only one user holds control over final edit or revision, both users may "mark" specific features or items of interest on the display for further review or comment. This software module is now installed and undergoing advanced testing at the offices of Nautical Data International in St. John's and at CHS Headquarters in Ottawa.
The New Brunswick Geographic Information Corporation (NBGIC) is committed to bringing its entire province-wide 1:10,000 digital mapping coverage to a common level of topological structuring and consistency during the 1995/96 government fiscal year. Rather than performing the work in-house, the residual editing and substantial re-processing involved will be completed through a series of production contracts to GIS service firms. In addition, all processing, file inspection and verification activities comprising the Project's Quality Control are being contracted to a separate independent consulting firm.
There are very stringent penalty clauses associated with late delivery of the files from both the contractors to the Inspector, and from the Inspector to NBGIC. With over 1800 separate files to be received, inspected, assessed and either approved or passed back to one of the contractors for re-work, the prospects of encountering delays due to disorganized handling, learning-curve difficulties, poor inter-organization communications or unnecessary duplication of effort are significant.
The conventional workflow and handling procedures developed for this project are based in part on a previous province-wide effort involving inspection and correction of DTM coverage. While this project is now underway, members of the Research team have been examining procedures with a view to streamlining (or even totally redefining) selected processes which could be handled within single LAN, over a proprietary LAN-Interconnect service or across the Internet.
Preliminary findings indicate that many of the workflow improvements may be achieved through the use of a shared server, improved file "check-in/checkout" procedures, and a common project bulletin board. Enhanced WWW and Lotus Notes server -based systems are both being investigated over the coming year and findings will be reported in a future paper.
While groupware products and CSCW approaches clearly represent a significant advancement in the larger industrial community, their effective application to problems in geomatics data collection and management requires further study and testing. Further research into collaborative production in GIS could have a significant impact on the streamlining or even outright redefinition of future production, quality control and updating activities in major mapping and charting organizations around the world. By examining the re-engineering of generic mapping processes to take advantage of new broadband communications services, such investigations tie together current developments and interest in geographic information systems, broadband communications and spatial data infrastructure.
Baecker, R.M. [1993]. Readings in Groupware and Computer-Supported Cooperative Work: Assisting Human-Human Collaboration. Morgan-Kaufman Publishers, San Francisco, CA. USA.
Chapanis, A. [1975]. "Interactive Human Communication". Scientific American, Vol 232, No. 3, pp. 36-42. W.H. Freeman and Company.
Coleman, D.J. and J.D. McLaughlin [1988]. "The Shift to Digital Mapping: Points of Impact on Management." Journal of Surveying Engineering, Vol. 114, No. 2, pp. 59-70. May. American Society of Civil Engineering.
Coleman, David D. [1993]. Groupware 93: Proceedings of the Second Conference. D.D. Coleman, ed. Morgan-Kaufman Publishers, San Francisco, CA. USA.
Coleman, D.J. [1994a]. "Geographic Information Systems Performance in a Broadband Communications Environment". Unpublished Ph.D. Dissertation. Department of Surveying and Spatial Information Science, University of Tasmania, Hobart, Tasmania, Australia. February.
Coleman, D.J. [1994b]. Internal Workshop Summary: "ChartNet Workshop on Collaborative Production" Summary prepared for the Canadian Hydrographic Service and Nautical Data International, Ottawa, Ontario, in support of a CANARIE-funded project on application of broadband networks to electronic chart production. July.
Craig, W.J. and P. Tessar and N. Ali Khan [1991]. "Sharing Graphic Data Files in an Open System Environment." URISA Journal, vol. 3, no. 1, pp. 20-32. Spring.
Englebart, D. [1963]. "A Conceptual Framework for the Augmentation of Man's Intellect." in Vistas in Information Handling, Vol. 1, P. Howerton, ed., pp. 1-29. Spartan Books, Washington, D.C.
Grief, I. and S. Sarin [1987]. "Data Sharing in a Work Group." TOOIS, Vol. 5, No. 2, pp. 187-211. April.
Grudin, J. (1990). "Groupware and Cooperative Work: Problems and Prospects." In Readings in Groupware and Computer Supported Cooperative Work Assisting Human-Human Collaboration. Ed. Baecker, Ronald M. Morgan Kaufman Publishers Inc., San Mateo, California. P. 97
Licklider, J.C.R. and Vezza, A. [1978]. "Applications of Information Networks". Proceedings of the IEEE, Vol. 66, No. 11, pp. 1330-1346.
Marshak, D. [1992] "The Notes Phenomenon: The Industry Reacts to Lotus Notes". Office Computing Report, Vol. 15, No. 12, pp. 3 - 24. Patricia Seybold Group, Boston, Massachusetts, U.S.A. December.
McGrath, J. E. [1984]. "Methods for the Study of Groups." In Readings in Groupware and Computer Supported Cooperative Work Assisting Human-Human Collaboration. Ed. Baecker, Ronald M. Morgan Kaufman Publishers Inc., San Mateo, California. P. 200.
Newton, P.W. and P.R. Zwart and M. Cavill [1992]. "Inhibitors and Facilitators in High-Speed Networking of Spatial Information Systems." in Networking Spatial Information Systems, P.W. Newton, P.R. Zwart and M.E. Cavill, eds. Belhaven Press, London.
Sproull. L. and S. Kiesler [1991]. Connections: New Ways of Working in the Networked Organization. MIT Press, Cambridge, MA., U.S.A.
David J. Coleman, Ph.D., P.Eng.
Rupert Brooks
Department of Geodesy and Geomatics Engineering
University of New Brunswick
P.O. Box 4400, Fredericton, N.B.
CANADA E3B 5A3
David Coleman is Director of the Geographical Engineering Laboratory at UNB. He obtained his Ph.D. in Surveying and Spatial Information Science from the University of Tasmania (Australia) in 1994, where his research dealt with determination and prediction of GIS performance across broadband wide-area networks. Prior to returning for post-graduate studies in Tasmania and at UNB, he held senior executive-level positions in the mapping and GIS industry in both Canada and the United States over a 15-year period. He may be reached by E-mail at <dcoleman@unb.ca>.
Rupert Brooks is a senior undergraduate student in the process of completing his B.Sc. thesis in Surveying Engineering and, as of September, 1995, enjoying his first month of marriage. His Internet E-mail address is <c3wl@unb.ca>.