Democratizing GIS: Are We There Yet?

David Paige, President, DataLOGIC, Inc. 1120 Bromley Road, Avondale Estates, Georgia 30002, U.S.A.

Roy Mead, Ph.D., 8204 Trolley Square Crossing, Atlanta, Georgia 30306, U.S.A.

Abstract

Over the past several years we have used the term "democratization" to describe the point at which a technology has been fully adopted by the people in an organization and used in their daily work. This paper describes the GIS democratization process in the Southern Region of the U.S. Forest Service. Our strategy of democratizing GIS is centered on delivering information to our customers to support collaborative problem solving in a timely manner. Put simply, it focuses on providing GIS information to people who want it, when they want it.

We illustrate our democratization experience by examining the deployment of three key GIS technology applications. This deployment is part of a common approach in which business processes, data and data standards, hardware and connectivity, and user skills are brought together in the proper sequences. This paper is directed toward organizations which have moved beyond the typical start-up activities of building databases, and provides suggestions on how to approach and measure GIS democratization.

GIS is a new order of things in the U.S. Forest Service, at least at the enterprise level, where GIS is widely used, with standard processes and data. Three regions of the Service currently have what could be considered operational enterprise GIS's - the Southern, the Pacific Northwest, and the Alaska regions. Although developed independently, the three systems have basic similarities in construction and function.

The Southern Region of the Forest Service manages over 12 million acres (5 million hectares) of public lands in 13 states from Oklahoma to Virginia, and in the Commonwealth of Puerto Rico. Recently, its enterprise vector GIS database, collected at the 1:24,000 scale, was completed after 8 years of work. The region is also in the midst of a massive 4 year phase-in of a Unix based client-server computer system. Agency-wide GIS capability was the driving force for this effort, and it has been integrated with office automation and other information systems as well, to provide all employees with a common computing environment.

Figure 1 illustrates the coordination between all aspects of our ongoing technology upgrade at the organizational level of the Southern Region. Depicted specifically are the aspects of enterprise GIS deployment, but the sequencing is applicable in general to technology roll-outs in the Southern Region. Comprehensive Information Needs Assessments (INA's) took place first to define user groups and their business needs. Identified needs were based on existing business practices and on new practices born out of new technology capabilities. GIS enterprise core data needs were also identified in the INA's and data collection efforts were funded accordingly. Pilot sites were then equipped with computers to begin the development of user applications, and standards for collecting, maintaining, analyzing and distributing data.

Key region-wide GIS applications were then developed as waves of new equipment were installed, and agency investment in employee skills rose to support the increasing number of operational GIS sites. Ancillary data acquisitions followed the core data collection to augment total GIS functionality as new business needs emerged.

We believe that this complex sequence of events is necessary to facilitate the deployment of the GIS technology in the Southern Region. But technology deployment and actual customer use are not the same. Our intent is to democratize GIS information, that is, to make available to all users whatever information they want, when they want it. The delivery system is a combination of the GIS data, applications, computers, telecommunications networks, and skilled people.

So, the question is "How do we democratize GIS in our organization and how do we know when it is democratized?" To answer that question, we need to first look at how innovations spread through organizations.

First some definitions. In this paper we use the term "diffusion" for the process by which GIS use propagates through an organization, by a pattern of adoption by individuals. Adoption is the acceptance and use of GIS by individuals in their day-to-day work. Democratization, finally, is the cumulative measure of adoption of GIS in the organization. We have found that understanding our diffusion processes was fundamental for democratizing the GIS technology in the Southern Region.

Diffusion research is a body of science with roots in the 1940's, and now thousands of articles and texts are available in this field. Rogers (1983) defines diffusion as "an innovation that is communicated through certain channels over time among the members of a social system".

Knowing the diffusion channels or "pathways" in your organization allows you to successfully deploy an innovative technology. These organizational pathways are composed of:

The types and levels of encouragement available in the organization to promote adoption, such as peer recognition, awards, and management mandate.

Basic organizational commitment or willingness to support an innovation by funding, equipment, skills acquisition, etc.

The distribution and characteristics of individuals in the user community. The community can be segmented into category groups based on individual skills, such as expert and novice GIS users. Innovativeness can also be used to describe groups, such as innovators, early adoptors, and laggards, as shown in Figure 2. Knowledge of the diffusion pathways of your organization is essential to develop a successful democratization strategy.

Communication modes which determine the efficiency or speed of innovation spread. Both the type of connectivity (hierarchical or interconnected) and media (electronic, paper publishing, etc.) are part of this aspect of the organizational culture.

The diffusion pathways of an organization determine how individual adoption spreads. Adoption is an individual's choice to use an innovation in their daily work. On this individual level, the facets of adoption are:

The innovativeness of the user - the willingness of an individual to embrace an innovation.

The four tests of adoption -- items the user must answer in the affirmative for his or her adoption of an innovation:

Does it meets my business needs?
Is it available and reliable?
Is it easy to use?
Is it fast?

Adoption thresholds - these thresholds apply to each of the four tests of adoption, and vary by the innovativeness of each user. For example, an "innovator" might be satisfied with a lower ease of use than a "late majority", but less satisfied with a slow response time.

We believe a successful democratization strategy is based upon familiarity with our organizational pathways and the facets of individual adoption, and most importantly, on working relationships with the innovators and early adopters in our organization. Next, we present an example of a GIS democratization experience in the Southern Region. It is based on developing an integrated GIS data maintenance and delivery environment.

THE GIS MAINTENANCE AND DELIVERY (M&D) ENVIRONMENT

The Atlanta regional office administers 13 national forest offices throughout the Southeast, and the national forest offices in turn administer 99 ranger district offices. In order to provide high quality GIS performance at the ranger district offices, district GIS databases are maintained on local computer processors. Due to the commercial communications infrastructure in the Southeast, costs of centralizing computing would be prohibitive, based on the intense network loads created by a fully functional GIS community. So, we must somehow provide GIS support to our regional users for timely delivery of fresh information in this distributed computing environment.

The Marc editing program, the MTV metadata program, and the SDE network GIS server programs form the foundation of the GIS Management and Distribution environment in the Southern Region. The three interdependent applications are in different stages of deployment, and consequently of democratization as well.

First released in 1995, Marc allows our district editors to efficiently update our 99 vertically coincident GIS databases distributed throughout the Region. Marc was designed to specifically interface with MTV and automatically update its metadata tables as GIS feature edits are committed.

The MTV metadata application is of pivotal importance in the Southern Region for dispensing information from a GIS physically distributed across 13 states. Now in beta test, MTV is an ORACLE application that meets Federal Geographic Data Committee requirements. It will house and report out the metadata associated with all GIS datasets. ORACLE is the agency's relational database management system, and so will serve MTV and other agency ORACLE applications at each district office. Just as MTV interacts with Marc, it must also communicate with SDE.

Our implementation of SDE includes two interfaces - one for external Internet users and one for internal users running ArcView, and is currently in the alpha testing stage. SDE will warehouse and serve queries on a complete copy of the 12 million acre distributed GIS database. In order for SDE to maintain fresh GIS data in its warehouse, it must periodically reach out over the Forest Service wide area network and upload current copies of district datasets. Different layers will of course need to be updated at different frequencies based on the editing activity. This update process or regional "roundup" first contacts each district instance of MTV to determine if any edits of the layer of interest have taken place since the last roundup for that layer. If no edits have occurred, then the SDE copy of the data is deemed current and needs no update. If MTV reports that the data has been recently edited, then the district dataset is uploaded to the SDE warehouse and replaces the previous version. Then, the next district is contacted, in turn, until the entire regional GIS roundup is complete.

So, as the GIS M&D environment, the three key applications must work together and meet the four user adoption tests - to be available and reliable, easy to use, fast, and to meet user business needs.

Again, a democratization strategy is the approach or plan for diffusion of an innovation, based upon specific knowledge of the organization's people and culture. Our democratization strategy for the GIS M&D environment includes this sequence of tactics:

1. Target key individuals who will embrace good ideas and lead with very little encouragement. These are members of the "innovator" group.

2. Involve early adopters to refine the M&D environment, making sure that all four tests of adoption are met. Our veteran GIS database managers and editors are the keepers of business processes standards and maintain the information. Many are early adopters -- they refine our GIS business processes as the technology advances.

3. Use increasing levels of encouragement such as peer recognition and rewards to involve the early majority and some of the late majority.

4. At some point, the organizational energy required to encourage more complete adoption by the laggards becomes prohibitive. In Figure 2 this point is labeled the "mandate point", and an management mandate should be considered. When a mandate is made, the organization's encouragement effort drops to nearly zero. The intent of mandate is to "bring aboard" all who have not adopted the innovation, and its effectiveness is based on the how well mandates work in the organizational culture. Where there are few consequences for non-compliance, there will be less adoption among the laggards.

Referring to Figure 2, we see that the Marc, MTV, and SDE applications are all at different levels of democratization, due mainly to their different stages of development. We estimate that all three applications will be operating together by December 1998, thereby offering our GIS users a fully functional GIS environment. All three are proceeding upward on the diffusion curve as time progresses toward full democratization. No mandate has yet been issued, as we believe our user community is rich in innovators, early adopters and majorities, and has few if any laggards. Of course each user community has its own distinctively shaped diffusion curve, which again points out the importance of knowing the makeup of that community.

So, no, we aren't there yet, but we are moving forward. If we find that diffusion proceeds slower than expected, we should look first at how the innovation may not be meeting all four tests of adoption, and use the power of the early adopters, the keepers of the process, to help refine our offerings. We should honor our innovators for their willingness to risk, our majorities for their solid strengths, our laggards for keeping life interesting. And above all, let us remember that we are managing the "creation of a new order of things".

Machiavelli, N. 1513. The Prince.

Masser, I. and H. J. Onsrud, (eds.). 1993. Diffusion and Use of Geographic Information Technologies. Kluwer Academic Oublishers, London. ISBN 0-7923-2190-1.

Mead, R. and R. Johnston. 1994. Field-Level Diffusion Eases GIS Implementation Efforts. GIS World Vol. 7 No. 11.

Rogers, E. M. 1983. "Diffusion of Innovations", The Free Press, New York. ISBN 0-02-926650-5

Tobin, D. R.. 1993. Re-Educating the Corporation. Oliver Wight Publications, Essex Junction, Vermont. ISBN 0-939246-48-1.

Mr. Frye is a graduate of San Diego State University and California Polytechnic State University San Luis Obispo, in biology. He has been a fisheries and wildlife biologist with the Forest Service for 12 years. Since 1992, he has been in GIS and new technology management in the Forest Service regional office in Atlanta.

Mr. Paige is a graduate of the University of California Santa Barbara and the Georgia Institute of Technology, and specializes in assisting clients integrate their business processes with new technologies. Dr. Mead has a Ph.D. from the University of Minnesota in remote sensing and is a recognized expert in the application of raster and vector GIS technologies to the management of natural resources.