Current Vision and Implementation Strategy

CBI's Current Vision, long and short term DSS Direction

After two years in formulation BRD is very close to an agreed on Framework Document for DSS. In many ways it can be said that the development of Spatially based Decision Support Systems parallels the original development, deployment and acceptance of many new information technologies, GIS a good earlier example.

Over the past decade and a half GIS have migrated from mainframe extensions of automated cartography to powerful workstation based tools for the manipulation of topologically based spatial data.

To an extent, SDSS (PBDSS, GEDSS) are poised to extend that functionality with the inclusion of specific scientific models and visualization tools that can allow decision makers at various levels in the decision process to examine the problems they confront with increasing degrees of flexible complexity.

It has been said that DSS's do not provide just answers, per se, they allow the human deciders to 'think with the data'.

The work of the Aurora Project and the IGDS is pointing towards an increasingly sophisticated path for the achievement of this kind of functionality.

What has been discussed at recent meetings (Denver July 2000, ) is a model of decision support that goes beyond the use of a GIS to display the relationships of spatial scientific understanding, requiring the close collaboration of GIS specialist and Scientific mater expert. The 'loose coupling' (Dr. Paul Densham, in conversation ) has drawbacks in that it require each science model to be brought in the GIS decision space one at a time, to work on a specific project at a time, with little in the way of true extendibility.

Only recently has the GIS world moved away itself from such a pathway, increasingly specific modeling classes of functionality are included in the core functionality of fully featured GIS software. Examples include location-allocation modules and business-oriented demographic tools. What has made this possible is the inclusion of the underlying algorithms into the GIS topological data structures. These models can be called directly into the GIS interface (without extensive reprogramming) to work on applicable spatial data and instantly provide the what if functionality that is so much at the heart of the DSS concept.

To date though, macro spatial ecological-behavioral models per se, those of immediate interest to land management decision makers, are not inventoried and as easily available to practitioners as they might be. Such Functionality is not commonly included maintaining the reliance, in the natural resources realm, largely on the loose coupling mentioned above. Discussions by IGDS and others now center on a concept that would allow the extensions of models that in turn would fuel DSS capabilities by pursuing a strategy that has already been employed for spatial Data. Conceptually think of a 3D pyramid whose facets represent the necessary infrastruture to fully empower distributed work with Spatial DSS systems and other tools, at various scales, and with out the continuing need to reprogram a data,model, computational capabilities, locally, in each place or time that it is required.

CBI will also pursue the identification of more purely biological models and tools and linkages to partners in these areas. Many of these tools and models are not spatially focussed, representing the more underlying theoretical and primary causal mechanisms of biological systems, (genetic and chemical determinants), and here too the inventory and implementation of knowledge based support achieved it's full potential.

So while pursuing an eventually mature integrated knowledge, rule, model based Decision Support scenario as it's goal, CBI will attempt to identify and utilize smaller pieces of that design

To that end, one of CBI's interim (near term) goal will be to pursue the identification and assemblage of a suite of modular, and initially, primarily spatially focussed DSS tools that while not providing the fully featured modular, model base and rule base product envisioned in the long term, it does take positive step of concentrating in a single interface, a growing suite of tools and services, that have been identified as desireable in a DSS context. Furthermore, the assemblage of such a suite decreases the need for their recreation in each instance of a DSS project, a reduction in cost of development. Although still requiring the loose coupling with scientific expertise, and for the moment, still localized to particular platforms. The existence and wider adoption of such a quasi-standard assemblage, would enhance exchange of models where they do exist since in cases where they have been created, they will be available for use elsewhere.

Currently CBI is pursuing such an assemblage, focused for several reasons on ESRI ArcView extensions like Smart Places and Community Viz and possibly the US Forest Services EMDs The primary reason is the widely deployed user base of ArcView in the USGS, and it's federal partners, as well as academic institutions, and NGOs, increasing the likelihood of robust and to an extent integrated modules accessed through a well known common user interface.

We have been working with PIERC in HI, to have them use and evaluate these available 'common' tools and provide the scientific expertise for the requisite loose linkage to place appropriate ecological models in their research arena.

Hawaii was chosen as a good fit to test the usefulness of decision support tools in the hands of government and community groups with differing valuations of outcomes.

CBI will continue to seek other partners and venues in which to vet such tools as we identify Clearinghouse for Information about DSS products and progress, with in government, academia and the private sector providing links where they exist to already established sites providing similar information