The objective of ADEPT Digital Earth (DE) development is to imagine, implement and evaluate next generation user interface paradigms for interacting with geospatial digital library (GDL) technology. Focii include the application of existing DE and related technologies to known problem areas, expanding the potential user base of DE + GDL technology, and the creation of customer feedback loops to drive further research and development in this area.

UN Aarhus Convention: "Each Party shall ensure that environmental information progressively becomes available in electronic databases which are easily accessible to the public through public telecommunications networks."

Three theorems of spatial user interfaces:

All data is related to other data
These relationships can be encoded spatially
Ergo, all data may be usefully spatialized

Approach

Our approach will be to work with multiple digital earth systems, so as to examine the issues involved with GDL visualization generally. We will work towards the creation of a generalized layer or set of middleware services that allow the rendering of library objects in any DE system.

The Virtual South Coast subproject is a lens for focusing work on actual applications of digital earth + geospatial digital library technology. The near-term objectives are to create a detailed, navigable in real-time model of the South Coast region (defined as approximately Point Conception to Carpinteria), and use this model as the basis for investigating user interface techniques. Library search result set visualization, gazetteer integration, collection-level metadata visualization and eventual experiments with spatial hypertext are all features planned in this approach.

Theorem: DE based applications are going to be region based, since that's where political, environmental, social & urban design solutions are needed.

For some applications and some collections, a visualization of the entire planet is preferable to the regional approach. We will continue to work with the Visual GIS group at Georgia Tech in applying their system to these applications.

internal objectives

• create customer feedback loop to drive DE dev
• have a good visual model to build parameter change demos on
• provide value-add for new ADL MW nodes
• platform for generalized investigation of spatial data & hypertext
• find the killer app for digital earth technology
• marketing & community visibility
• develop local techniques that will apply to entire DE

external objectives

• planning & development process support. focus on housing
• virtual tourism & city marketing
• education
• support for the local sustainability mafia
• support for individual researchers
• web journal/blog & digital picture collection interties.

possible collections

• housing information
• government & political information
• history & culture
• environmental information
• local maps & imagery subcollection
• personal collections (documents, annotations, pictures, journal entries)

ArcADL

First, client that goes from easy search and loads data directly into 2D. (dependency for alpha: one collection with the new access report to test it against) Later, when ArcView 9 and ArcGlobe is released, one-touch library data into terrain view problem should be solveable with minimal additions or changes to this client.

VirSoCo

a few static models with at least terrain, buildings extruded from footprints (possibly with textures), roads, imagery, and some trees. Campus first, then a full regional model, then perhaps downtown with other areas or neighborhoods as necessary.

some kind of parameter change demonstration, built on top of the static model

local collection development (metadata, database & ingest)

VTP to library interties (query based on region visible and selected point or area, result set visualization first in terrain view then maybe in globe)

spatial hypertext experiments

local collection metadata, DB & ingest mechanisms (ingest & collection creation dependent on middleware services)

VGIS deliverables

Launch simple query against middleware from VGIS, and display either footprint geometry or icons representing the results. The result would look something like this, modulo the link lines connecting the viewpoints.

More robust ingest functionality, perhaps via a GUI dataset assembly tool that automates what the scripts do now.

Educational environment features like annotations and viewpoint manipulation. (either through visible link widgets, or through a more traditional viewpoint selection mechanism like a list tied to pageup & pagedown)

Easier building geometry authoring, maybe from footprint data.

docs & tutorials

0A - Finish ArcADL demo.
0  - Finish ArcADL alpha. (dep: collection w/ access reports to test against)

1 - Get text labels working in VTP
1A - improvements to viewpoint handling in VTP

2 - Improve model with new & better aligned imagery for downtown &
campus

3 - Turn imagery->VTP ingest notes into something like actual
documentation (maybe)

4 - Evaluate new VGIS tools

5 - Gazetteer based labels & ADL search results visualization demo
(results viz in globe view and in terrain view)

6 - "parameter change" demo & model/library integration proposal

7 - Flesh out VirSoCo RFC

8 - Deliver ArcADL to VTP/TOPP, discuss directions

9 - geospatialized hypertext demo of some kind (spider dumps nongeoreferenced
pages into a DL collection, they're initially randomly spatialized).  dependent
on having basic hypertext visualization toolkit in VTP.

10 - full VTP usability pass

11 - e-z ingest from VB, and/or ingest from VTP (drag & drop?)
including obj placement.

12 - (Greg & Gary's idea) terravision-like, tile-based searches.

existing collaborators

possible future collaborators