Download
this paper from http://www.innovativegis.com/basis/,
select ‘Online Papers’ item
GIS in the
<click here> for
a printer-friendly version (.pdf)
Panelists:
·
Peter Batty, Former Chief Technology Officer of Intergraph
Corporation (see http://geothought.blogspot.com/2007/09/gis-in-rockies-panel-on-service.html
for posting of comments)
·
Joseph K. Berry, Keck Scholar in Geosciences,
·
Jack Dangermond, President, Environmental Systems Research Institute
(ESRI)
·
William Gail, PhD, Director, Strategic Development - Microsoft
Virtual Earth
·
Geoff Zeiss, Director of Technology, Infrastructure Solutions
Division, Autodesk, Inc.
·
Andy Zetlan, National Director of Utility Industry Solutions,
Oracle Corporation
(see http://apb.directionsmag.com/archives/3332-SOA-for-GIS.html
for Directions magazine Blog discussion)
Question: The creation of
spatially aware enterprises is at the forefront of GIS application
development. However, most enterprises have disparate legacy applications
and data sets as well as evolving business requirements that require
integration. One solution in the marketplace is the creation of service
oriented architecture to support these needs.
·
Do
you agree that service oriented architecture is the key to enterprise data
integration and interoperability?
·
If
so, how do you see geospatial technology evolving to support the concept of
location services and data interoperability?
·
If
not, what is the alternative?
Each
panelist will have five minutes to answer the above questions and then
there will be a general discussion with members of the audience.
Response by Joseph K.
Berry
Let
me start with a couple of disclaimers …as the token academic on the panel, my
read of the question likely will be a bit askew from the others. Also, my programming skills have atrophied to
a point that programmers tactfully suggest that I shouldn’t get anywhere near
computer code. As such, my thoughts will
avoid a technical critique and focus on the broader scope of the ‘…how GIS will evolve…’ portion of the
question.
Most
definitions of Service-Oriented Architecture boil down to ‘…a developer technology
that supports heterogeneous, loosely-coupled services.’ This means a lot to computer scientists, but
what does it mean to the rest of us, or more importantly, to GIS technology.
In
a sense, Service-Oriented Architecture (allow me to call it ‘SOA’ for short) is
sort of like standardization for data exchange, except it standardizes program
integration. Like the Geography Network that
is designed to grab a piece of data from here and another from there, SOA is
designed to hook together disparate programs. For example, this means that an
application can seamlessly mix stuff from Excel with a routing model and then visualize
the result as a series of charts, tables and maps. The result is a comprehensive solution rather
than a segmented string of the piecemeal parts of a solution.
What
used to be independent stovepipes, or silos of processing, are bundled into
pool of exchanged data and analytics that holistically address a sequence of
steps of a complete application. SOA
promises to provide the processing flexibility needed to fully integrate GIS
with the wealth of other computer-based technologies. No longer will our field be ‘down the hall
and to the right’ in its walled fortress (or small cubicle) but its
visualization and analysis tools readily accessible to other programs.
How
this vision plays out is yet to be determined.
The idea of program integration has been around quite awhile …beginning
with Modular Programming in the 1980s through Object-oriented Programming in
the 90s to the current Web Services applications. The dream of ‘plug ‘n play’ programming has
universal appeal. However, its practical
reality seems to be much different from the vision.
Tim
Bray of Sun Microsystems commenting on the term SOA says that “…I don’t really
like to use it any more. It has, I think
become damaged, weakened by over hype, over use, over promise, under
deliver. I can't explain what the difference
is between SOA and Web Services and I'm not sure I've met anybody who actually
can without going into paragraphs and paragraphs of prose. … If you are confused about what SOA really
means, it's because the world is.”1
More
SOA bashing comes from Gregor Hohpe, co-author of the book Enterprise
Integration Patterns who suggests “…some ‘alternative’ meanings for SOA,
such as: same old architecture, some other architecture, SOAP without the P and
stupid over-hyped acronym.”2
While
such comments imply that the term SOA is overused, does it mean that the
concept is dead on arrival? Maybe the
best way to describe SOA is by comparing it to its alternatives. Its polar opposite is the old stand-alone
system with self-contained, proprietary software that is constrained to a
single hardware environment. Next are Legacy
Systems involving homespun corporate local area network solutions. Then there are Internet Portals, such
as the Department of Interior’s Geospatial
One Stop, that offers multiple views and applications but does not adhere
to the same level of interoperability standards as envisioned for SOA. While there is a myriad of stand-alone,
legacy and portal applications, I am not personally aware of any organization implementing
non-trivial GIS operations in a SOA framework, such as geocoding and simple
buffering.
As
a curiosity, how many in the audience have adopted the SOA framework and
implemented a mapping solution? …or plan
to in the next year? If this was a
general IT meeting I suspect that a lot more hands would be raised (note: no one in the audience indicated that they
have adopted; 3 hands out of 200+ were raised to indicate that they planned to
adopt SOA in the near future). So
what makes GIS a laggard in SOA adoption— the purported wave of computer
science’s future? I suspect three
primary factors are at play— data, demands, and people.
Dave
Bouwman with Sanborn GIS Solutions states that “While this style of
architecture works well for a lot of business applications, GIS is different in
that there is massively more data involved, and transporting this (as binary or
xml) is inefficient. In order to avoid
huge data volumes “on the wire”, you need to build one-stop services that take
simple inputs, perform multi-step complex analyses, and return relatively
simple results. Thus, almost all services are custom, which means there is
relatively little re-use beyond the scope of the original application— which
negates some of the SOA hype. Of course this reuse can be done at the object
level, but that’s not accessible in the “SOA” framework until you wrap a web
service in front of it.”3
Coupling
the multi-gigabyte appetites that clog the network with the huge processing
demands of GIS solutions creates a technical bottleneck that is orders of
magnitude larger than traditional non-spatial applications. But the perfect storm arises when the ‘people-factor’
enters the mix. Historically, there have
been IT folks who have little understanding and even less patience with oddball
spatial data. In opposition there were
GIS folks with a geo-centric focus that fueled a ‘built here’ attitude skirting
the larger computer industry.
The
fact that the SOA question is being discussed at a general GIS meeting, points
to today’s convergent trajectory of these two camps— mainstream GeoScience and
GeoExploration tools. Traditional
computer companies like Google, Microsoft and Yahoo are entering the waters of
geotechnology at the GeoExploration shallow end. Conversely, GIS vendors with deep keels in
GeoScience are capitalizing on computer science advances for improved
performance, interoperability and visualization.
An
important lesson learned by the GeoScience camp is that data has to be
integrated with a solution and not left as an afterthought for users to cobble
together. Another lesson has been that
user interfaces need to be intuitive, uncluttered and consistent across the
industry. Additionally, the abstract 2D
pastel map is giving way to 3D visualization and virtual reality rendering— a
bit of influence from our CAD cousins and the gaming industry.
But
what are the take-aways for traditional computer science vendors? First and foremost is an active awareness of geotechnology,
both in terms of its technical requirements and its business potential. Under the current yardstick of ‘eyeball
contacts,’ the GeoExploration tools have been wildly successful.
While
Service-Oriented Architecture is an important Evolutionary Step in the Expression and Access of
geospatial data, it is not revolutionary. SOA is child of computer science that is
being adapted to geotechnology, not the reverse. It greatly facilitates the development of GIS
solutions by providing an environment that brings benefits, but also poses
challenges.
My
30-plus years of tinkering with GIS has seen its environments evolve from
mainframe computers to mini-computers, to personal computers, to distributed
computing, to web services and now the offshoot of Service-Oriented
Architecture. What is common at each of these steps is that GIS has gained
increasing general awareness and acceptance.
In the 1970s there was just a small cadre of a few thousand folks in the
world that even had a hint of what “digital maps” were and how they might be
used.
Fast-forwarding
to today sees 3D visualization as the current lubricant driving GIS’s
broad acceptance with hundreds of millions of folks viewing ‘a CNN zoom-in’ to a
news hot spot. Society has come to
accept —no demand— digital map
processing in everything we do from getting driving directions, to seeing your
neighbor’s property valuation, to zooming-in on a beachfront resort in Belize. Maps used to be an abstract collection of
lines, symbols and colors sandwiched on a handy, foldable piece of paper. With any luck, in the not too distant future we
will all have a 'decoder ring' that projects a hologram map with real-time fly-through
along the lines of R2D2’s imaging of Princess Leia in Star Wars.
But
at the core, have these technological advancements really changed mapping? …or has
just mapping’s expression and access changed?
Current
Revolutionary Steps in
Analytics and Concepts are underway like the energized paddling
beneath a seemingly serene swan (current GIS expression and access). Recall from your academic days, the
philosopher’s progression…
from Data
(facts),
to Information (facts within a
context),
to Knowledge (interrelationships
among relevant facts), and finally
to Wisdom (actionable knowledge).
Most
GeoExploration applications simply assemble spatial data into a graphic
form. While it might be a
knock-your-socks-off graphic, the distillation of the data to information is
left to visceral viewing and human interpretation and judgment.
For
example, a mash-up of a set of virtual pins representing crimes in a city can
be poked into a Google Earth display.
Interpretation and assessment of the general pattern, however, is left
for the brain to construe. But there is
a multitude of analytics that can be brought into play that translates the spatial
data into information, knowledge and wisdom needed for decision-making. Geo-query can segment by the type of crime;
density analysis can isolate unusually high and low pockets of crime;
coincident statistics can search for correlation with other data layers;
effective distance can determine proximity to key features; spatial data mining
can derive prediction models.
While
the leap from mapping to map analysis might be well known to those in
GeoScience, it represents a bold new frontier to the GeoExploration camp. I am likely reading too much into the panel’s
question, but the mere fact that the SOA discussion focuses on integrating
programs to operate on geospatial data suggests future development of solutions
that stimulate spatial reasoning through 'thinking with maps' (information and
knowledge) rather than just visualizing data— a significant movement beyond
mapping.
In
the longer run, I see the current advances in spatial data expression, access and
analysis being closely followed by a GeoScience lead revolution in concepts on three main
fronts; 1) the nearly 400 year old Cartesian referencing system will be
replaced by a hexagon/polyhedral system (think hexagon design of a soccer
ball cover) that seamlessly addresses a three-dimensional curved globe; 2) our
current reliance on discrete spatial objects (points, lines and polygons) will
be extended to continuous surfaces and volumes; and 3) our current
modeling capabilities focusing on coincident 2D map layers will be extended
to characterizing 3D correlations and flows (particularly important
in defense, environmental, health, resources and other applications
heavily reliant on spatial relationships and interactions).
A
fourth, and even more radical revolutionary front, is the full integration
of space and time forming geo-temporal data structures supporting geotechnology.
But such a crystal-balling diatribe is moving well astray from even the broad ‘…how GIS will evolve…’ portion of the
question.
In
conclusion, I believe we need to keep SOA in perspective as an important
evolutionary step in the expression and access of spatial data that
facilitates revolutionary steps in GeoScience.
My 30+ years in GIS has tempered a quick embrace of any GIS
end-all, silver bullet. However, there appears
to be a very interesting three phase cycle in GIS's evolution— a mapping
focus followed by a data/structure focus and then an analysis
focus.
In
the 1970s, Automated Cartography was the rage (mapping focus). In the 1980s, Spatial Database Management and
geo-query took center stage (data/structure focus). In the 1990s, GIS Modeling was at the
forefront (analysis focus). Today, Multimedia
Mapping and visualization in the form of GeoExploration tools are capturing all
of the attention (mapping focus). I
sense that SOA and a desire to extend and integrate programs, not to mention
the melding of GIS and traditional computer industries, are early nudges in the
rounding of the tanker toward a GeoScience overhaul of its underlying concepts
(data/structure focus) and analytic tools (analysis focus) ...then maybe we can
get around to the holographic decoder ring (mapping focus) by mid-century.
Thank
you.
Note: a .pdf text version of this presentation is posted at www.innovativegis.com/basis, under the ‘Online Papers’ item.
______________________
1 As quoted by Ben
Ellingson, Contributor,
2 As quoted by Rich
Seeley, News Writer,
3 Personal
communication with Dave Bouwman, Sanborn GIS Solutions,