1S year.
rmance
ing the
nust be
ber will
: With a
ents on
, to the
ogies to
ed onto
t. What
ct few
itutions
ll - and
up to
out the
at there
ed GIS,
nt and
. Many
ms are
to their
vide a
om for
with a
r (PC).
he cost
oftware
ce and
ow not
Cs and
A desktop GIS was often a scaled-down version of the more
powerful workstation-based systems, but now this is not
necessarily the case. PC based systems often provided all the
functions, or even if the desktop version was an unabridged
one, the amount of data involved was normally reduced.
Desk-top mapping systems are special types of desktop GIS,
which are getting more popular these days because of their
low cost and ease in use. Desk-top mapping systems do not
provide the full capability of data input and spatial analysis,
but are very good learning tools and provide excellent support
for data retrieval and thematic mapping.
A unique feature of desktop GIS is that they are much
cheaper, in the order of about one-quarter of the cost of full-
fledged workstation versions. Therefore the cost of choosing
the wrong system is much reduced, and the user has more
freedom to "evaluate" different software before choosing the
final product. Moreover, some of the vendors provide
demonstration copies either free of charge or at very attractive
prices. It is also easier to borrow a system from the existing
users for "test driving" because these systems are often easier
to use and the total system is more portable. For example,
lap-top versions of them can be carried around very easily.
3. Desktop GIS Requirements
The decision to implement a desktop GIS is not necessarily an
obvious conclusion to a series of judgment issues. As with any
business decision, the migration to or the implementation of
desktop GIS can be an unnecessarily expensive proposition if
a well thought out plan has not been developed. Some of the
issues that should be examined in the development of the plan
should include the following questions:
e What is the purpose of implementing desktop GIS?
* Is the organization prepared to commit the necessary
human, financial and management resources?
e Is the organization stable enough at this time?
e Is there sufficient staff to undergo GIS training?
* Is there a complete inventory of existing hardware,
software, networks and peripherals?
* Has a user needs study been done to determine
requirements?
* Has the issue of system administration, configuration and
security been investigated?
* How is the desktop GIS to be technically supported?
* Has a pilot project been done and evaluated?
* Have technical leaders been identified?
4. Desktop GIS Project Elements
The development cycle of GIS is quite well-known. It starts
with the definition of the internal requirements, the evaluation
of systems that could perform the job, the procurement of the
system, its customization, and its daily maintenance. The
development cycle of desktop GIS is not very different from
the classic model, except that the depth and breadth that one
would pursue in each stage of the cycle would be somewhat
reduced because of the smaller size of the system involved.
307
The lifecycle of a GIS, as with most other computer based
developments, has a number of steps or stages. Most often the
major steps are the following:
* information collection and planning
* system implementation
*hardware procurement
*software procurement
*data procurement
straining
application development
* test and evaluate
* system operation and maintenance
These steps are the most common broad order of
implementing any GIS. Desktop GIS development is often not
complex and so it is unlikely that a project would deviate
much from these steps in the order given.
This paper describes only briefly the task of system
evaluation. This applies to the hardware system, the software
system and the GIS system. Each of these elements should be
evaluated against a predetermined criteria that should be
developed by the project leader. Criteria for system evaluation
may be found in the current literature and will not be
elaborated on here. The remainder of this paper will now
describe each implementation step in more detail. Since most
of the elements required for a desktop GIS may be purchased
off the shelf, the assumption that the project will do the
minimum amount of custom development will be made for the
following sections.
4.1 Information Collection and Planning
One of the first things that anyone contemplating the
development of a GIS project must do is to familiarize
themselves with the technology and the benefits. This
includes assessing the project requirements to see if GIS is the
solution. In most requirements where there is a spatial data
component, it is hard to make a case where GIS would not be
useful. Most often the costs of a desktop GIS are such a small
investment that the cost benefit outweighs any disadvantages.
The project leader, if he is not familiar with the technology
must review the literature and any text books available on the
subject. He should attend several GIS conferences, exhibits
and read the recent trade magazines to come up to speed with
the latest technology and applications. Another thing that he
should do is to take several generic GIS training courses, such
as the ones given by academic institutions or the GIS Division
of Geomatics Canada. These generic courses will give a non-
biased view of the technology that would not be available
from a course given by a GIS vendor.
Some potential users requiring larger GIS configuration would
go to lengths to study the technology during the planning stage
of the development cycle. One example involves a study
lasting years resulting in a specification spanning volumes of
documentation. A desktop GIS buyer would have no need for
this kind of thoroughness because the cost of these studies
would far exceed the cost of the desktop GIS itself. However,
the spirit of acquiring appropriate knowledge about the GIS
technology before jumping into it remains the same.
International Archives of Photogrammetry and Remote Sensing. Vol. XXXI, Part B2. Vienna 1996
