The products and benefits of remote sensing have long been known to the military. Military and law enforcement agencies require and use sophisticated GIS systems, but rarely are they used for multi-functional applications outside their realm. The dual use of these systems, as well as new and replacement systems, has the potential to produce significant economies of scale.
Take for example a GIS-based emergency management system located in California. To be responsive to emergency situations, this type of system would have to create and maintain an up-to-date comprehensive spatial database. One of its many emergency response missions might be to provide remote sensing-based oil spill diffusion model data to assist in control and clean-up efforts. Another might be to provide near real-time image maps for forest fire fighting efforts. Although this kind of system might be viewed as critical, rarely would it be used on a full- time basis. As its secondary mission, if the system was centered out of the University of California, Santa Barbara, faculty and students could use the systems to do research in remote sensing applications. If the system were designed to meet multi-function applications, it could also be used to provide GIS products to commercial companies for their applications.
Lockheed Martin, Management & Data Systems (M&DS) is building GIS and multi-spectral image processing systems for military and dual-use, non-military applications. The systems range from those that are office based to totally self-contained, air transportable versions. The keys to M&DS' success include: the exploitation of integrating commercial-off-the-shelf (COTS) hardware and software, enhancing and customizing the COTS products to meet specific user requirements, and the development of a capability to input an extremely wide variety of both raster and vector data formats. Options for direct downlink imagery receipt and other near real-time, remote sensing inputs and on-board communication are also important system features.
In designing systems for multi-purpose GIS applications, the greatest challenges come from data management, data quality management and validation, and the selection of data types to be applied to any given model. In the real world it seems that the ideal data for any given model is never available. Models must be designed to ensure they are not dependent on only one form of data. If a helicopter landing zone model must be run to support flood evacuation or relief efforts and the perfect data type isn't available, it is critical that alternatives be used to produce the needed results.
This presentation will address designing systems for multi-purpose GIS applications, with specific emphasis on data management and the selection of data types to be applied to any given model.
I. INTRODUCTION In recent years it has become clear that geographic data, including both raster and vector, can play a major role in managing emergency situations generated from natural and man-made disasters. To be efficient, the mitigation of the effects of disasters must be a combination of pre- disaster planning and post-disaster situation management and clean-up. Because emergency management is primarily a part-time job and because state-of-the-art Geographic Information Systems (GIS) can be very expensive, obtaining the resources necessary to build, operate and maintain GIS emergency response systems is a challenge. To meet this challenge with success, design reuse and alternative or supplemental system functions can play a major role. II. RESOURCE ALLOCATION We are clearly in an era when the resources available for national, state, provincial and local governments are being stressed by shrinking budget allocations and increased demands. This same phenomena is also true in commercial enterprises and the academic world. These diverging supply and demand curves can be further complicated by the introduction of new technologies. When a technology is replaced by something new, typically there is some form of one for one change from old to new. The demonstration of how the new technology is better than the old is relatively easy because one can be compared to the other. When there is no replacement of an old technology, a further burden is placed on resources. For those who need or could use the new technology there is also the need to convince resource allocators that it is necessary. The use of spatial/geographic information is a classic example of a new technology which, for the most part, does not replace an older technology. Although there is a growing understanding of its applications, demonstrating this is not always easy. The allocation of resources to functions which are not full- time and not completely understood, tend to be among those deferred or given the lowest implementation priority. The challenge of justifying the need for resources must directly address these issues. If the application technology is not understood, demonstrations of clear, non-technical applications should be used to support resource requests. When a function is not a full- time requirement, alternative use(s) of the resources will be of significant assistance in their justification. Resource allocators react favorably to those programs that are innovative in technology reuse, have multi-functional applications, and are clearly understood.
III. DUAL-USE EMERGENCY MANAGEMENT SYSTEMS
Nobody wants to be placed in an emergency management
situation, but both natural and man-made disasters are a
fact of life. Safety goes a long way in the prevention of
man-made catastrophes, but despite all precautions
occasionally they occur. Knowledge and understanding of
natural disaster prediction matures constantly, yet
hurricanes, typhoons, floods, earthquakes, blizzards,
forest fires, etc., will always be with us. Minimizing
the loss of lives and material damage associated with
disasters is for some a full-time job, but the majority of
the efforts are part-time.
As the technologies associated with spatial/geographic
information systems mature, their application in support
of emergency management is becoming better understood.
Finding and applying dual uses for GIS emergency
management systems (EMS) is not difficult. The following
examples are provided to assist in building the
justification necessary to support the creation of these
kinds of systems. These examples are based on a scenario
where the, EMS needed to support the state/province
requirements is a combination of an office-based and
mobile system.
A. Emergency Management and Academia
Many states and provinces have colleges and universities
that offer studies at both the undergraduate and graduate
level in the fields of geography and remote sensing. The
facilities, input data, and automated data processing
equipment necessary to support these studies are very
expensive. A properly designed multi-workstation, mobile
(tractor trailer with expandable side(s)) GIS EMS with a
large screen display for Emergency Operation Center (EOC)
briefings could be home based at the state university.
With an office-based GIS data server at the headquarters
of an emergency management agency, the mobile system could
be a remote client linked to the server and could be used
as a classroom/laboratory. Part of the curriculum could
include course(s) in emergency management and disaster
situation mitigation and could include the design and
operation of the system. Studies could also include
requirements to develop and refine GIS model applications.
The students and faculty could also share in the
responsibility of updating and validating data. Research
could be done, using the mobile system, on everything from
business students developing commercial applications of
GIS data to geography studies in all aspects of spatial
data, including environmental, population growth and land
use analyses. This system, although small compared to
many classrooms, would provide a near real world
environment.
B. Emergency Management and Economic Development
To be effective, it is critical that a provincial or state
EMS have the most comprehensive set of terrain data
possible for their area of responsibility. As the
government's economic development efforts focus on
building and attracting businesses, GIS data can be used
for many functions. These include: potential location
identification, factoring in support requirements like
transportation needs and material resources location;
environmental impact studies; demographic analysis; and
many others. Just as a mobile system could set up as a
client, so could other government agencies. The data for
one could be used to feed the other and vice versa.
C. Emergency Management and Law Enforcement
Establishing law enforcement agencies as clients linked to
an emergency management spatial data server would provide
the interactive posture critical in disaster planning and
relief operations. It would also provide the law
enforcement agencies the opportunity to use the spatial
data for everything from tracking crime trends by incident
location, aiding in personnel searches and traffic
control, to using future hyper- spectral imagery for drug
crop location identification.
D. Emergency Management and Miscellaneous Other Agencies
An often forgotten element of law enforcement are the men
and women of the fish and game agencies. Linking GPS data
to terrain information could provide significant
management and operational assistance to these people who
cover hundreds of square miles of territory, often alone.
Going back to the concept of setting up the emergency
management operation as the spatial data server, other
potential government agencies could use the data for many
different purposes. Transportation agencies could run
route or mobility models for maximizing the efficiency of
route selection. They could also use the data for future
route analysis and selection. Trash and waste collection
agencies could also maximize the efficiency of collection
routes. Demographics could be tracked by personnel
agencies. The applications of natural resource agencies
are many. Park and recreation divisions could create
special maps and overlays to support the tourism industry.
They could also establish secondary clients off their
client at parks and recreation sites and provide tourists
an on-demand map generation capability. This list goes on
and on with the imagination as the only limitation.
In the United States, there is a very logical linkage
between the state national guard and emergency management
because national guard forces are under the direct control
of the state governor. For those states with a
topographic guard unit, a properly designed terrain
information system could be used for emergency management
and at the same time fulfill its military mission.
When searching out dual-use resource justification, the
boundaries of state or province should not be a barrier.
Establishing joint support agreements for the use of
assets like a mobile EMS makes excellent economic sense.
Although this section has concentrated on using a GIS EMS
as the spatial database server, this should not be taken
to mean that the EMS could not be a client. If the
mission and responsibility of a state or provincial
geographic agency is to develop and maintain the
government's geographic database, then it may be very
logical for the EMS to be the client.
IV. ACQUISITION COST EFFICIENCY
In addition to obtaining cost efficiency through dual use,
minimizing the overall acquisition cost of an EMS is very
important. There are several methods which can be used to
achieve acquisition efficiency. These include, but are by
no means limited to, the use of Commercial-Off-The-Shelf
(COTS) products, software reuse, hardware design reuse and
leveraging similar acquisition efforts.
Lockheed Martin, Management & Data Systems' (M&DS), Fort
Washington Operations (FWO) has built and fielded GIS
systems that exploit these savings principles to the
customer's advantage. The challenges in developing
systems through a Non-Developmental Item (NDI) acquisition
process are many. Rarely, if ever, would one find an
existing EMS that exactly met another EMS user's specific
requirements. Although the vast majority of the design
might be similar, a system design to support hurricane
disaster management will be different from one used in
fighting forest fires. Similarly, most individual COTS
products, including both hardware and software, do not
match one for one, their capabilities versus the user's
requirements. Therefore modification becomes necessary.
To ensure product support integrity is maintained, direct
modifications should be minimized, while extensions and
enhancement are exploited. Unfortunately, many people are
under the misconception that buying the individual pieces
and integrating them together is a simple process. This
could not be further from the truth.
A. COTS Product Integration
The typical GIS systems integrated by Lockheed Martin's
FWO include one or more of the following major hardware
components: UNIX workstations, with peripherals (CD- ROM,
magneto optical drives, 3.5" floppy drives, etc.); large
format ("D" and "E" size) colour printers and scanners;
PCs; external mass storage devices; CD recorders;
Uninterruptible Power Supply (UPS); and external
communications devices. "Fig. 1" shows one hardware
configuration diagram. Optimizing the configuration of
these components to provide operational flexibility and
redundancy is a major task.
Similarly, configuring and optimizing multiple COTS
software application packages to provide the specific user
functions associated with "Fig. 1" is a monumental task.
Table I lists some of the general types of COTS software
used in this form of GIS design.
Table I
TYPICAL GIS COTS SOFTWARE
- GIS vector processing application
- Image processing application
- Graphics application
- Scanner software
- Printer software
- CD ROM reader/writer software
- Multiple communications software
- DBMS
- Word processor
- Spreadsheet
- Security software
- Diagnostic software (multiple)
B. Software Reuse
There are two real challenges associated with integrating
multiple COTS application packages into any system. The
first is to optimize their interoperation, and the second
is to create a human factors environment (Graphical User
Interface {GUI}) that allows an easy understanding of the
overall system use. The integration of these COTS
packages allows customization of the software so that the
use of multiple applications to run a model or create a
product is transparent to the end user.
The creation of software reuse libraries is growing in
both government agencies and commercial companies. There
is also a significant amount of shareware available on the
Internet. As part of an acquisition process, the reuse of
application extensions to COTS and their overall
integration should be serious decision factors. "Fig. 2"
depicts a very complex GIS integration effort that has
significant reuse potential. Although it was developed
for military applications, with minor modification, most
of the GIS models could be applied to emergency management
systems.
C. Hardware Design Reuse Although there are fewer instances for hardware design reuse than those in the area of software, there are still opportunities for savings. A properly designed mobile EMS should consider shock and vibration isolation systems for commercial products. The reuse of isolation designs eliminates the expense of ruggedizing individual COTS components, allowing deployment of EMSs in areas where off-road use is likely. For shelterized systems there are also reuse designs for communications, power and Environmental Control Unit (ECU) interfaces available from military systems. Perhaps the greatest potential savings could come from the reuse of shelter/system designs that have received certification for air transportation. For many applications transportation by aircraft like a C-130 is a highly desired feature. D. Leveraging Similar Developments This element of acquisition cost efficiency is nothing more than combining the three previous factors and researching what others are doing. Although many examples come to mind, one which clearly captures all of the above would be the procurement of an Army High Mobility Multipurpose Wheeled Vehicle (HMMWV) GIS for applications in forest fire command and control. V. DATA MANAGEMENT It may seem to be a contradiction in terms, but building systems with significant flexibility can create management problems. The terrain information systems developed by FWO, for the military, have the capability to import data in over 25 different formats and export them in 12 formats, including Digital Geographic Information Exchange Standard (DIGEST) formats. Tables II and III are import/export capabilities samples. At an initial glance, most people would say this capability provides the systems with extensive flexibility, and it does. However, it creates two challenges, data management and determining which data provides the highest quality terrain analysis model outputs. Table II DATA IMPORT CAPABILITIES - DMA ITD Vector Product Format (VPF) Feature Attribute Coding Catalog (FACC) - DMA Digital Terrain Elevation Data (DTED) Levels I and II - DMA Digital Chart of the World (DCW) - Landsat (BIL, BIP, BSQ, Standard/Fast Formats) - DMA Arc Digitized Raster Graphics (ADRG) - DMA Compressed ADRG (CADRG) - Spot (Pan, XS, Quadmap) - National Imagery Transfer Format (NITF) - Arc Digitized Raster Imagery (ADRI) - Water Resources Data Base (WRDB) - Scanned Maps - DMA Vector Smart Map - ITD Standard Linear Format (SLF) Defense Mapping Agency Feature File (DMAFF) - Digital Feature Attribute Data (DFAD) - Vertical Obstruction Data (VOD) - Digital Line Graph (DLG) - Digital Elevation Model (DEM) - ArcInfo Grid Format - AVHRR Airborne Vehicle High Resolution Radar - Tagged Image File Format (TIFF) - ASCII - PostScript - Computer Graphic Metafile (CGM) - Arc Export Interchange Format - DMA Gazeteer Data Base Table III DATA EXPORT CAPABILITIES - CGM - VPF - NITF - PostScript - SLF - TIFF - ASCII - Grid - Hewlett Packard Graphic Language (HPGL) - HPGL2 - HPRTL - Arc Export Interchange Format Unfortunately, the issues surrounding data management are often the last factors considered in a system design. The use and quality desired/required from products needs to be an up- front design consideration. The data management problems are best exemplified by the fact that different spatial data formats use significantly different thermatic layer schema, and any one feature has the possibility of fitting the criteria of multiple layers. For example: Interim Terrain Data (ITD) has 6 layers; Vector Smart Map (VSM) has 8 layers; Digital Topographic Data (DTOP) has 14 layers; and features such as ditches, dams and wells may be placed in two or more layers. In system design, the quality of the terrain analysis model outputs is critical. When more than one type of data is available for use in running a model, determining the priority of the data to be used versus the output quality can be a major design consideration. For example, if slope is a factor in a model for cross-country mobility of environmental effects, should the model use the slope data in ITD or DTOP if both are available? VI. SUMMARY GIS-based disaster monitoring and mitigation systems are ideal candidates for dual use functionality. They are also the type of systems to which the acquisition cost savings techniques of COTS integration, software reuse, hardware design reuse, and leveraging similar development efforts can be applied. In applying cost savings techniques, agencies developing a GIS system need to remember that integrating multiple hardware and software elements into an efficient system is a very complex process. Additionally, data management should be considered as a critical design matter.