Land Mine Detection: DOD's Research Program Needs a Comprehensive
Evaluation Strategy (17-APR-01, GAO-01-239).			 
								 
Recent U.S. military operations have shown that land mines	 
continue to pose a significant threat to U.S. forces. U.S. land  
mine detection capabilities are limited and largely unchanged	 
since the Second World War. Improving the Department of Defense's
(DOD) land mine detection capability is a challenging		 
technological issues. In this report,GAO reviewed DOD's strategy 
for identifying the most promising land mine detection		 
technologies. GAO found that DOD's ability to make progress in	 
substantially improving its land mine detection capabilities may 
be limited because DOD lacks an effective strategy for		 
identifying and evaluating the most promising technologies. While
DOD maintains an extensive program of outreach to external	 
researchers and other nations' military research organizations,  
it does not use an effective methodology to evaluate all	 
technological options to guide its investment decisions. Although
DOD is investing in several technologies aimed toward developing 
a better solution to the mine detection problem, it is not clear 
that DOD had selected the most promising technologies. Because	 
DOD has not systematically assessed potential land mine detection
technologies against mission needs, GAO conducted its own	 
assessment. GAO found that the technologies DOD is exploring are 
limited in their ability to meet mission needs or are greatly	 
uncertain in their potential. GAO identified other technologies  
that might address DOD's needs, but because they are in immature 
states of development, there is uncertainty about whether they	 
are more promising than the approaches that DOD is exploring.	 
-------------------------Indexing Terms------------------------- 
REPORTNUM:   GAO-01-239 					        
    ACCNO:   A00845						        
  TITLE:     Land Mine Detection: DOD's Research Program Needs a      
             Comprehensive Evaluation Strategy                                
     DATE:   04/17/2001 
  SUBJECT:   Defense capabilities				 
	     Military research and development			 
	     Weapons						 

******************************************************************
** This file contains an ASCII representation of the text of a  **
** GAO Testimony.                                               **
**                                                              **
** No attempt has been made to display graphic images, although **
** figure captions are reproduced.  Tables are included, but    **
** may not resemble those in the printed version.               **
**                                                              **
** Please see the PDF (Portable Document Format) file, when     **
** available, for a complete electronic file of the printed     **
** document's contents.                                         **
**                                                              **
******************************************************************
GAO-01-239

Report to the Chairman, Subcommittee on Military Research and Development,
Committee on Armed Services, House of Representatives

United States General Accounting Office

GAO

April 2001 LAND MINE DETECTION

DOD's Research Program Needs a Comprehensive Evaluation Strategy

GAO- 01- 239

Page i GAO- 01- 239 Land Mine Detection Letter 1

Appendix I Scope And Methodology 21

Appendix II Land Mine Detection Mission Requirements 23

Appendix III Land Mine Detection Technologies 26

Bibliography 35

Tables

Table 1: Ten Candidate Technologies with Known Operational Limitations 14
Table 2: Potentially Promising Technologies Funded by DOD 16 Table 3:
Examples of Environmental Limitations and Uncertainties

Associated with Potentially Promising Technologies 17 Table 4: Criteria for
Target and Operational Parameters 24 Table 5: Criteria for Environmental
Parameters 25

Figures

Figure 1: Candidate Technologies for Land Mine Detection 13 Figure 2:
Technologies for Land Mine Detection that Send

Electromagnetic Energy 27

Abbreviations

DARPA Defense Advanced Research Projects Agency DOD Department of Defense EM
electromagnetic FCS Future Combat Systems GSTAMIDS Ground Stand- off Mine
Detection System HSTAMIDS Handheld Stand- off Mine Detection System IR
infrared LIDAR Light Detection and Ranging MURI Multidisciplinary University
Research Initiative NVESD Night Vision and Electronics Sensors Directorate
Contents

Page ii GAO- 01- 239 Land Mine Detection

SASO Stability and Support Operations SPIE The International Society for
Optical Engineering TNA thermal neutron analysis

Page 1 GAO- 01- 239 Land Mine Detection

April 17, 2001 The Honorable Duncan Hunter Chairman, Subcommittee on
Military Research

and Development Committee on Armed Services House of Representatives

Dear Mr. Chairman: Recent U. S. military operations, such as those in the
Balkans, have shown that land mines continue to pose a significant threat to
U. S. forces. U. S. land mine detection capabilities are limited and largely
unchanged since the Second World War. A U. S. military that now uses million
dollar cruise missiles, tens of million dollar aircraft, and billions of
dollar ships still generally detects land mines with a metal detector and a
probe. The Department of Defense (DOD) has an extensive research program
aimed at developing new detectors to improve its capabilities.

Improving DOD's land mine detection capability is a challenging
technological issue. Because of the threat that land mines pose to U. S.
armed forces, you requested that we assess the abilities of competing
technological options to address DOD's mission needs for land mine
detection. Specifically, our objectives were to determine whether DOD (1)
employs an effective strategy for identifying and evaluating the most
promising land mine detection technologies and (2) is investing in the most
promising technologies to fully address mission needs.

To evaluate DOD's strategy for identifying the most promising land mine
detection technologies, we reviewed regulations, policies, and procedures,
and interviewed DOD officials. To determine if DOD is investing in the most
promising technologies to fully address mission needs, we identified
prospective technological options by reviewing the literature connected with
land mine detection technologies, interviewing researchers from universities
and corporations and other federal agencies, and reviewing proposals that
had been submitted in response to DOD solicitations to fund land mine
detection research. We then developed and applied a systematic framework to
evaluate the prospects of these technological options for meeting
countermine mission needs to determine which were the most promising. (See
app. I for a detailed discussion of our scope and methodology.)

United States General Accounting Office Washington, DC 20548

Page 2 GAO- 01- 239 Land Mine Detection

DOD's ability to make progress in substantially improving its land mine
detection capabilities may be limited because DOD lacks an effective
strategy for identifying and evaluating the most promising technologies.
While DOD maintains an extensive program of outreach to external researchers
and other nations' military research organizations, it does not use an
effective methodology to evaluate all technological options to guide its
investment decisions. More specifically, DOD has not identified all relevant
mission needs to guide its research programs and does not systematically
evaluate the broad range of potential technologies that could address those
mission needs. In addition, its productive program of basic research for
addressing fundamental science- based questions is threatened by a proposed
curtailment of funding. Lastly, because DOD's testing plans do not require
adequate testing of land mine detectors currently in development, the extent
of performance limitations in the various operating conditions under which
they are expected to be used will not be fully understood.

Although DOD is investing in several technologies aimed toward developing a
better solution to the mine detection problem, it is not clear that DOD has
selected the most promising technologies. Because DOD has not systematically
assessed potential land mine detection technologies against mission needs,
we conducted our own assessment. Our evaluation reveals that the
technologies DOD is exploring are limited in their ability to meet mission
needs or are greatly uncertain in their potential capabilities. We
identified other technologies that might address DOD's needs, but because
they are in immature states of development, there is uncertainty about
whether they are more promising than the approaches that DOD is exploring.
Further, because of all these uncertainties, it is also not clear whether
combining two or more technologies will allow DOD to fully meet its mine
detection needs.

To improve the Department's ability to identify and pursue the most
promising technologies for land mine detection, we are recommending that the
Secretary of Defense direct the establishment of a comprehensive research
program to periodically evaluate all applicable land mine detection
technologies against a complete set of mission- based criteria, such as
target signatures, operational requirements and expected environmental
conditions, and provide a sustained level of basic research to sufficiently
address scientific uncertainties. We are also recommending that the
Secretary of Defense require the services to provide adequate test
conditions for systems in development that better reflect the operating
environment in which they will likely have to operate. DOD concurred and
elaborated on its current programs. Results in Brief

Page 3 GAO- 01- 239 Land Mine Detection

Since the advent of modern warfare, the presence of mines and minefields has
hampered the freedom of movement of military forces. The origins of mine
warfare may be traced back to crude explosive devices used during the Civil
War. Since that time, the use of land mines has increased to a point where
there are now over 750 types of land mines, ranging in sophistication from
simple pressure- triggered explosives to more sophisticated devices that use
advanced sensors. It is estimated that there are about 127 million land
mines buried in 55 countries.

Land mines are considered to be a valuable military asset since, by slowing,
channeling, and possibly killing opponents, they multiply the combat impact
of defending forces. Their attractiveness to smaller military and
paramilitary organizations, such as those in the Third World, is further
enhanced because they do not require complex logistics support and are
readily available and inexpensive. Virtually every combatant can make
effective mines, and they will continue to be a viable weapon for the
future.

U. S. forces must be prepared to operate in a mined environment across the
spectrum of military operations, from peacetime activities to large- scale
combat operations. Detection is a key component of countermine efforts. In
combat operations, the countermine mission revolves around speed and
mobility. Mines hinder maneuver commanders' ability to accomplish their
missions because unit commanders need to know where mines are located so
they can avoid or neutralize them. In peacekeeping operations, mines are
used against U. S. forces to slow or stop daily operations. This gives
insurgents a way to control traffic flow of defense forces and affect the
morale of both the military and civilian population.

Since World War II, the U. S. military's primary land mine detection tool
has been the hand- held metal detector used in conjunction with a manual
probe. This method is slow, labor intensive, and dangerous because the
operator is in close proximity to the explosive. The Army has also recently
acquired a small number of vehicle- based metal detectors from South Africa
to be used in route clearing operations and to be issued to units, as
needed, on a contingency basis.

Metal detectors are also sensitive to trace metal elements and debris, which
are found in most soils. This limitation leads to a high level of false
alarms since operators often cannot distinguish between a metal fragment and
a mine. False alarms translate into increased workload and time because each
detection must be treated as if it were an explosive. The wide use of mines
with little to no metal content also presents a significant Background

Page 4 GAO- 01- 239 Land Mine Detection

problem for metal detectors. For example, according to DOD intelligence
reports, about 75 percent of the land mines in Bosnia are low- metallic and
some former Yugoslav mines containing no metal were known to have been
manufactured. In fact, the Army has stated that the inability to effectively
detect low metal and non- metallic mines remains a major operational
deficiency for U. S. forces.

Given the limitations of the metal detector, DOD has been conducting
research and development since World War II to improve its land mine
detection capability. For example, during the 1940s the United States began
research to develop a detector capable of finding nonmetallic mines. Since
then, DOD has embarked on a number of unsuccessful efforts to develop a
nonmetallic detector and to field a vehicle- based land mine detector. DOD
now has new programs to develop a vehicle- based detector and an improved
hand- held detector. DOD expects to field these new systems, both with
nonmetallic capability, within the next 3 years. Airborne detectors are also
being developed by both the Army and the Marine Corps for reconnaissance
missions to locate minefields.

Countermine research and development, which includes land mine detection, is
funded by a number of DOD organizations and coordinated through a newly
established Unexploded Ordnance Center of Excellence. The Army is designated
as the lead agency for DOD's countermine research, with most of its
detection research funding being managed by the Night Vision and Electronic
Sensors Directorate (NVESD) and the Project Manager for Mines, Countermine
and Demolitions. The Marine Corps and the Navy are also supporting a limited
number of land mine detection research efforts. Additionally, the Defense
Advanced Research Projects Agency (DARPA) has been involved with a number of
land mine detection programs throughout the years.

In fiscal years 1998 through 2000, DOD funded over $360 million in
countermine- related research and development projects, of which
approximately $160 million was aimed specifically toward land mine
detection. DOD sponsored an additional $47 million in research during this
period for unexploded ordnance detection (which includes land mines) in
support of other DOD missions such as humanitarian demining and
environmental cleanup. Because of the basic nature of detection, these other
efforts indirectly supported the countermine mission. Overall, DOD funding
levels for countermine research have been sporadic over the years. Major
countermine research initiatives and fieldings of new detectors have
coincided with U. S. military actions, such as the Korean War, the Vietnam
War, Operation Desert Storm, and the recent

Page 5 GAO- 01- 239 Land Mine Detection

peacekeeping operations in the Balkans. Following each influx of countermine
research funding has been a corresponding lull in activity.

A countermine program assessment conducted for the Army in 1993 concluded
that whereas mine developments have benefited from the infusion of leap
ahead technologies, countermine tools have been essentially product improved
counterparts of World War II ideas. However, according to DOD, countermine
development is a slow process because of the technological challenges
inherent to land mine detection. Not only must a detector be able to find
mines quickly and safely through large variety of soils and at varying
depths in battlefield conditions with clutter and even countermeasures, but
it must also be able to discriminate between mines (which vary considerably
in size, shape, and component materials) and other buried objects.

DOD's ability to develop meaningful land mine detection solutions is limited
by the absence of an effective strategy to guide its research and
development program. DOD maintains frequent contact with the external
research community to constantly learn about new detection approaches and
technologies. However, it has not developed a comprehensive set of mission
needs to guide its research programs and does not systematically evaluate
the broad range of potential technologies that could address those mission
needs. In addition, its resources for conducting critical basic research for
addressing fundamental science- based questions are threatened. Lastly,
because DOD's testing plans do not require adequate testing of land mine
detectors in development, the extent of performance limitations in the
variety of operating conditions under which they are expected to be used
will not be fully understood.

DOD has not developed a comprehensive and specific set of mission- based
criteria that reflect the needs of U. S. forces, upon which to base its
investments in new technologies in land mine detection. Although DOD's
overall acquisition process sets out a needs- based framework to conduct
research and development, DOD has not developed a complete statement of
needs at the early stages of research when technologies are first
investigated and selected. The process calls for an evolutionary definition
of needs, meaning that statements of needs start in very general terms and
become increasingly specific as programs mature. Early stages of research
are generated from and guided by general statements of needs supplemented
through collaboration between the combat users and the research communities.
DOD Does Not

Employ An Effective Strategy For Identifying Promising Land Mine Detection
Technologies

DOD Has Not Adequately Specified Mission Needs to Guide Its Research

Page 6 GAO- 01- 239 Land Mine Detection

In the case of land mine detection, the Army stated a general need of having
its forces be able to operate freely in a mined environment. This need has
received a broad definition, as “capabilities for rapid, remote or
standoff surveillance, reconnaissance, detection, and neutralization of
mines.” Further specification of the need is left to representatives
of the user community and researchers to determine. It is only with respect
to specific systems at later stages of the acquisition cycle that more
formalized and specific requirements were established to guide decisions
about further funding.

Although we found that a comprehesive set of specific measurable criteria
representing mission needs had not been developed, we did find some specific
criteria in use to guide research efforts, such as rates of advance and
standoff distances. However, a number of these criteria were established by
DOD to reflect incremental improvements over the current capabilities of
technologies rather than to reflect the optimal needs of combat engineers.
For example, the Army was using performance goals to guide its forward
looking mine detection sensors program. The objective of this program was to
investigate and develop mine detection technologies to increase standoff and
speed for route clearance missions beyond current capabilities. Performance
goals included developing a system with a standoff of greater than 20 meters
with a rate of advance of 20 kilometers per hour. However, these goals were
primarily driven by the capabilities and limitations of the systems being
considered. According to an Army researcher, they were based on what
existing technologies could achieve in a limited time period (3 years) and
not on what the combat engineers would ultimately need. During our
assessment of technologies, which is described in the next section of this
report, we found that the standoff desired by combat engineers was almost 50
meters for route clearance missions with a rate of advance of 40 kilometers
per hour.

One barrier to DOD's developing a comprehensive set of mission needs is
large gaps in information about target signature characteristics and
environmental conditions. For example, significant information gaps exist
about the rate at which land mines leak explosive vapors and the
environmental pathways that the vapors take once they are released. Also,
knowledge gaps about soil characteristics in future battlefields limit DOD's
ability to fully specify mission needs and knowledgeably select among
competing technologies. They also reduce the pace of technological
innovation by hampering researchers from predicting how their devices will
function. DOD is currently funding research to answer several important
questions in these areas. But, as discussed below, continued DOD funding is
threatened.

Page 7 GAO- 01- 239 Land Mine Detection

Just as DOD has failed to adequately specify countermine mission needs for
assessing promising technologies, we found that it had not systematically
assessed the strengths and the limitations of underlying technologies to
meet mission needs. DOD employs a number of mechanisms to obtain ideas for
promising land mine detection solutions. These include attending and
sponsoring technical conferences, arranging informal system demonstrations,
convening workshops, and publishing formal solicitations for research
proposals. However, DOD does not systematically evaluate the merits of the
wide variety of underlying technologies against a comprehensive set of
mission needs to identify the most promising candidates for a focused and
sustained research program. Instead, it generally evaluates the merits of
specific systems proposed by developers against time- driven requirements of
its research programs.

One way DOD identifies land mine detection ideas is through sponsoring and
attending international technical conferences on land mine detection
technologies. For example, it sponsors an annual conference on unexploded
ordnance detection and clearance, at which countermine related detection is
a major focus. Additionally, DOD research officials have chaired mine
detection conferences within annual sensing technology symposia of the
International Society for Optical Engineering (SPIE) since 1995. The most
recent SPIE conference on mine detection, held in April 2000, included over
130 technical presentations by researchers from DOD and other organizations
worldwide. SPIE provides DOD land mine research officials an opportunity to
network with researchers working in different areas of sensing technologies.
DOD also identifies new technologies through reviewing researchers' ideas
outside of the formal solicitation process by occasionally allowing
researchers to demonstrate their ideas at DOD facilities. Technical
workshops are another mechanism used by DOD to identify new ideas. For
example, DOD's Unexploded Ordnance Center of Excellence held a workshop, in
part, to identify new land mine detection technologies in 1998. This
workshop, largely attended by DOD staff and contractors, explored
technological approaches that were not receiving a lot of attention. The
report of the workshop pointed out several potential paths for future
investment for land mine detection.

Of all the mechanisms DOD uses to identify new technologies, issuing
announcements in the Commerce Business Daily is its principal means for
communicating its research needs to the outside research community and
receiving ideas and approaches to improve land mine detection capabilities.
In our interviews with non- government researchers, we found that they use
DOD's announcements as their principal means for DOD Does Not

Systematically Evaluate Technological Options for Land Mine Detection
Solutions

Page 8 GAO- 01- 239 Land Mine Detection

familiarizing themselves about DOD's needs. In connection with our efforts
to identify candidate technologies for land mine detection, we searched
databases, such as the Commerce Business Daily, containing DOD
announcements. We found that the Army placed 20 of the 25 announcements we
identified from 1997 through 2000. NVESD accounted for 17 of the
solicitations.

DOD did not perform a systematic evaluation of all the responses to its
announcements against a common set of mission- based criteria to determine
which were the most promising. Instead, we found that typically responses
that had a reasonable chance of meeting time- constrained program milestones
were more likely to receive funding. This thrust is indicated by the
following statement from a recent report of the DOD Center of Excellence:

Countermine research and development detection funding is concentrated on
four primary technologies There has been increasing emphasis on radar and
active electromagnetics as the technologies showing the greatest short term
promise for the reliable detection of land mines (emphasis added). 1

At NVESD, which has the largest share of countermine detection research,
programs are generally time- limited. As a result, evaluations of proposals
are largely based on the maturity of the idea. An example is the Future
Combat Systems (FCS) Mine Detection and Neutralization program, which is
funded at about $21 million over 3 years. This program is designed to have a
system ready for testing by fiscal year 2002, only 3 years after the program
started. This pace is necessary to meet the Army's overall goals for
fielding FCS. NVESD officials told us that this time constraint means they
are more apt to fund the more mature ideas. This time constraint could
therefore result in not selecting potentially promising technologies that
might involve more risk. Although NVESD officials stated that they are
receptive to less developed ideas that show promise, the requirements of the
program may make this difficult to do.

We found that DOD did not supplement its frequent announcements with
periodic reviews of the underlying technologies that the responses were

1 Joint Unexploded Ordnance Coordination Office, UXO Center of Excellence
Annual Report for 1999, Apr. 2000, p. 17.

Page 9 GAO- 01- 239 Land Mine Detection

based on. 2 Such a review would evaluate their future prospects and could
suggest a long- term sustained research program in a technological area that
required several thrusts, whereas the individual project proposals might
appear to have doubtful value in themselves. Along a similar vein, in 1998 a
Defense Science Board task force that evaluated DOD's efforts in a closely
related area of research and development also recommended a two- track
approach for research and development. 3 The Board found that,

“there has been too little attention given to some techniques which
may provide capabilities important for particular sites” and
recommended that DOD institute a program parallel to the
“baseline” program that “would

involve an aggressive research and development effort to explore some
avenues which have received too little attention in the past.”

Numerous questions about the physics- based capabilities of the various
detection technologies make it difficult, if not impossible, to evaluate
them against mission needs at the present time. Although DOD has invested
funds in basic research to address some of its questions, its efforts are
expected to end after fiscal year 2001. In addition to providing support to
technology evaluations, a sustained basic research program is needed to
support DOD's ongoing efforts to develop better systems.

Independent evaluations, as well as our assessment of candidate land mine
detection technologies, which is presented in the next section of this
report, have revealed many uncertainties about the strengths and limitations
of each of the applicable technologies with respect to addressing
countermine mission needs. In addition, DOD has noted a number of
fundamental science- based questions regarding detection technologies. For
example, 3 years ago the Center of Excellence, through a series of
workshops, identified 81 broad research needs critical to improving
detection capabilities. Examples of research needs included an improved
understanding of the impact of environmental conditions on many of the
technologies examined and better characterization of clutter, which
contributes to the problem of false alarms currently plaguing a

2 The Defense Threat Reduction Agency recently presented a limited
assessment of alternative landmine detection technologies in its recent
review of militarily critical technologies. This assessment reviewed the
strengths and limitations of the principal technologies in terms of broad
militarily relevant needs (e. g. speed and effectiveness).

3 Office of the Under Secretary of Defense for Acquisition and Technology,
Report of the Defense Science Board Task Force on Unexploded Ordnance (UXO)
Clearance, Active Range UXO Clearance, and Explosive Ordnance Disposal
Programs, April 1998. DOD's Productive Program

of Fundamental Research is Threatened

Page 10 GAO- 01- 239 Land Mine Detection

number of technologies. Some of the needs have been addressed since the
workshops. For example, the Center sponsored follow- on workshops and
independent studies of radar and metal detectors to address research
questions specific to these technologies. However, DOD officials told us
that the broad set of needs has not been systematically addressed and that
many questions still remain. Also, over the past 3 years, DOD has invested
about $4 million annually in basic research directed at answering
fundamental science- based questions supporting land mine detection. This
work has been managed by the Army Research Office, with funding provided by
both the Army and DOD through its Multidisciplinary University Research
Initiative. However, this research program is expected to end after fiscal
year 2001.

According to DOD, this basic research has been valuable to its land mine
detection program. For example, the 1999 Center of Excellence annual report
states that the basic research program has improved physics- based modeling
so that it is now possible to examine realistic problems that include soil
interactions with buried targets. 4 The results of this modeling have
yielded insights into limitations of sensor performance in various
environments. The report concludes that this modeling work needs to be
continued and expanded to systematically study soil effects. In fact, the
report recommends continued investment in basic research to increase
understanding of phenomenology associated with detection technologies,
stating that the greatest value of basic research comes from a sustained
effort.

DOD's policy is that systems be tested under those realistic conditions that
most stress them. According to DOD, this testing is to demonstrate that all
technical risk areas have been identified and reduced. However, because of
questions about the physics- based strengths and weaknesses of land mine
detection technologies, there is uncertainty about how well the detectors
currently in development will function in the various environmental
conditions expected in countermine operations. Some of these questions could
be answered through thorough developmental testing. However, DOD's testing
plans do not adequately subject its detectors to the multitude of conditions
necessary to address these performance uncertainties.

4 UXO Center of Excellence Annual Report for 1999 (Joint Unexploded Ordnance
Coordination Office, April 5, 2000). Land Mine Detectors Are

Not Subjected to Adequate Testing to Reduce Uncertainties

Page 11 GAO- 01- 239 Land Mine Detection

We reviewed the Army's testing plans for two land mine detection systems
currently in development to determine whether the test protocols were
designed on a framework of identifying and minimizing technical risks
stemming from the uncertainties detailed above. These are the Handheld
Stand- off Mine Detection System (HSTAMIDS) hand- held detector and the
Ground Stand- off Mine Detection System (GSTAMIDS) vehicle- based detector.
We found that the testing plans were not designed around the breadth of
environmental conditions expected for those systems or around anticipated
limitations and uncertainties. Rather, testing is to be conducted at only a
limited number of locations and under ambient climatic conditions. As such,
knowledge about the performance of these detectors in the variety of soil
types and weather conditions expected in worldwide military operations is
likely to be limited.

For example, the performance of ground penetrating radar, a primary sensor
in both the HSTAMIDS and the GSTAMIDS, is questionable in saturated soils,
such as what might occur after a heavy rain. However, neither the HSTAMIDS
nor GSTAMIDS testing plans specifically call for testing in wet conditions.
The only way this condition would be tested is if there is heavy rain on or
just before the days that testing is to occur. As such, knowledge about the
performance of these detectors in a variety of conditions is likely to be
limited.

Incomplete knowledge of the properties of candidate land mine detection
technologies makes it difficult to assess whether DOD is investing in the
most promising technologies to address countermine detection missions.
Because DOD had not performed a systematic assessment of potentially
applicable technologies against military countermine mission needs, we
performed our own evaluation. Through a broad and systematic review of
technological candidates, we identified nine technologies with potential
applicability, five of which DOD is currently exploring. However,
insufficient information about these nine technologies prevented us from
definitively concluding that any could address any of the missions.
Additionally, because of these uncertainties, we could not conclude whether
a “sensor fusion” approach involving a combination of two or
more of the technologies would yield an adequate solution.

We conducted a broad search for potential technological candidates for
solutions to the countermine problem, and then evaluated the candidates
against a set of mission- based criteria to determine which candidates were
promising for further research. A more detailed description of our
methodology is presented in appendix I. For criteria, we identified It is
Uncertain

Whether DOD is Investing in the Most Promising Technologies For Land Mine
Detection

Page 12 GAO- 01- 239 Land Mine Detection

operational needs for each of five different types of critical countermine
missions: (1) breaching, (2) route clearance, (3) area clearance, (4)
tactical reconnaissance, and (5) reconnaissance supporting stability and
support operations during peacetime. 5 A more detailed description of these
missions is presented in appendix II.

We then developed a set of technical criteria to specifically define
detection requirements for each mission. The criteria we developed were
based on target parameters, operational parameters, and environmental
parameters. Target parameters describe the physical characteristics of land
mines and the methods by which they are emplaced. These include such
characteristics as land mine sizes and shapes, metallic content, explosive
content, burial depths and the length of time mines have been buried.
Operational parameters describe the operational needs of the military as
they relate to countermine operations involving mine detection. These
factors include speed of advance, detection distance from the mine (called
stand- off), and the level of precision in identifying the exact position of
hidden mines. Target and operational parameters can vary among the five
types of missions. Environmental parameters, unlike target and operational
parameters, do not vary based on the type of mission. Rather environmental
parameters are site- specific. They are natural and man- made conditions in
and around the battlefield that affect mine detection. These parameters
cover a wide array of atmospheric, surface, and sub- surface environmental
conditions, such as air temperature, dust or fog obscuration, surface snow,
varying soil types and post- blast explosive residue. A more detailed
description of the criteria used in our evaluation is presented in appendix
II.

Our search yielded 19 technological candidates, which span a wide variety of
different physical principles 6 and are shown in figure 1.

5 Stability and Support Operations (SASO), previously known as operations
other than war, involve the use of military capabilities for any purpose
other than war and include such actions as humanitarian assistance,
peacekeeping operations, and support to counterdrug operations.

6 This list of 19 reflects our pooling of similar approaches. Given the
large number of similar projects in the landmine detection field, we found
it necessary to combine similar technologies. For example, we identified a
single acoustic or seismic approach after evaluating four different
technologies and concluding that one of them most closely addressed the
evaluation criteria.

Page 13 GAO- 01- 239 Land Mine Detection

Figure 1: Candidate Technologies for Land Mine Detection

Note: “EM” is an abbreviation for “electromagnetic”.
Source: GAO analysis.

As shown in figure 1, the majority (15) of the technologies use energy from
the electromagnetic (EM) spectrum, either to detect emissions from the mine
or to project energy at the mine and detect a reflection. The energies used
in these technologies span the entire EM spectrum, from radio waves
(characterized by long wavelengths/ low frequencies) to gamma rays (short
wavelengths/ high frequencies). Of the remaining four technologies not
directly utilizing EM energy, two (biosensors and trace vapor detectors)
operate by using a chemical or biological reaction to detect explosive vapor
that is emitted from mines into the surrounding soil or the air directly
above the ground. Another one is based on sending neutrons toward the
target. The last technology works by sending acoustic or seismic energy
toward a target and receiving an acoustic or seismic

1. EM signatures 2. Infrared 3. Passive millimeter wave 4. Passive microwave

1. Conductivity/ resistivity 2. Electromagnetic induction 3. Electromagnetic
radiography 4. Gamma ray imaging 5. LIDAR 6. Microwave enhanced IR 7.
Quadrupole resonance 8. Radar 9. Terahertz imaging 10. X- ray backscatter
11. X- ray fluorescence

1. Acoustic/ seismic 2. Biosensors 3. Neutron activation analysis 4. Trace
vapor Nineteen Technologies Nineteen Technologies

Technologies that exploit properties of the EM spectrum Technologies that
exploit

properties of the EM spectrum Other Technologies Other Technologies Passive
EM Passive EM Technologies that send EM energy

Technologies that send EM energy

Page 14 GAO- 01- 239 Land Mine Detection

reflection. A more detailed discussion of these 19 technologies is included
in appendix III.

When we evaluated the 19 technologies against the operational parameters, we
found that 10 had one or more physics- based limitations that would prevent
them from achieving any of the five countermine missions by themselves (see
table 1). 7 As can be seen from table 1, standoff and speed are the most
challenging attributes of a detection system that would meet DOD's
countermine mission needs. Nine technologies failed to meet the standoff
criterion, and four failed to meet the speed criterion for any of the five
missions.

Table 1: Ten Candidate Technologies with Known Operational Limitations
Technology Known Target and Operational

Limitations

Conductivity/ resistivity Standoff, speed Metal detectors Standoff Neutron
activation analysis Standoff, speed Gamma ray imaging Standoff, speed X- ray
backscatter Standoff, depth Quadrupole resonance Standoff EM signatures
Standoff, mine types Passive microwave Standoff, speed Trace vapor Standoff
Microwave enhanced infrared Mine types

Source: GAO analysis.

We judged that the remaining nine technologies were “potentially

promising” because we did not conclusively identify any definitive
operational limitations to preclude their use in one or more countermine
missions. 8 For all of these nine technologies, our ability to determine
their operational capabilities was reduced by significant uncertainty as to
their capabilities. Some, such as ground penetrating radar and acoustic
technologies, have been studied for many years. Yet continuing improvements
to the sensors and the critical mathematical equations that interpret the
raw data coming from the sensors made it difficult for us to

7 DOD is currently contracting with researchers in 6 of these 10
technologies. 8 Moreover, seven of the nine technologies had no physics-
based limitation for any of the five countermine missions.

Page 15 GAO- 01- 239 Land Mine Detection

predict the absolute limits of their capabilities. Our inability to draw a
conclusion about these technologies is supported by reports from the
Institute for Defense Analyses and other organizations that have found
similar uncertainty about their prospects. The critical issue for radar is
whether it will ever be capable of doing a good enough job discriminating
between targets and natural clutter to allow an acceptable rate of advance.
The issue of clutter is the fundamental problem for many sensor approaches.

Our uncertainty about three technologies, terahertz imaging, x- ray
fluorescence and electromagnetic radiography was different because their
capabilities were not as well- studied. As a result, there was not enough
information for us to determine whether they could meet mission- based
criteria. In addition, DOD officials told us that they believe that two of
them (terahertz imaging and x- ray fluorescence) have fundamental
limitations that rule them out for countermine missions. They claimed that
terahertz energy is unable to penetrate deep enough through the soil and
that x- ray fluorescence has inadequate standoff. However, we were not able
to resolve these issues.

We believe that the lack of consensus about the capabilities of most of the
nine technologies is due, in part, to a basic lack of knowledge about the
upper limits of their capabilities. The only way to determine whether these
technologies can be employed in a detector that meets countermine mission
needs is through a systematic research program.

DOD is currently investing in five of the nine technologies (see table 2),
and it recently stopped funding a project in one of them (passive millimeter
wave).

Page 16 GAO- 01- 239 Land Mine Detection

Table 2: Potentially Promising Technologies Funded by DOD Technology DOD
Funded

Acoustic/ seismic Yes Biosensors Yes Infrared, multi/ hyperspectral Yes
LIDAR Yes Radar Yes Electromagnetic radiography No Passive millimeter wave
No Terahertz imaging No X- ray fluorescence No

Source: GAO analysis.

In our review of the ability of the nine technologies to operate in
different environmental conditions, we could not, with certainty, identify
absolute limitations on the ability of four to operate in expected
environmental conditions. However, all nine have uncertainties about the
range of environmental conditions in which they can adequately perform. The
most significant uncertainties relate to performance in various surface and
subsurface conditions, such as water saturated soil and differing soil
types. In most cases, these uncertainties have not been adequately studied.
Examples of environmental limitations and uncertainties for the nine
technologies are presented in table 3.

Page 17 GAO- 01- 239 Land Mine Detection

Table 3: Examples of Environmental Limitations and Uncertainties Associated
with Potentially Promising Technologies

Technology Known Limitation Uncertainty

Acoustic/ seismic Surface water Saturated soil Biosensors None Soil types
Electromagnetic radiography None Vegetation Infrared, multi/ hyperspectral
Snow cover Rough surfaces LIDAR Precipitation Post- blast Residue Passive
millimeter wave None Snow Cover Radar Saturated soil Vegetation Terahertz
imaging None Saturated Soil 9

X- ray fluorescence Surface water Soil Types Source: GAO analysis.

The uncertainties about the various detection technologies also prevented us
from determining if the technologies could be combined to meet mission
needs. While most of the 19 technologies cannot meet operational and
environmental mission needs, in theory a combination of different sensors
might solve the countermine problem. This type of arrangement, known as
sensor fusion, combines different approaches to compensate for the
limitations of them individually. Canada and the Army are developing systems
that use some form of sensor fusion. Canada's Defense Research Establishment
in Suffield, Alberta, has produced a multisensor land mine detector that
employs thermal neutron activation (TNA), a type of neutron activation
analysis, as a confirmation detector in a system that also employs a metal
detector, infrared (IR), and ground penetrating radar to scan for mines. The
TNA sensor is used to confirm or reject suspect targets that the three
scanning sensors detect. The Army is developing a detector (HSTAMIDS) that
uses sensor fusion to take advantage of the strengths of both metal detector
and radar approaches. In this configuration, the radar is used to improve
the metal detector's performance with mines that employ small amounts of
metal. However, neither of these systems (Canada's and the Army's) will meet
the countermine mission needs stated previously because their component
sensors are limited. Any detection system utilizing sensor fusion would
somehow need to overcome limitations, such as standoff and speed, in
underlying technologies. As pointed out previously, the capability of the

9 While we could not conclude that there any environmental limitations with
terahertz imaging, Army research officials told us that they believe that
saturated soil is a limitation, rather than an uncertainty for this
technology.

Page 18 GAO- 01- 239 Land Mine Detection

identified technologies to meet mission needs is uncertain. Another
consideration in developing a sensor fusion solution is that it would
require significant advances in signal processing.

It is unclear whether DOD's research investments are in those technologies
that, either individually or in combination, have the greatest chance of
leading to solutions that address the U. S. military's countermine mission
needs given the lack of knowledge about the strengths and the limitations of
the various detection technologies. DOD's strategy of working toward
incrementally improving capabilities over current detectors may result in
improvements over current capabilities. However, without a systematic and
comprehensive evaluation of potential technologies based on a complete set
of mission- based needs, DOD does not know if it has invested its funds
wisely to address the needs of the military.

DOD's testing plans for its land mine detection systems in development do
not provide assurance that these systems will perform adequately under most
expected conditions. Demarcating the acceptable operating conditions of a
system is a critical part of research and development. This is important not
only for determining if developmental systems will meet mission needs but
also for defining the operational limitations so that users can make
informed decisions about their use. Therefore, systems should be tested
under those conditions that most stress them. Given the numerous
environmental and climatic conditions that can be expected to affect the
performance of any land mine detector, a robust program of developmental
testing is essential to fully understand the strengths and limitations in
performance under realistic conditions. Failing to test under a plan
specifically designed around the expected environmental and climatic
conditions of use as well as the anticipated limitations of the technologies
could increase the risk of fielding the system.

To improve the Department's ability to identify and pursue the most
promising technologies for land mine detection, we recommend that the
Secretary of Defense (1) direct the establishment of a long- range research
program to periodically evaluate all applicable land mine detection
technologies against a complete set of mission- based criteria and (2)
provide a sustained level of basic research to sufficiently address
scientific uncertainties. Mission- based criteria could include target
signatures, operational requirements, and expected environmental conditions.
We also recommend that the Secretary of Defense require the services to
Conclusions

Recommendations For Executive Action

Page 19 GAO- 01- 239 Land Mine Detection

provide adequate testing conditions for land mine detection systems in
development that better reflect the operating environment in which they will
likely have to operate.

DOD provided written comments on a draft of this report (see app. IV). DOD
concurred with each of our three recommendations and augmented its
concurrence with additional comments. DOD's comments describe and illustrate
the lack of a focused and systematic approach underlying DOD's research
programs for land mine detectors. It is not clear from DOD's response what,
if any, measures it plans to take to implement our recommendations.

In responding to our first recommendation, DOD states that the Army pursues
a systematic research, development, and acquisition program to address land
mine detection needs. However, we found that its approach lacked elements
critical to the success of this program, such as the use of a comprehensive
set of mission- based criteria and a systematic evaluation of the capability
of competing alternative technologies to address these criteria. In fact,
the Army Science Board study cited by DOD in its comments to us also
recommended that “operational needs and priorities need to be clearly
thought through and quantified.” There is nothing in DOD's comments
that is directed toward bridging these gaps. Therefore, we continue to
believe that the changes that we have recommended are required.

Regarding our second recommendation, DOD describes the benefits provided by
its current basic research program, but does not commit to continuing
funding for basic research for land mine detection after this fiscal year.
As we discuss in this report, we believe it is extremely important for DOD
to continue with a sustained program of basic research to support its land
mine detection program given the extent of the uncertainties surrounding the
various technologies. This point was also made by the Army Science Board
panel.

In response to our third recommendation, DOD states that the testing plans
we reviewed were not detailed enough to allow us to reach our conclusions,
and it describes certain activities that it is engaged in to incorporate
realistic environmental conditions into its testing programs for HSTAMIDS
and GSTAMIDS. However, we believe that the described activities further
illustrate the lack of a systematic strategy to guide testing during product
development. DOD acknowledged the threat to the performance of metal
detectors from soils that are rich in iron oxide and Agency Comments

And Our Evaluation

Page 20 GAO- 01- 239 Land Mine Detection

pointed out that it is seeking to identify a “suitable site to test
the HSTAMIDS system in unique soil environments such as laterite.” We
feel that this is an important step in the development of this system. But
we believe that this step, along with tests in saturated soils and snowy
conditions, should have been taken much earlier, before a large commitment
had been made to this system. Testing programs should also be driven by a
systematic mission- based evaluation framework. Such an approach should
delineate at the earliest stages of development the expected environmental
operating conditions based on mission needs. An analysis should then be made
to identify for testing those conditions that pose substantial challenges or
uncertainties for detector performance. Without such a framework, there is a
risk that uncertainties about the performance of these systems will remain
after they have been fielded and that significant testing will ostensibly be
conducted by users rather than by testers.

We are sending a copy of this report to the Honorable Mitchell E. Daniels,
Jr., Director, Office of Management and Budget; the Honorable Donald H.
Rumsfeld, Secretary of Defense; the Honorable Joseph W. Westphal, Acting
Secretary of the Army; the Honorable Robert B. Pirie, Jr., Acting Secretary
of the Navy; General James L. Jones, Commandant of the Marine Corps; and
other interested congressional committees and parties. We will also make
copies available to others upon request.

Please contact me on (202) 512- 2700 if you or your staff have any questions
concerning this report. Major contributors to this report were Kwai- Cheung
Chan, Dan Engelberg, Cary Russell, and John Oppenheim.

Sincerely yours, Nancy Kingsbury Managing Director, Applied Research and
Methods

Appendix I: Scope And Methodology Page 21 GAO- 01- 239 Land Mine Detection

To determine whether the Department of Defense (DOD) employs an effective
strategy for identifying the most promising land mine detection
technologies, we reviewed literature related to research program design and
met with experts in this area. We interviewed officials from the Army, the
Navy, the Marine Corps and the Defense Advanced Research Projects Agency
(DARPA) responsible for running land mine detection research programs. We
also reviewed DOD policy and doctrine related to this area including the
Defense Technology Area Plan, the Army Science and Technology Master Plan,
and Countermine Modernization Plans.

To determine whether DOD is investing in the most promising technologies to
fully address mission needs, we evaluated the set of potential land mine
detection technologies identified through a systematic search against a set
of criteria derived from mission needs. We first designed a framework for
evaluating potential technologies. This framework assisted in identifying
the most promising technologies and research gaps for further investigation.
Through our discussions with DOD, we found out that such a framework had not
previously been created.

Because our framework was mission directed, we identified a set of critical
countermine missions that involve detecting land mines by systematically
interviewing Army and Marine Corps combat engineers to determine how
countermine activities fit into a variety of combat scenarios and reviewing
Army and Marine Corps doctrine that discuss mine threats to U. S. forces and
corresponding countermine tactics. Next, through a review of documents and
discussions with Army and Marine Corps combat engineers, we identified
technical criteria that define detection requirements for each mission.
Officials representing the two organizations responsible for combat engineer
requirements, the Army Engineer School and the Marine Corps Combat
Development Command, reviewed and agreed with the set of criteria we
developed. The critical missions and the set of criteria we developed are
discussed in appendix II.

We then identified conventional and alternative technologies that could have
value in terms of performing these land mine detection missions. We
distinguished between technologies and systems. “Technologies are
approaches by which principles of physics are exploited to achieve
tasks.” 1

1 Kerner, David, et al. Anti- Personnel Landmine (APL) Detection Technology
Survey and Assessment. Prepared for the Defense Threat Reduction Agency.
DynMeridian. Alexandria, VA. Mar. 1999, p. 21. Appendix I: Scope And
Methodology

Appendix I: Scope And Methodology Page 22 GAO- 01- 239 Land Mine Detection

Systems are implementations of technologies. By developing a methodology
that was based on identifying and characterizing technologies, rather than
systems, we sought to go beyond the strengths and limitations of current
devices and thereby provide information on which to base a future- oriented
research program. We identified candidate technologies in three ways: One
way was to review literature on land mine detection and interview
researchers and other experts in the land mine detection field. 2 Another
way was to interview experts in related fields, such as geophysics and civil
engineering, that involve similar activities (i. e., looking for hidden
subsurface objects). In this, our goal was to find out if those fields use
any tools that have not been explored by DOD. The final way was to review
proposals that had been submitted to DOD in response to recent solicitations
for funding. The technologies we identified are presented in appendix III.

We evaluated each of the identified technologies against the set of mission
criteria to determine which were promising for land mine detection. We
identified “potentially promising” technologies by eliminating
those that have limitations that would preclude their meeting mission goals.
In performing this evaluation, we attended conferences and workshops,
reviewed published and unpublished technical literature, interviewed
developers of land mine detection systems, and contracted with an expert in
the field of land mine detection technologies to review our conclusions. We
also obtained comments from technical experts from the Army. Finally, we
determined which of the “potentially promising” technologies DOD
was exploring by reviewing agency documents and interviewing DOD officials.

We performed our work from November 1999 to February 2001 in accordance with
generally accepted government auditing standards.

2 The primary sources of literature we reviewed are contained in the
Bibliography.

Appendix II: Land Mine Detection Mission Requirements

Page 23 GAO- 01- 239 Land Mine Detection

Using our methodology, we identified land mine detection requirements. The
five critical countermine missions that involve land mine detection are (1)
breaching, (2) route clearance, (3) area clearance, (4) tactical
reconnaissance, and (5) stability and support operations (SASO)
reconnaissance. Breaching is the rapid creation of safe paths through a
minefield to project combat forces to the other side. This mission is
usually conducted while the force is under enemy fire. Route clearance is
the detection and removal of mines along pre- existing roads and trails to
allow for the passage of logistics and support forces. Area clearance is the
detection and removal of mines in a designated area of operations to permit
use by military forces. Tactical reconnaissance is performed to identify
mine threats just prior to and throughout combat operations. SASO
reconnaissance is used to assist in making decisions about where to locate
forces and for planning area clearance operations. A principal difference
between tactical and SASO reconnaissance is the time required for performing
the mission. Because SASO reconnaissance involves peacetime operations, the
speed at which it is conducted is not as critical as that for tactical
reconnaissance.

We developed a set of technical criteria to specifically define detection
requirements for each mission and grouped the criteria into target
parameters, operational parameters, and environmental parameters. Target
parameters describe the physical characteristics of land mines and the way
they are emplaced. Given that there are over 750 types of land mines
available worldwide, the target characteristics vary considerably. The
parameters we identified are presented in table 4.

Operational parameters describe the operational needs of the military as
they relate to countermine operations involving mine detection. Our set of
operational parameters are also presented in table 4. One critical
operational criterion for a mine detector is speed of advance. For time
critical missions, like breaching and route clearance, a detector needs to
function effectively at the military forces' operational speeds. The ability
of a detector to keep up with the required rate of advance is dependent on
two factors: its scanning speed (the time to search a given area for mines)
and its false alarm rate, which is based on the number of times a detector
indicates the presence of a mine where one does not exist. False alarms
reduce the rate of advance because combat forces must stop to confirm
whether an alarm is actually a mine.

Another key operational parameter is standoff, which is the distance a mine
detector (and its operator) can be from a mine and still be able to detect
it. The minimum standoff required is the lethal radius of a mine, Appendix
II: Land Mine Detection Mission

Requirements

Appendix II: Land Mine Detection Mission Requirements

Page 24 GAO- 01- 239 Land Mine Detection

which is about 35 meters (for an antitank sized mine). This distance
requirement increases as speed increases to allow for reaction time once an
alarm is sounded. In cases of minefield reconnaissance performed by airborne
detectors, the standoff required is the minimum altitude necessary to
provide safety for the aircraft from enemy ground fire. One final
operational parameter is the ability of a detector to accurately locate the
position of a buried mine. This is important for reducing the time necessary
to remove or otherwise neutralize the mine and the safety risk associated
with manually probing the ground to find the exact mine position.

Table 4: Criteria for Target and Operational Parameters Parameters Criteria

Target parameters Mine types Round/ flat to long/ thin in shape.

Two inches to 14 inches in diameter/ length. Metal, low- metal, and non-
metal content. Explosive content from 7 grams to 25 kilograms. Variety of
explosive types to include TNT, RDX, Tetryl, PETN and Composition B. Aged as
well as recently buried mines Mines/ minefields in place from hours to
years. Buried as well as surface mines From surface laid down to burial
depths of 6

inches Operational parameters Rate of advance 40 kph or more for breaching
and route

clearance. Speed requirement for other missions varies considerably, but can
be significantly slower than 40 kph. Standoff From 35 meters to almost 50
meters for groundbased detectors, 1,500 meters or more for

airborne detectors. Precision of location of mine Identify position of
buried mines to within 25

centimeters. Other operational considerations Consideration of other
limitations as such as

weight, power requirements, and use of radioactive source.

Source: GAO analysis.

The environmental parameters we identified are presented in table 5. These
are natural and man- made conditions in and around the battlefield that
affect mine detection and are grouped into atmospheric, surface, subsurface,
and other environmental conditions. While the target and operational
parameters can vary among the five mission types, the environmental
parameters are not mission- specific. Rather environmental parameters are
site- specific.

Appendix II: Land Mine Detection Mission Requirements

Page 25 GAO- 01- 239 Land Mine Detection

Table 5: Criteria for Environmental Parameters Condition Criteria

Atmospheric Conditions Air temperatures between –25 degrees F to 120
degrees F. Sustained wind speeds up to 46 miles per hour (gusts up to 61
miles per hour). Obscuration from fog, dust, sand, rain, or snow. Surface
Conditions Ice or snow cover.

Surface water from puddles to rivers and rice paddies. Vegetation from short
grass to broad leafy plants. Rocky or uneven surfaces. Subsurface Conditions
Variety of soil types to include clay, sandy/ loamy, volcanic as

well as man- made conditions such as road substrate. Soil moisture content
from dry/ arid to saturated. Existence of natural underground clutter such
as rocks and roots. Other Conditions Time of day from high sun to complete
darkness.

Presence of man- made clutter such as radio frequency interference and post-
blast explosive residue.

Source: GAO analysis.

Appendix III: Land Mine Detection Technologies

Page 26 GAO- 01- 239 Land Mine Detection

In this appendix, we briefly describe the land mine detection technologies
and projects that we identified through our methodology. We grouped the
individual projects and lines of effort on the basis of their underlying
technological approach. Our grouping resulted in 19 distinct approaches.

These technologies vary in their maturity. Some, such as metal detectors and
radar, have been explored by many researchers for many years. Much less is
known about others such as electromagnetic radiography and microwave
enhanced infrared. Others, such as x- ray fluoresence, have been used in
other applications but have received relatively little attention thus far in
this application.

The technologies use different principles. Fifteen of the 19 technologies
are based on receiving electromagnetic (EM) energy from the target. 1 Eleven
of the 15 EM technologies are based on sending energy (in one case energy in
the form of neutrons) into the ground. The remaining four EM technologies
are “passive electromagnetic”; they are based on receiving
energy that is emitted by the land mine. These four technologies are similar
in principle; their relative strengths and limitations with respect to
addressing countermine missions arise from the different types of energy
that they receive. The final 4 of the 19 technologies are primarily not
electromagnetic. Two capture and analyze the explosive that the mine
releases into the ground or air, one is based on acoustic or seismic energy
reflected off of the target, and one is based on sending neutrons toward the
target.

1 The four that do not operate by receiving electromagnetic energy are
acoustic/ seismic, trace vapor, neutron activation analysis and biosensors.
However, certain implementations of acoustic/ seismic can be designed to
utilize electromagnetic energy. Appendix III: Land Mine Detection

Technologies

Appendix III: Land Mine Detection Technologies

Page 27 GAO- 01- 239 Land Mine Detection

Eleven technologies use electromagnetic energy and operate under three
different approaches (see fig. 2).

Figure 2: Technologies for Land Mine Detection that Send Electromagnetic
Energy

Source: GAO analysis

Four operate by sending EM energy into the ground, reflecting off the mine.

Five operate by sending EM energy into the ground, creating an effect on the
explosive substance. Whereas four of the five act on the explosive within
the mine casing, one relies on detecting released explosive molecules.
Technologies That Are

Based on Sending Electromagnetic Energy

1. Conductivity/ resistivity 2. Metal detectors 1. LIDAR

2. Radar 3. Terahertz imaging 4. X ray backscatter

1. Electromagnetic radiography 2. Gamma ray imaging 3. Microwave enhanced
infrared 4. Quadrupole resonance 5. X- ray fluorescence Those that reflect

energy off the mine Those that reflect energy off the mine Those that detect
an electromagnetic field

Those that detect an electromagnetic field Those that react with the
explosive Those that react with the explosive Technologies that send
electromagnetic energy

Technologies that send electromagnetic energy

Appendix III: Land Mine Detection Technologies

Page 28 GAO- 01- 239 Land Mine Detection

Two operate by detecting differences in the low frequency electromagnetic
field around the mine.

Four of the 11 active EM technologies (radar, terahertz imaging, LIDAR, and
x- ray backscatter) are based on projecting energy into the ground and
reflecting off the land mine. The presence of a mine or other buried object
is detected from differences in the electromagnetic properties of the target
and those of the surrounding ground. The relative strengths and limitations
of these technologies vary with their wavelengths. Managing the trade- off
between depth of penetration and resolution is one of the central research
concerns in this area. The choice of frequency is important; lower
frequencies allow better ground penetration but will suffer from poor
spatial resolution. Radar's relatively long wavelength (it operates in the
microwave part of the electromagnetic spectrum) allows it to penetrate the
ground deeply enough to reach buried mines. This ability, along with the
fact that it can detect plastic mines, has made radar the focus of much
research and development in the United States and in other nations. For
example, DOD has incorporated radar into its hand- held system, Handheld
Stand- off Mine Detection System (HSTAMIDS). However, whether a system based
on radar will meet countermine mission needs remains in dispute. The poor
spatial resolution of radar, which makes it difficult at best to distinguish
between buried mines and other objects of a similar size and shape, is the
largest obstacle. Another issue is its inability to penetrate soils that are
saturated with water.

The other technologies have greater resolution but have a corresponding loss
of depth penetration. Because LIDAR has a shorter wavelength than radar, it
has a limited ability to detect buried mines. X- ray backscatter can provide
detailed images of shallowly buried mines due to the extremely short
wavelength of the x- rays. It operates by detecting the difference in the
atomic number between the ground and the mine target. However, the
applicability of this technology is limited due to the limited penetration
of the x- rays into the ground. In theory, terahertz imaging should have a
similar limitation. However, a researcher studying the feasibility of
creating images of mines in the terahertz part of the spectrum told us that
Technologies That Reflect

Energy Off the Mine

Appendix III: Land Mine Detection Technologies

Page 29 GAO- 01- 239 Land Mine Detection

his system might be able to penetrate more deeply by increasing the power of
the energy. 2

Another general approach involves projecting energy into the ground that
reacts with the molecules of the explosive, which send a signal that is
received by the detector. Because it reacts with the explosive, rather than
the container, the approach has the advantage of more specifically targeting
land mines and being less prone to the clutter problem that hinders other
active electromagnetic approaches. However, technologies that adopt this
approach tend to be more complex and expensive.

We identified five distinct technologies that have been advanced that
utilize this general approach. One of them, quadrupole resonance, is a
relatively mature technology in land mine applications and systems have been
built around it. Less is known about the other four technologies and how to
apply them to detect land mines and what their capabilities are for
addressing countermine missions. These four are electromagnetic radiography,
microwave enhanced infrared, x- ray fluorescence, and gamma ray imaging.
Therefore, our assessments are less complete for these than for the other
more well- studied approaches.

? Quadrupole resonance has been explored for identifying explosives for
several years. Much of the basic research was conducted at the Naval
Research Laboratory. Quadrupole resonance detectors are also being developed
to screen for explosives at airports. In quadrupole resonance, a pulse of
long wavelength energy causes the nitrogen nuclei in the explosives to emit
a pulse of energy that is characteristic of the molecule. For example, the
nitrogen atoms in TNT emit a unique pulse that can be picked up by the
detector. One limitation of quadrupole resonance with respect to countermine
missions is that the detector head must be close to the target. The speed at
which quadrupole resonance can operate is in question. Current systems are
fairly slow. In addition, research questions currently exist in several
areas, including how to overcome interference from other sources of energy
and how to configure a quadrupole resonance detector to detect TNT. Despite
these limitations and questions, DOD is developing systems that use this
technology. The Marine Corps is

2 Terahertz, or 10 -12 is at the long wavelength end of the infrared part of
the spectrum. A concern with compensating in this way, however, is that the
amount of energy reflected off of the surface of the ground also increases
dramatically. Technologies that create a

detectable reaction with the explosive

Appendix III: Land Mine Detection Technologies

Page 30 GAO- 01- 239 Land Mine Detection

developing a hand- held device that uses quadrupole resonance and the Army
is developing a land mine detection vehicle that would use an array of
quadrupole resonance detectors across the front to confirm targets presented
by sensors that use either radar or metal detector.

In conversations with individual systems developers, we identified four
other examples of this land mine detection approach. The first two
technologies are based on scanning the ground with long wavelength
microwaves. This energy excites the explosive molecules that emit a signal
that is detected. The other two technologies using this approach send
shorter wavelength energy toward the target.

? Electromagnetic radiography operates by scanning the ground with long
wavelength microwaves. According to one developer, when it is struck by this
energy; the target radiates back in a particular way; exciting molecules at
atomic levels. The molecules respond with spin effects that produce “a

spectrographic signature of the target substance.” As noted
previously, very little is known at the present time about what the limits
are of this technology in terms of the operational requirements and
environmental conditions for countermine applications.

? Microwave enhanced infrared detection operates by sending long wavelength
microwaves into the ground and then detecting a “unique

thermal signature and infrared spectra of chemical explosives.” One
limitation with this approach is it cannot be used to detect metallic mines
because the microwave energy cannot penetrate metal. In addition, the speed
at which it can operate and the standoff distance are both highly uncertain.

? The third technology illuminates the ground with x- rays that causes a
series of changes in the electron configuration of the target atoms that
results in the release of an x- ray photon (x- ray fluorescence). Unlike the
other technologies in this category x- ray fluorescence detects molecules of
explosive that are emitted from the mine. The amount of fluorescence is
dependent on the target molecule. A critical issue in dispute at the present
time is whether x- ray fluorescence can work at the distances required to
address countermine missions The short wavelength of the x- rays used has a
corresponding high degree of scattering. Several experts we spoke to
expressed reservations about standoff for this technology, although the
system developer claims to have surmounted this limitation.

? The fourth technology is gamma ray imaging. The basis of this technique is
an electron accelerator that produces gamma rays that “interact with
the chemical elements in explosives to generate a unique signature.”
Because of the scattering of the (short wavelength), x- ray and gamma ray
detectors operating on these principles must be in close proximity to the
target.

Appendix III: Land Mine Detection Technologies

Page 31 GAO- 01- 239 Land Mine Detection

According to a developer, the detector must be within one foot of the
target. Another obstacle is that the detector would require an extremely
large source of energy to create the gamma rays.

We identified two technologies that are based on detecting an
electromagnetic field.

? The first is electromagnetic induction. As discussed in the background
section, metal detectors that utilize this approach are the principal means
for detecting land mines at the present time. Metal detectors generate a
magnetic field that reacts with electric and/ or magnetic properties of the
target. This reaction causes the generation of a second magnetic field,
which is received by the detector. 3 The restriction to metallic objects is
a limitation given the increasing development of mines with extremely small
amounts of metal. Increasing the sensitivity of a metal detector to detect
extremely small amounts of metal in these mines leads to its detecting other
objects in the ground. Metal detectors are also limited by the need to be
relatively close to the mine target in order to operate effectively.

? The second technology is conductivity/ resistivity that involves applying
current to the ground using a set of electrodes and measuring the voltage
developed between other electrodes. The voltage measured at the electrodes
would be affected by the objects in the ground, including land mines. The
conductivity technique was originally developed to locate minerals, oil
deposits, and groundwater supplies. The need to place the electrodes in or
on the ground is a concern for land mine detection applications of this
technology.

We identified four technologies that have been proposed which do not
actively illuminate the target, but are based on detecting energy emitted or
reflected by the mine. Three detect the energy naturally released by
objects. They are ostensibly cameras that operate in a very similar fashion
to video cameras, although they view not red, green, and blue frequencies,
but other parts of the spectrum. Land mine detectors that use passive
sensing principles spot either (1) a contrast between the energy emitted or
reflected from the mine and that of the background or (2) the contrast

3 Magnetic devices are another type of passive metal detector that sense
perturbations in the earth's magnetic field caused by the presence of
ferrous objects, such as iron. Because they are employed almost exclusively
to detect magnetic objects, magnetometers are less useful as mine detectors.
Technologies that detect

an electromagnetic field Passive Electromagnetic Technologies

Appendix III: Land Mine Detection Technologies

Page 32 GAO- 01- 239 Land Mine Detection

between the (disturbed) soil immediately surrounding a buried mine and the
top layer of soil. They can be designed to pick up this energy difference in
different wavelength bands. Passive detectors have been designed or proposed
to operate in different parts of the EM spectrum. We identified technologies
that operate using infrared, millimeter wave, and microwave principles.
Infrared, millimeter, and microwave techniques have different strengths and
limitations. The trade- offs between scattering and resolution that exist
with the active backscatter approaches (radar and LIDAR) also exist for
passive EM technologies. For example, the longer wavelengths of microwave
and millimeter waves allow them to penetrate through clouds, smoke, dust,
dry leaves, and a thin layer of dry soil but provide more limited resolution
of targets.

These four technologies are capable of greater standoff than others. Several
nations are developing systems that use IR detection to detect minefields
(tactical reconnaissance). Systems are also being developed to gather
information in several infrared wavelength bands at the same time (“
multi- spectral infrared”). This approach increases the amount of
information available to distinguish mine targets from the background. The
Marine Corps is conducting research in this area.

One of the constraints with infrared detection systems is that the mines'
signature against the background will tend to be reduced at certain times
during the day. To overcome this limitation, researchers funded by DOD's
Multidisciplinary University Research Initiative (MURI) recently
investigated amplifying the infrared signal by heating the ground with
microwave energy. Their early findings suggest that microwave heating
enhances the infrared signature of objects buried under smooth surfaces.
However, much work remains. Given continued funding, they plan to add
increasing complexity to their experimentation by testing with rough
surfaces, random shapes, and different mine and soil characteristics. They
will need to conduct additional research to determine whether the rate of
heating is consistent with the speed required to meet most countermine
missions.

The fourth passive electromagnetic approach is based on detecting the energy
produced by the circuitry of advanced mines that contain sophisticated
fuses. DOD has recently funded work on this approach as part of the MURI
initiative. Apart from the limited applicability of this technology,
questions remain concerning how feasible it is and how easily a detector
operating on these principles might be fooled with a decoy.

Appendix III: Land Mine Detection Technologies

Page 33 GAO- 01- 239 Land Mine Detection

We identified four technologies that are not based on electromagnetic
principles. They are acoustic/ seismic, neutron activation, trace vapor and
biosensors. Sensors that utilize an acoustic/ seismic approach operate by
creating an acoustic or seismic wave in the ground that reflects off the
mine. The energy can be delivered in a number of different ways such as a
loudspeaker, a seismic source coupled with the ground, and a laser striking
the ground over the mine. In addition, there are different ways of receiving
the signal from the target (electromagnetically through a doppler radar or
doppler laser device or acoustically through a microphone). Numerous
questions remain about whether an acoustic/ seismic approach can meet the
operational needs for countermine missions and the environmental factors
that would influence its employment.

Although we identified no certain, absolute limitations to an acoustic/
seismic approach meeting countermine missions, we did identify significant
concerns. Acoustic waves are capable of imaging buried land mines. However,
clutter is a major concern with acoustic approaches. Interference from
rocks, vegetation, and other naturally objects in the environment alter the
waves as they travel in the ground. Additional work needs to be conducted to
assess the limits of an acoustic/ seismic approach for detecting land mines.
An acoustic system is one of the technologies that the Army is currently
exploring for the Ground Stand- off Mine Detection System (GSTAMIDS).

Neutron activation analysis techniques operate on the principle that mine
explosives have a much higher concentration of certain elements like
nitrogen and hydrogen than naturally occurring objects. There are several
neutron- based techniques for detecting these explosive properties in bulk
form. All systems are composed of at least a neutron source –
continuous or pulsed, emitting in bursts – to produce the neutrons
that have to be directed into the ground, and a detector to characterize the
outgoing radiation, usually gamma rays, resulting from the interaction of
the neutrons with the soil and the substances it contains (e. g. the
explosive). Neutron activation analysis cannot be used as a standoff
detector. Our review indicated that neutron activation analysis must operate
directly over the mine target. The limited speed of this technology is
another restriction for most missions. In addition, unanswered questions
about this technology concern the depth of penetration and whether it can be
used to detect smaller anti- personnel mines. Because of these limitations
and questions, neutron activation analysis is currently envisioned as having
a role as a confirmation detector alongside faster sensors on systems that
are remotely piloted. For example, as described above, Canada's military has
developed a vehicle that incorporates thermal neutron activation as a Other
Technologies

Appendix III: Land Mine Detection Technologies

Page 34 GAO- 01- 239 Land Mine Detection

confirmation sensor. The vehicle would need to stop only when one of the
scanning sensors indicated a possible mine target.

The other two technologies are trace vapor and biosensors. Trace vapor
detectors involve sensing molecules of the explosive that emanate from the
buried mine and then analyzing them. There are several different approaches
for capturing and analyzing these molecules. In 1997, DARPA initiated a
research program aimed at detecting land mines via their chemical
signatures, referred to as the “electronic dog's nose” program.
The program was established because DARPA believed that the technologies DOD
was developing (metal detectors, radar and infrared) were limited in that
they were not seeking features unique to land mines and were susceptible to
high false alarm rates from natural and man made clutter. Through this
program, DARPA hoped to change the overall philosophy of mine detection in
DOD by detecting the explosive, a unique feature of land mines. This work
has been transitioned over to the Army. However, the role of trace vapor
detectors in most countermine missions is likely to remain limited due to
the limited standoff that can be achieved. The central feature of the
biosensor technology approach is a living animal. Current examples of
biosensors are dogs, bees, and microbes that detect explosives. Many
research questions remain with these approaches.

Appendix IV: Comments From the Department of Defense

Page 35 GAO- 01- 239 Land Mine Detection

Appendix IV: Comments From the Department of Defense

Appendix IV: Comments From the Department of Defense

Page 36 GAO- 01- 239 Land Mine Detection

Now on p. 18.

Appendix IV: Comments From the Department of Defense

Page 37 GAO- 01- 239 Land Mine Detection

Now on pp. 18- 19. Now on p. 18.

Bibliography Page 38 GAO- 01- 239 Land Mine Detection

Andrews, Anne et al., Research on Ground- Penetrating Radar for Detection of
Mines and Unexploded Ordnance: Current Status and Research Strategy,
Institute for Defense Analyses, 1999.

Bruschini, Claudio and Bertrand Gros. A Survey of Current Sensor Technology
Research for the Detection of Landmines, LAMI- DeTeC. Lausanne, Switzerland,
1997.

Bruschini, Claudio and Bertrand Gros. A Survey of Research on Sensor
Technology for Landmine Detection. The Journal of Humanitarian Demining,
Issue 2.1 (Feb. 1998).

Bruschini, Claudio, Karin De Bruyn, Hichem Sahli, and Jan Cornelis. Study on
the State of the Art in the EU Related to Humanitarian Demining Technology,
Products and Practice. ï¿½cole Polytechnique Fï¿½dï¿½rale de Lausanne and Vrije
Universiteit Brusel. Brussels, Belgium, 1999.

Carruthers, Al, Scoping Study for Humanitarian Demining Technologies,
Medicine Hat, Canada: Canadian Centre for Mine Action Technologies, 1999.

Craib, J. A., Survey of Mine Clearance Technology, Conducted for the United
Nations University and the United Nations Department of Humanitarian
Affairs, 1994.

Evaluation of Unexploded Ordnance Detection and Interrogation Technologies.
Prepared for Panama Canal Treaty Implementation Plan Agency. U. S. Army
Environmental Center and Naval Explosive Ordnance Disposal Technology
Division, 1997.

Garwin, Richard L. and Jo L. Husbands. Progress in Humanitarian Demining:
Technical and Policy Challenges. Prepared for the Xth Annual Amaldi
Conference. Paris, France, 1997.

Groot, J. S. and Y. H. L. Janssen. Remote Land Mine( Field) Detection, An
Overview of Techniques. TNO Defence Research. The Hague, The Netherlands.,
1994.

Gros, Bertrand and Claudio Bruschini. Sensor Technologies for the Detection
of Antipersonnel Mines, A Survey of Current Research and System
Developments. EPFL- LAMI DeTeC. Lausanne, Switzerland., 1996. Bibliography

Bibliography Page 39 GAO- 01- 239 Land Mine Detection

Havlï¿½k, Stefan and Peter Licko. Humanitarian Demining: The Challenge for
Robotic Research. The Journal of Humanitarian Demining, Issue 2.2 (May
1998).

Healey, A. J. and W. T. Webber. Sensors for the Detection of Land- based
Munitions. Naval Postgraduate School. Monterey, CA., 1995

Heberlein, David C., Progress in Metal- Detection Techniques for Detecting
and Identifying Landmines and Unexploded Ordnance, Institute for Defense
Analyses, 2000.

Horowitz, Paul, et al ., New Technological Approaches to Humanitarian
Demining, The MITRE Corporation, 1996.

Hussein, Esam M. A. and Edward J. Waller. Landmine Detection: The Problem
and the Challenge. Laboratory for Threat Material Detection, Department of
Mechnaical Engineering, University of New Brunswick. Fredericton, NB,
Canada. 1999.

Janzon, Bo, International Workshop of Technical Experts on Ordnance Recovery
and Disposal in the Framework of International Demining Operations (report),
National Defence Research Establishment, Stockholm, Sweden, 1994.

Johnson, B. et al., A Research and Development Strategy for Unexploded
Ordnance Sensing, Massachusetts Institute of Technology, 1996.

Kerner, David, et al. Anti- Personnel Landmine (APL) Detection Technology
Survey and Assessment. Prepared for the Defense Threat Reduction Agency.
DynMeridian. Alexandria, VA., 1999.

McFee, John, et al., CRAD Countermine R& D Study – Final Report,
Defense Research Establishment Suffield, 1994.

Mï¿½chler, Ph. Detection Technologies for Anti- Personnel Mines. LAMIDeTeC.
Lausanne, Switzerland, 1995.

Scroggins, Debra M., Technology Assessment for the Detection of Buried
Metallic and Non- metallic Cased Ordnance, Naval Explosive Ordnance Disposal
Technology Center, Indian Head, MD, 1993.

Bibliography Page 40 GAO- 01- 239 Land Mine Detection

Sensor Technology Assessment for Ordnance and Explosive Waste Detection and
Location. Prepared for U. S. Army Corps of Engineers and Army Yuma Proving
Ground. Jet Propulsion Laboratory, California Institute of Technology.
Pasadena, CA. 1995.

Tsipis, Kosta. Report on the Landmine Brainstorming Workshop of August 25-
30, 1996. Program in Science and Technology for International Security,
Massachusetts Institute of Technology. Cambridge, MA., 1996.

(713065)

The first copy of each GAO report is free. Additional copies of reports are
$2 each. A check or money order should be made out to the Superintendent of
Documents. VISA and MasterCard credit cards are also accepted.

Orders for 100 or more copies to be mailed to a single address are
discounted 25 percent.

Orders by mail:

U. S. General Accounting Office P. O. Box 37050 Washington, DC 20013

Orders by visiting:

Room 1100 700 4 th St., NW (corner of 4 th and G Sts. NW) Washington, DC
20013

Orders by phone:

(202) 512- 6000 fax: (202) 512- 6061 TDD (202) 512- 2537

Each day, GAO issues a list of newly available reports and testimony. To
receive facsimile copies of the daily list or any list from the past 30
days, please call (202) 512- 6000 using a touchtone phone. A recorded menu
will provide information on how to obtain these lists.

Orders by Internet

For information on how to access GAO reports on the Internet, send an email
message with “info” in the body to:

Info@ www. gao. gov or visit GAO's World Wide Web home page at: http:// www.
gao. gov

Contact one:

? Web site: http:// www. gao. gov/ fraudnet/ fraudnet. htm

? E- mail: fraudnet@ gao. gov

? 1- 800- 424- 5454 (automated answering system) Ordering Information

To Report Fraud, Waste, and Abuse in Federal Programs
*** End of document ***