Ada 430098
Ada 430098
Ada 430098
w
Force. Dr. Lew Sloter will serve as the Executive Secretary, and LtCol Tony Yang will
serve as the DSB Secretariat Representative.
The Task Force will be operated in accordance with the provision of P.L. 92-463,
the "Federal Advisory Committee Act", and DoD Directive 5105.4, the "DoD Federal
Advisory Committee Management Program". It is not anticipated that this Task Force
will need to go into any "particular matters" within in the meaning of Section 208 of Title
18, U.S. Code, nor will it cause any member to be placed in the position of acting as a
procurement official.
f. S. Gansler
Annex B.
__________________________________________________Task Force Membership
185
ANNEX B.
TASK FORCE MEMBERSHIP ____________________________
Co-Chairs
Dr. Anita Jones University of Virginia
Mr. Larry Lynn Private Consultant
Executive Secretary
Dr. Lewis Sloter ODUSD(S&T)/Weapons Systems
Military Applications Panel
Gen (R) Mike Carns (Co-Chair),
USAF
Private Consultant
Mr. Frank Kendall (Co-Chair) Private Consultant
Mr. Ed Brady Strategic Perspectives, Inc.
Gen (R) Wesley Clark, USA Stephens Group
Gen (R) Wayne Downing, USA Downing & Associates, Inc.
Dr. Michael Frankel SRI International
Gen (R) Dave Maddox, USA Private Consultant
Mr. Walt Morrow MIT Lincoln Laboratory
RADM (R) Marc Pelaez, USN Newport News Shipbuilding
Dr. Robert Rankine Private Consultant
MG (R) Robert Scales, USA Walden e-Learning Inc.
Gen (R) Larry Welch, USAF Institute for Defense Analyses
Investment Strategy Panel
Dr. Ted Gold (Co-Chair) Institute for Defense Analyses
Dr. Lydia Thomas (Co-Chair) Mitretek Systems
Mr. Denis Bovin Bear, Stearns & Co. Inc.
Dr. Jacques Gansler University of Maryland
Dr. Bill Graham National Security Research
Dr. Barry Horowitz Concept Five Technologies
Dr. Ron Kerber Private Consultant
Mr. George Singley Hicks and Associates, Inc.
MajGen (R) Jasper Welch,
USAF
Jasper Welch Associates
DSB Summer Study on
Defense Science and Technology ___________________________________________
186
Technology Panel
Dr. Regina Dugan (Co-Chair) Dugan Ventures
Dr. Peter Lee (Co-Chair) Carnegie Mellon University
Dr. Ruth David ANSER
Dr. Larry Dubois SRI International
Dr. Ken Gabriel Carnegie Mellon University
Dr. James Heath UCLA
Mr. William Koos MILCOM
Dr. Gregory Kovacs, MD Stanford University
Dr. Patrick Lincoln SRI International
Dr. Stephen Rockwood SAIC
Dr. Michael Roukes California Institute of Technology
Pat Scannon, MD XOMA, Ltd.
Judy Swain, MD Stanford University
Mr. Dick Urban Charles Stark Draper Laboratory
Dr. Haydn Wadley University of Virginia
Mr. Owen Wormser C3i
DSB Representative
LtCol Roger Basl, USAF DSB
Government Advisors
CAPT John Costello, USN Joint Staff J-8
Dr. Paris Genalis Naval Warfare
Ms. Vivian George Night Vision Lab
Dr. Genevieve Haddad AFOSR/NL
CAPT Shaun Jones, USN, MD
Col Mike Perrin, USMC Joint Staff J-8
Ms. Mary Scott National Ground Intelligence Center
Staff
Ms. Barbara Bicksler Strategic Analysis, Inc.
Ms. Julie Evans Strategic Analysis, Inc.
Mr. Bob Piccerillo Strategic Analysis, Inc.
Mr. Brad Smith Strategic Analysis, Inc.
Ms. Stacie Smith Strategic Analysis, Inc.
Mr. Theodore Stump Strategic Analysis, Inc.
Annex C.
____________________________________________Presentations to the Task Force
187
ANNEX C.
PRESENTATIONS TO THE TASK FORCE____________________
Plenary Briefings
March 1, 2001
Mr. John Sullivan CIA Future
Mr. Ken Knight DIA
Dr. Rick Smalley Nanotechnology
March 29-30, 2001
Gen Gregory Martin, USAF
Commander USAFE
Kosovo Air Operations
LtGen Maxwell Bailey, USAF Commander AFSOC
LTG Ron Adams SFOR Commander
ADM Thomas Fargo, USN CINCPACFLT
RADM Albert Konetzni, USN Commander SUBPAC
BrigGen William Shelton, USAF AFSPACECOM
BrigGen Dan Leaf, USAF Recent Commander 31
st
Fighter
Wing, Aviano AB, Italy
LtGen Joe Hurd, USAF Recent Commander 7
th
AF Osan,
Korea
MG Dell Dailey, USA JSOC
MajGen (S) Gary Winterberger,
USAF
NATO AWACS Commander
May 3-4, 2001
Mr. Andy Marshall Discussion
June 28-29, 2001
Hon. Pete Aldridge USD (AT&L)
July 26-27, 2001
Dr. Dean Kamen Innovative Invention and the DoD
August 13-24, 2001
Ed Wagamon Russian Experience in Chechnya
Bob Nelson Concealment, Deception, etc. in
Kosovo
Mr. Luc Barthelet Computer Simulation and the
Sims
DSB Summer Study on
Defense Science and Technology ___________________________________________
188
Military Applications Panel
May 3-4, 2001
Dr. Delores Etter DoD S&T Investment Strategy
LTC Bruce Reider, USA Military Innovations
GEN Wesley Clark, USA (Ret) Discussion
Dr. Allen Adler Robotics and FCS
Dr. Ted Gold Views on Transformation
Dr. Wick Murray Innovation
Mr. Steve Rast and Mr. Tom
Lamp
DARPA UCAV Program
Dr. Douglas Gage Software for Robotics
June 28-29, 2001
ADM Art Cebrowski, USN Network Centric Warfare
Dr. Mike Frankel Telecommunications
RADM R.G. Sprigg, USN Navy Warfare Development
Command
Engine for Transformation
ADM Bill Owens, USN (Ret)
MG Bob Scales, USA (Ret) The Future of Warfare
Dr. George Ullrich Advanced Weapons: Capability
Objectives and Technology
July 26-27, 2001
Mr. Andy Krepenivitch MilitaryTransformation
Implications for Science and
Technology
VADM Mayer, USN JFORSCOM
Technology Panel
May 3-4, 2001
Dr. Greg Papadopoulos IT
Dr. Joanna Aizenerg Bio Tech: Biomaterials
Dr. Roger Breeze Bio Tech: Agricultural Biology
Dr. Gordon Ringold Bio Tech: Pharmacology
Dr. Patrick Walsh Bio Tech: Pharmacology
Prof Allan Hoffman Bio Tech: Biomaterials
Dr. Pat Scannon Bio Tech: Pharmacology
Annex C.
____________________________________________Presentations to the Task Force
189
Technology Panel (continued)
June 7-8, 2001
Dr. Bill Press Leveraging the Long-Term
Defense Technology Investment
Dr. Tom McGill Todays Lethal MOUT
Environment
Dr. Steve Wax Lightweight Concepts for
Personal Protection
Dr. Ephrahim Garcia Exoskeletons for Enhancing
Human Perfomance
Dr. Jordan Pollack Automated Design of Throw
Away Robots
Dr. Michael Cima Ceramic Materials
Dr. Robert Full Robotic Locomotion
Dr. Michael Macedonia What the DoD Needs to know
(Modeling and Simulation)
Ms. J.C. Herz Social Ecology of Technological
Changes
Mr. Jesse Schell Massively Multi-Player Games
Mr. Chris Stapleton Interactive Imagination
Dr. Ken Forbus AI & Cognitive Science for Use in
Military Training
Dr. Stuart Wolf Potential Impact to Military of
Quantum Computing and
Communications
Dr. David DiVencenzo Quantum Computing
Dr. David Awsehalom Optical Coupling Concepts to
Spin Systems
Dr. Eli Yablonovitch Quantum Device Concepts
June 28-29, 2001
Dr. Paul Kozemchak Investments and Innovation in the
DoD from 1945-2000
Dr. Allen Adler DARPA FCS Program
Dr. Rodney Brooks AI & Humanoid Robotics
Dr. Graeme Hirst Natural Language Processing
Dr. Michael Goldblatt DARPA
DSB Summer Study on
Defense Science and Technology ___________________________________________
190
Investment Strategies Panel
May 3-4, 2001
Dr. Delores Etter DoD S&T Investment Strategy
June 7-8, 2001
Dr. William Rue System Architecture Experiences
from the Business Community
Dr. Roger McCarthy and
Mr. John Johnston
Application of Commercial
Technologies to the Land Warrior
Mr. John Neer Commercial and national Security
Investment in Space Science and
Technology
Mr. Dick Paul Boeing Perspective
Dr. Robert Hermann Industry Perspective
Mr. Jack Hammond and
Mr. Dale von Haase
Lockheed Martin Perspective
Mr. Wade Schott GD Perspective
June 28-29, 2001
ADM Art Cebrowski, USN Network Centric Warfare
Dr. Mike Frankel Telecommunications
RADM R.G. Sprigg, USN Navy Warfare Development
Command
Engine for Transformation
ADM Bill Owens, USN (Ret)
MG Bob Scales, USA (Ret) The Future of Warfare
Dr. George Ullrich Advanced Weapons: Capability
Objectives and Technology
July 26-27, 2001
Mr. Tom Perdue Technology Transition
Annex D.
__________________________________________ Military Priorities Questionnaire
191
ANNEX D.
MILITARY PRIORITIES QUESTIONNAIRE
PROVIDED TO COMBATANT COMMANDERS ________________
1. Briefly describe your current appointment and its
operational accountabilities.
2. If you could have your wish/capability right now in your
current operational appointment, what would it be? What
kind of altered outputs would you want to be able to
achieve? In what new ways would you like to be able to
employ it and to what effect?
3. What is your toughest/most demanding employment
challenge right now? What worries you the most? If that
employment challenge could be solved right now, precisely
what capability would you like to remedy this operational
employment deficiency? What are your thoughts as to how
it might be employed (conceptually) and to what effect
(outcome)?
4. Operationally, where are you most vulnerable? What
capability would make a huge difference in mitigating this
vulnerability? Think way out of the boxwhat would
really make a difference?
5. We have a fair degree (and growing) capability in stealth
and a fair degree of precision. How much better does it
need to be? How about electronic warfare and the
requirement for electronic preparation of the battlefield and
enroute support (escort)? Is this an under appreciated
deficiency or is it about righton course? What would
you like to have in a more perfect world of instant response
that you do not now see coming?
6. How about C4ISR? Given the reality of high demand/low
density in certain capability sectors, how could we think of
this problem in a different way such that we werent so
reliant on this set of very expensive scarce assets? How
about space doing the job? How good would it have to be?
What would you as a field commander want C4ISR to be
DSB Summer Study on
Defense Science and Technology ___________________________________________
192
able to do and on what kind of a timeline? What
capabilities would you be comfortable with being
exclusively provided from space? Primarily from space?
Secondarily/back up in space?
7. Talk about sensor-to-shooter, real or virtual. Given that
you would like command-in-the-loop, how would you want
sensor-to-shooter to work? How about accountability,
recall, etc? What leap capability would materially improve
matters?
8. How important is battle damage assessment (BDA) to you?
Does it trouble you that BDA is even necessary, tying up
assets and perturbing planning/strike cycles? Suppose
better, more reliable weapons were available with deadly
accuracy? At what point would you be willing to say: with
that probability of accuracy and in view of that functional
reliability, I dont need BDA; I know its dead.
9. How satisfied are you with intelligence and geo-spatial
(mapping; geodesy; situational awareness) support? How
would you like it to be? Be specific. How happy are you
with the current capability to deliver relevant intelligence to
you, your subordinates/execution cells, and your field
units? Is it at the level needed? Be specific. Are
intelligence arrangements with coalition partners
satisfactory? If not, what would you want to happen to
make you more effective?
10. Re mapping and geodesy support. Are you satisfied? Is
the necessity for mensuration of targets a problem and
inhibitor to rapid targeting/attack/reattack? How good
should it be? How about location/height formation (digital
terrain & elevation data [DTED])? What accuracy would
make a major difference to operational employment
(DTED 1 thru 4)?
11. Please discuss command and control. How do you feel
with respect to operating inside your adversarys decision
cycle? Where are we now? Where would you rather be?
Be specific, both in terms of time and capability. Do you
think centralized command ad control is best or, in the
future, are we better off with a common relevant
operational picture at all levels and at all echelons of joint
command? In such a circumstance would you go into the
mission orders business? How about the dark sidethe
Annex D.
__________________________________________ Military Priorities Questionnaire
193
adversary breaking into the net, either through capture or
electronic theft? On balance, where do you want to go?
12. Do you trust your communications/information data
linksinformation assurance issues? How confident are
you that we have impenetrable systems such that they can
be trusted to execute critical missions on time, on target
(synergy/synchronization issues)? In short, if you could
have your way, what sort of command and control set up
and capability would you want now if you could have it?
13. How about information operations: offensive information
warfare, defensive information warfare, and information
security? Recognizing that no commander seems happy in
this area, tell us of your needs, wants, and fears. How do
you want it differently? What kind of capabilities at the
field levelyour levelwould really make a big
difference in your ability to be dominant and effective?
Think big.
14. One final area. How about administrative concerns:
personnel matters, documentation, records keeping, data
crunching, medical, and other support? Consider in
deployed garrision and in field applications, but particularly
on the remote battlefield. Do you feel that support
technology is moving fast enough? Tell us what
perfection would be in this area. How you want it to be.
15. Bottom line. If you had to pick 3-5-7-10 (pick a number)
technology/capability leap improvements in warfighting,
what would they be? Why did you prioritize them as you
did? What is so importantoutcome wiseand which
ones are needed the soonest and why? On the other hand,
what capabilities do we now have that you are responsible
for and that you think are low contributors to combat
outcomes, could be terminated, and the funds rolled into
building new capabilities on your priority list?
DSB Summer Study on
Defense Science and Technology ___________________________________________
194
________________________________________________________ Annex E. Power
195
ANNEX E.
POWER ____________________________________________
Electrical power for the military falls into three broad categories:
Portable power for applications such as communications,
computers, and the Global Positioning System (<100 watts).
Mobile power for applications such as generators, auxiliary
power units, and tactical operations centers (kilowatts).
Stationary power for applications such as bases and ship
propulsion (megawatts to gigawatts).
In addition there are many specialized and very challenging
applications, for example, individual soldier cooling which requires about
200 watts (W) continuously for extended periods of time. Energy
scavenging, or harvesting, from the environment (using power from the
sun, wind, and tides) is another niche application, for both small (remote
sensors) and large stationary installations.
Advances in large-scale stationary power systems, including the use of
more efficient turbines, fuel cells, and potentially nuclear power, are
currently be undertaken through both government funding (primarily from
the Department of Energy and MITI/Japan) and commercial funding.
Cogeneration, enhancements in reliability (e.g., through the development
of advanced materials, sensors, and controls) and efficiency (e.g., from
operation at higher temperatures or the use of bottoming cycles), and
multi-fuel use are all issues currently being addressed. While funding
from the Department of Defense might accelerate the search for solutions
and demonstrate applicability in specific environments, lack of such
funding is certainly not limiting progress at present. Based on the results
of the recent Defense Science Board Study on Fuel Efficiency, however,
more attention in this area is anticipated.
Similarly, advances in mobile power generation are occurring at a
rapid rate due to rising fuel costs overseas. With the advent of new, highly
efficient diesel engines in the European automobile and truck markets as
well as advanced hybrid power trains developed in Japan and North
America, and the current interest in creating fuel cell systems for
automobiles and homes, much of the work in mobile power systems is
DSB Summer Study on
Defense Science and Technology ___________________________________________
196
already being conducted without significant funding from the DoD. It is
estimated that in 2001 over $1 billion of venture capital funds will flow
into the energy sector. These new systems offer smaller, lighter, and more
efficient operation at lower cost and with less environmental impact than
ever before.
The commercial market is already pursuing issues such as rapid start-
up, acoustic and thermal signatures, operation in extreme environments
(e.g., temperature, humidity, altitude), environmental emissions, cost, and
reliability under adverse conditions due both to government mandate
(particularly in Japan and Europe) and to enhanced consumer interest. In
many cases military specifications for such power systems are no more
stringent than those imposed by the automobile industry. The key for the
DoD is to ensure that such systems will operate on fuels found worldwide
(e.g., logistics fuels) which are generally not as clean or as well refined as
those found in North America, Japan or Western Europe. Thus, funding
for fuel reformers, multi-fuel systems and systems integration may be an
appropriate DoD contribution.
The biggest power challenge facing todays military is in portable
power sources. The impact of this issue is graphically demonstrated in
Figure E-1, where a weight trade-off between batteries and other mission-
critical items is shown. Anecdotal remarks from soldiers indicate that they
sometimes take less food and water on a mission in order to carry more
batteries. Due to the proliferation of battery types and sizes in use (which
number in the thousands across the DoD), resupplying batteries in the field
is even more difficult and probably more expensive than resupplying fuel.
Note that while the cost of fuel at a depot is less than $1.50 per gallon, the
estimated cost on the front line is in excess of $500 per gallon. To date, a
similar analysis has not been performed to access the cost of delivering
batteries to soldiers in battle.
Although substantial advances in battery technology have occurred
over the last few years, improvements in energy density have come only at
the rate of about 5 to 10 percent per year. This rate pales in comparison
to, for example, Moores law. One major advantage of the breakthroughs
in microelectronics is that the power usage per transistor has dropped
dramatically over the last two decades. Unfortunately this drop has been
more than offset by the increase in the number of transistors in any given
device, which has grown orders of magnitude faster. Thus todays
microprocessors can require upwards of 100 W of power. This gap
between the need or demand for the function of the device and the ability
to supply power in a compact form is continuing to grow. Thus as the size
of individual microelectronic components continues to shrink, the weight
________________________________________________________ Annex E. Power
197
and volume of the power source can dominate the system. Figure E-2
shows the component breakdown of a typical cell phone.
Figure E-1. Todays weight trade-off: Adding electronics and batteries is at
the expense of other mission-essential items.
Figure E-2. Batteries are clearly the size and weight drivers for portable
electronic devices as evidenced by this data from a Motorola
Micro TAC Lite cellular telephone.
Despite the large role that todays batteries play in portable electronic
devices, there is minimal commercial or government investment in leap
ahead battery technologies. Current improvements (particularly advances
in lithium and lithium-ion rechargeable batteries) are completely driven by
the commercial sector, with the Japanese and Koreans in the lead. There
64.7%
Electronics and
Enclosure
46.7%
Electronics
and Enclosure
53.3%
Battery
35.3%
Battery
Percent Volume Percent Weight
Laptop Computer
and 2 Day Simply
of Batteries
(Appro*. 12.-4 tbs.)
402 Rounds of
5.56 mmAnmo
(Tl-t30-Rouad
f/bgazin*$)
H H H
EBB
BEE
BEE
4 Day Supply
of Food
(72 14b. fJKES)
Protective
3
Canteen
$
Compass
First Aid Kit
DSB Summer Study on
Defense Science and Technology ___________________________________________
198
are a few new ideas for improved anode, cathode, and electrolyte
materials, but these developments would provide less than a factor of two
increase in performance. The Army is looking for almost an order of
magnitude improvement in performance in the next 20 years (see Table E-
1).
Table E-1. Land Warrior Operational Requirements Document,
August 3, 1999 (Scott Feldman, Natick Soldier Center,
Natick, MA).
The intelligence community is the only part of the government
providing relatively modest funding on more advanced battery systems.
The word system must be stressed here. This community deals with a
very uneven duty cycle in many applications and is therefore considering
hybrid battery-battery or battery-ultracapacitor systems. In battery
chemistry there is generally a trade-off between high energy density and
high power density. For example, todays best rechargeable lithium ion
batteries have energy densities on the order of 120 W-hrs/kg and can
deliver about 100 W of power. In contrast, an ultracapacitor
(electrochemical double layer capacitor) holds only about 20 W-hrs/kg,
but can deliver several kilowatts/kg. Combining a low-rate, high-energy-
density battery with a high-rate ultracapacitor may provide the best
combination of performance features for many applications. For example,
a digital cell phone on standby for an extended period of time requires
very low power, but a lot of energy. In talk mode the power requirements
are high, but the duty cycle is short, so that the energy needs are generally
low.
Clearly, to address the challenge created by high-power electronics,
technologists can work either to drop the power requirements of the
system dramatically, to improve the performance of the power source to
meet new demands, or to do both in combination. Both avenues are
certainly being pursued. The commercial sector is currently the driver for
Units
Baseline
(1998) 2003 2005 2008 2018- 2125
Average Power Watts 23 16 8 4 2
Energy in 12.5 hr W hr 285 205 95 51 26
Mission Duration hrs 12.5 12 48 72 227
Mission Energy W hr 285 197 364 294 455
Mission Weight lbs 5.9 1.6 2 1 1
Energy Density W hrs/kg 106 271 400 646 1000
________________________________________________________ Annex E. Power
199
low-power electronics (though much of the current work can be traced to
earlier programs funded by the DoD). Consider, for example, the size,
weight and run time of a state-of-the-art cell phone (days) or how long the
battery lasts in a new personal data assistant (weeks). The use of low-
power microprocessors, more efficient displays, and advanced power
management tools allows laptop computers to run for up to eight hours.
The DoD recognizes these gains (see Table E-1 above) and will certainly
leverage the efforts of others. These gains will also help to solve the
soldier backpack weight problem. Although significant advances have
been madenext generation SINGARS radios are less than 1/10 the size
[548 in3 vs. 48 in3] and 1/7 the weight [3 lbs. vs. 22 lbs.] of current
models, for examplethere is still substantial room for improvement.
SINGARS radios are still more than 10 times the size and weight of even a
low-cost cell phone.
Some of the most popular current and potential future battery
chemistries are summarized in Table E-2. Most very-high-energy-density
systems are experimental or require an open system (e.g., they use air or
water from the environment at the cathode). The energy density is
deceptive since no (or minimal) cathode materials need be factored into
the weight of the cell. As can be seen in Figure E-3, the cathode
dominates the size and weight of a typical electrochemical cell. Open
systems, however, become heavier as usage increases due to the formation
of product salts at the cathode. For the most part these salts are quite inert
and can be dumped overboard as the system discharges. As the cell
sizes shrink, the current collector and packaging size and weight becomes
a greater fraction of the overall system size. Nevertheless, low-rate metal-
air systems are practical for hearing aid batteries. Zinc-air systems have
recently been introduced into the cell phone market with limited success.
Most open systems are far more experimental than their sealed
counterparts, and energy densities of practical systems are hard to estimate
since once a battery is started, it cannot be stopped. Note that sealed
systems are particularly important for underwater and space operations.
An interesting observation from Table E-2 is that the practical energy
density of a cell is only about one-quarter of that of the maximum
theoretical energy density. This is due to packaging, binders (inactive
materials), current collectors, and unknown factors. (See Figure E-3.)
The relative contributions of each of these factors varies dramatically
depending on the size, power requirements, and geometry of the cell.
DSB Summer Study on
Defense Science and Technology ___________________________________________
200
Table E-2. Energy densities of some common and experimental
battery systems.
Couple
Common Name
Voltage
(V)
Maximum
Specific
Energy
(Wh/kg)
Practical
1
Specific
Energy
(Wh/kg)
Practical
1
Energy
Density
(Wh/L)
Sealed Systems
Cd/HgO MerCad E
o
= 0.922 142 65 300
Pb/PbO2 Lead-Acid Voc = 2.18 182 34
2
85
2
Cd/NiO(OH) NiCad, Nickel-Cadmium,
Ni-Cd
Voc = 1.30 210 34
2
100
2
MH/NiO(OH) Nickel/Metal-Hydride, Ni-
MH
Voc = 1.35 220 54
2,7
180
2,7
Zn/HgO Mercury-Zinc Voc = 1.35 242 100 360
Zn/alk/MnO2 Alkaline Manganese Voc = 1.55 302 140 365
Zn/AgO Silver-Zinc Voc = 1.82 &
1.56
435 90
2
170
2
220 850
Li/I2 Lithium/Iodine E
o
= 2.78 555 260
3
940
3
LiC6/Li4CoO2 Lithium Ion Voc = 4.00 455 (630) 125
2,7
300
2,7
Li/MnO2
Lithium/Manganese
Dioxide
Voc = 3.15 900 280 600
Li/SO2 Lithium/Sulfur Dioxide Voc = 2.95 1110 300 500
260
4
350
4
Li/FeS2 Lithium/Iron Disulfide Voc = 1.80 &
1.60
1230 260
8
480
8
Li/SOCl2 Lithium/Thionyl Chloride E
o
= 3.70 1490 510 1020
660
5
1200
5
Li/CF Lithium/Carbon
Monofluoride
Voc = 3.00 2120 290
9
510
9
600
6
1000
6
820
10
1180
10
EH = 4.5 3200
Li/Cl2 Lithium/Chlorine E
o
= 3.98 2520
Li/S Lithium/"Sulfur" Voc = 2.20 3680
Li/SRIF
11
E
o
= 5.8 4750
Open Systems
12
Mg/Air E
o
=3.1 6800
Al/Air E
o
= 2.7 8130 300 330
Zn/Air (cell phone) E
o
=3.6 150
Li/Water E
o
= 3.1 8530
Li/Air E
o
= 3.4 13,000
Voc is the measured open circuit voltage
1 Based on "D" volume cells unless otherwise noted
2 Rechargeable cells
3 Heartpacer cells, 37oC
4 Army version (higher rate capability)
5 10,000 Ah, low rate cells (obsolete)
________________________________________________________ Annex E. Power
201
6 GJM 3/88 "DD" cells
7 "4/3A" cells
8 "AA" cells
9 Commercial "C" cells
10 GJM 1,200 Ah (low rate design)
11 SRI proprietary cathode material, all values are estimates
12 Maximum specific energies are based only on the weight of the anode material
Figure E-3. Weight breakdown of a typical sealed cell.
While todays small-scale systems use energy storage devices, larger-
scale systems convert hydrocarbon fuels (or, in very large systems, nuclear
fuel) to electrical energy. At present both the Department of Defense and
commercial industry are pushing to develop compact energy conversion
devices to run radios, cell phones, laptop computers, and the like. There
are a number of issues that must be addressed, however. The most
important consideration is the details of the missionshort vs. long,
continuous vs. intermittent duty cycle, maximum and minimum power
requirements, etc. As shown in Figure E-4, for short missions, batteries or
extremely small energy conversion devices (e.g., microturbines) may be
the best option to provide power, but for longer missions the size, weight,
and cost of the power source are dominated by the respective size, weight,
and cost of the fuel. In this case the energy density of the fuel and the
efficiency of energy conversion are key parameters.
P
o
u
c
h C
e
ll
Packaging
Packaging
Current Collector
Current Collector
Electrolyte
Cathode
Anode
Weight Breakdown of a Conventional Lithium Battery
current collector 2.7 g
anode 0.5
electrolyte 3.0
cathode 10.3
current collector 2.7
TOTAL 19.2 g (plus packaging)
DSB Summer Study on
Defense Science and Technology ___________________________________________
202
Figure E-4. The trade-off between energy conversion devices and energy
storage devices.
Government funding in the small power source area is really focused
on energy conversion devices (fuel cells, small engines, etc.) with very
little funding going into really advanced batteries (i.e., heavily
fluorinated systems, metal-air and metal-water chemistries, see Table E-
2). Primary sources of funds include the DARPA Palm Power Program
and NIST ATP. There is also growing interest in this area from the
Department of Energy. The driving force here is the very-high-energy-
density of many fuels.
For example, methanol has an energy density of about 5600 W-hrs/kg,
nearly 50 times that of a rechargeable lithium ion battery. Butane is more
than twice as energy-dense as methanol, logistics fuel (e.g., diesel or jet
fuel) is 2.5 times denser than methanol (about 13,000 W-hrs/kg), and the
energy density of hydrogen is a whopping 30,000 W-hrs/kg. On a per
weight basis, even 1 percent energy conversion of hydrogen leads to a
system substantially better than the best batteries in use today.
Unfortunately even liquid hydrogen is not very dense (about 71
g/liter), which means that the practical energy density of hydrogen is only
~2100 W-hrs/liter. Hydrogen can be stored at a density that approaches
that of liquid hydrogen and is on the order of a few percent (see Table E-
3). There is work currently being conducted on the use of chemical
hydrides such as MgH2, LiAlH4, alkyl silanes and amino boranes to
generate hydrogen in situ, but issues of control, generation rates, and
safety (e.g., chemical stability, thermal runaway) have yet to be solved
Mission Duration
(Energy Use Requirements)
Sy
ste
m
M
as
s
(or
Vo
lu
m
e
or
Co
st)
Fuel
Stack
Batteries - quantized
Fuel Cell, TPV,
AMTEC,
Engine, etc.
Slope = weight of fuel x fuel
energy density x energy
conversion efficiency
Intercept = weight
or size of stack,
motor, fuel tank,
controls, etc.
________________________________________________________ Annex E. Power
203
completely. [Note: experimental evidence indicating >>10% by weight
hydrogen storage on graphite nanofibers have never been reproduced.]
Throw in the size, weight and cost of a fuel cell (or other energy
conversion device), and the efficiency of energy conversion (~10 - 50%)
and one can see why people aren't yet jumping all over even hydrogen-
powered systems for portable devices. Restrictions are eased somewhat as
the size grows due to the approximate fixed size of the control electronics
and some of the packaging as well as the ease of storing high pressure
gaseous hydrogen in larger tanks.
As shown qualitatively in Figure E-4 and quantitatively in Table E-3,
fueled systems make the most sense for the longest missions where the
weight and size of the energy conversion device and the fuel tank become
negligible compared to the weight of the fuel, the energy density of the
fuel and the efficiency of energy conversion. Fueled systems also make
sense where issues such as acoustic and thermal signature, start-up time,
and air independent operation are not key concerns. For example, internal
combustion engines or microturbines, which tend to be small, lightweight,
and provide high power, but which are inefficient, noisy, and hot may
make sense to power micro-air vehicles.
One area that has not received much attention from the DoD is small
(portable) nuclear sources. While such systems have been used in space
exploration for years and have racked up an impressive reliability record
(over 250,000 hours of operation without a single failure), terrestrial
applications (other than niche applications such as powering pacemakers)
have not been forthcoming. Clearly public perception is a big issue,
although unbeknown to most, small radioactive sources are present in
smoke detectors, emergency exit signs, and other common items.
Although the energy density is quite high due to the long lifetime of the
source, most radiation-based systems tend to have quite low power
densities. (Shielding is typically very heavy.) To date no serious research
effort has been directed at how to build and optimize these devices.
Expanded funding by the DoD could make a significant impact in this
area, particularly in the development of sources better matched to energy
conversion materials (typically thermoelectrics) and in packaging and
shielding.
DSB Summer Study on
Defense Science and Technology ___________________________________________
204
Table E-3. Approximate energy density of small energy conversion
devices and fuels.
Energy
(kWh)
Energy
Density
(Wh/kg)
Power
Density
(W/kg)
Current/Near Term DMFC, 60W 1.44 215.4 9.23
4.9 553.8 6.59
3-5 yr Advanced DMFC, 20W 1.6 1558 19.05
5.06 2300 9.1
Theoretical Energy Density of MeOH 5600
Current/Near Term PEM Fuel Cell 1.6 235 14.67
5000 psi compressed H2, 100W 4.8 377 7.85
10 447 4.47
1-2 yr H2/PEM, 100W (10-12% by weight H2 storage) 1.6 256 16.08
4.8 439 9.13
10 537 5.37
Advanced (3-5 yrs) Solid Oxide Fuel Cell, JP-8, 20W 1.44 1506 20.9
4.87 3200 13.2
10 4235 8.4
Thermoelectrics (BiTe TEG operating at 4.5% efficient heat to
electricity)
16
Thermal Photovoltaics (TPV, 20W at 2% efficient fuel to electricity,
butane fuel)
Alkali Metal Thermal Energy Conversion (AMTEC)
Current, butane fuel (~10% efficient fuel to electricity) 1.46 502 6.9
5.2 760 2.9
10.4 842 1.6
Future, butane fuel (~15% efficient fuel to electricity) 1.4 598 8.5
5.2 1032 4.0
10.3 1195 2.3
Advanced (3-5 yrs) ICE/MicroStirling/Turbine, 20W 1.44 2900 40.3
4.8 3532 14.75
10 3711 7.4
MIT microturbine project (goal: 5% efficient conversoin of butane to
electricity)
600
Theoretical Energy Density of Diesel Fuel ~13,000
Nuclear
Current pacemaker battery (PuO2/BiTe), including source and
shielding
0.02
Space power (SiGe TEG, 0.07% efficient), including shielding 5.3
90
Sr + Icosohedral Boride, 1% efficient energy conversion ~100,000
28 year half-life, no shielding required for small systems
________________________________________________________ Annex E. Power
205
Finally, no discussion of power sources for the military would be
complete without at least a brief mention of energy scavenging techniques
(especially solar and thermal). While these technologies may not provide
the needed energy or power by themselves, when coupled with batteries
and/or ultracapacitors they may provide a good solution for systems
requiring long duration and very low or intermittent power (see Figure E-
5). There is certainly a lot of work being pursued in this area, most of it
outside of the DoD (at, for example, NREL and MITI/Japan).
Figure E-5. Energy generation from an array of solar cells or a power boot.
Energy scavenging is typically a low rate process requiring a
battery or ultracapacitor to store or convert the energy to a useful
form.
Solar cells are particularly useful for low-power applications (one can
already buy solar-power calculators and watches or power boosters for
cell phones, for example). For anything that requires serious power, size
becomes an issue. For example, a small laptop computer requires 30 W of
power. At 23% solar to electrical energy conversion efficiency (typical of
very high-end single crystal silicon), a panel of 15" x 15" exposed to direct
sunlight at noon is required to be effective. (About 1 kW per m
2
of
sunlight is about the maximum that hits the earth.) At other times of the
day (or with clouds, at night, indoors, etc.) a much larger array would be
needed. Having said that, using a solar panel to keep batteries topped off
will certainly extend the life of a system. Optimizing the mix of batteries
for continuous duty operation with ultracapacitors for intermittent power
needs (e.g., starting the hard drive of a laptop computer) will also improve
Low rale
&iergy
Use
F: ,-,er
Time
DSB Summer Study on
Defense Science and Technology ___________________________________________
206
system life. Other alternatives include wind-up power (commercially
available for radios and flashlights), the power boot being developed by
DARPA (power generation through walking, see Figure E-5) and
thermoelectrics (TEGs) for powering distributed sensors (DARPA/Office
of Naval Research) funding.
As one looks further out into the future, there are many wildcard
technologies that may or may not be practical sources of energy for the
military. These include molecular motors, biofuel cells, solid-state
engines, cold fusion, microemitter arrays, sonofusion, and many others.
In addition, there are potential breakthroughs in power distribution
technologies such as lightweight, low-cost plastic conductors, room-
temperature superconductors, and energy beaming. The DoD can either
take a leadership role in these areas and make a substantial impact on our
future, or just track developments as they unfold with an eye toward
capitalizing on possible breakthroughs.
It is very clear from the above that there is no one single solution to all
of the power problems facing the U.S. military (other than dropping the
power needs of specific devices, which is already happening). Further,
despite best efforts, power requirements in the military are likely to grow
in the future (e.g., due the need for increased soldier microclimate cooling
in response the growing threat of chemical and biological weapons or the
development of advanced electric battlefield vehicles and the use of
directed energy weapons). Of course the military has certain requirements
that most civilians do not need to consider (e.g., extremes of temperatures
(-20 to -60
o
C) and operation in space or under water. These requirements
almost certainly mean that some sort of energy storage device will be
required, independent of whether one uses a fuel cell, small generator, or
solar cell.
Annex F. Unmanned
_______________________________________________Ground Vehicle Technology
207
ANNEX F.
BACKGROUND AND CURRENT ASSESSMENT
OF UNMANNED GROUND VEHICLE TECHNOLOGY___________
Unmanned aerial platforms are a visible mainstay in current U.S.
operations. They are likely to remain so for the foreseeable future. As a
result, the technology and capabilities of such platforms are widely
understood throughout the defense establishment. Unmanned underwater
platforms are less so, but also less likely to play a defining role in overt
military operations. This said, there is great interest in the active use of
unmanned ground vehicles, which presents some important technological
challenges. These challenges and some of the DoDs programs to address
these challenges are described below.
Unmanned Ground Vehicles
Research and development related to mobile robots has been
conducted over the past 40 years. Currently, approximately $60 to $80
million of R&D is being spent annually by the DoD in this area. To date
few mobile robots have been added to the military inventory. Those that
have are primarily used for performing dangerous missions such as
explosive ordnance disposal or mine field clearing. These robots are
characterized by their relatively large size and their teleoperated brute
force approach to mission success. In the past couple of decades, as the
result of increased computation and sensing capabilities enabled by
microelectronics and MEMS technology, emphasis has focused upon the
development of robots for surveillance and reconnaissance missions.
Much smaller size, with on-board sensing, communications, and
processing characterize these robots.
Large Robots. The United States Marine Corps in coordination with
the Unmanned Ground Vehicle Joint Program Office, has developed and
transitioned to acquisition a Standardized Robotic System that is basically
a kit that allows currently fielded vehicles to be retrofitted for
teleoperation. Vehicles such as a D7G bulldozer, an M1 main battle tank
chassis, a HMMWV and an M-60 tank chassis have been teleoperated to
conduct breaching, route clearing, and area clearing missions. The Office
of Special Technology has developed a mini-flail system to clear anti-
personnel land mines. As a result of this effort the Engineering School
DSB Summer Study on
Defense Science and Technology ___________________________________________
208
developed a requirement for a 300 meter line-of-sight vehicle able to
neutralize anti-personnel landmines, perform wire breaching, dispense
obscurants, emplace demolitions, sweep runways, and create access lanes
through buildings or other antipersonnel obstacles. It is scheduled to reach
production in fiscal year 2005. Other large robots include the All-Purpose
Remote Transport System capable of remote operations in various mission
profiles, including clearing unexploded ordnance, and the Automated
Range Ordnance Mapping System, to retrieve buried ordnance and other
hazardous items.
Medium-Sized Robots. The DARPA Tactical Mobile Robots program
has invested approximately $50 million over the past five years and has
developed a variety of new robotics technologies and platforms designed
for dismounted operations by projecting operational influence and
situational awareness into previously denied areas. The robots range in
size from those that can be thrown to robots that weigh 30 to 60 pounds.
The stated technical objectives of the program are:
Machine Perception: Provide multi-sensor hazard detection
at 20 Hz, non-GPS position estimation, and multi-source
topological mapping with 90% accuracy.
Autonomy: Demonstrate robust traversal of complex terrain
with minimal (less than 1 command per 50 m) human
supervision. Develop and implement self-recovery
behaviors (self-righting, anti-handling, etc.) that compensate
for unanticipated events.
Mobility: Negotiate complex obstacles (stairs, rubble, etc.)
and barriers (barbed wire, fences, etc.) with greater
efficiency and less operational risk than human-oriented
alternatives.
Human-Robot Interface: Create a non-distracting, alert-
based system that allows a single operator to task, control,
and interact with no fewer than three different platforms
simultaneously.
Collaborative Robotic System Integration: Employ a
collaborative team of multiple robots that cooperate to
perform complex missions that exceed the capabilities of any
single platform. Develop and demonstrate collaborative
semi-autonomous behavior (docking, power exchange,
partner inspection, etc.) that allows teamed platforms to
function reliably in adverse conditions.
Annex F. Unmanned
_______________________________________________Ground Vehicle Technology
209
Miniature Robots. The DARPA Distributed Robotics program has
developed a variety of miniature robots that are approximately one inch to
five inches in size. Mobility modes such as wheels, tracks, jumping,
climbing, slithering, and swimming are being developed. One robot was
designed to be shot from a M203 Grenade Launcher through a building
window and then be teleoperated through a building using wheeled
locomotion and sending live video back to an operator outside the
building. This robot also has a spring-operated foot that allows the robot to
jump and potentially climb stairs. Another robot was developed with a
vortex generator that allows the robot to be attached to a vertical surface
and climb upward. It can also travel along ceilings. Also developed were
robots that can change shape from a track to a snake to a legged robot.
Working with the DARPA Software for Distributed Robotics program,
algorithms and networks are being developed to allow multiple miniature
robots to operate collaboratively to accomplish a specific mission. Other
DARPA programs are investigating biomimicry for robot locomotion and
control.
Miniature robots are appealing because they are portable, cheap, have
reduced visible and acoustic signatures, can be owned and operated by
individuals in a small team, and can perform tasks that larger robots or
humans cant perform. Experiments with operational forces have validated
their strong appeal and military applicability. Research and development
in miniature mobile robots is relatively new and has been driven primarily
by the entertainment industry, which is producing toys with greater and
greater sophistication. Lego, for example, has developed a kit from which
children can build a wide variety of robots. The modular approach allows
flexibility in platform design, selection of a variety of simple sensors,
alternative uses of motors for locomotion or actuation and a simple user
language that allows children to program robot behavior on a PC and
automatically download to the robot via infrared communications. These
robots are very crude, mechanically fragile, and offer little utility for
military operations.
Miniature robots, however, offer great promise for future military
operations. Used in a system in conjunction with other robots, unmanned
air vehicles, unmanned ground sensors, and small teams, miniature robots
can add value in unique ways. Their primary contribution in surveillance
and reconnaissance missions is their ability to get close to a target and
either provide high-resolution information about the target or serve as a
sacrificial target designator. They can also be used in a manner similar to
unattended ground sensors. But by moving they can gain better
surveillance vantage points or improved signal transmission/reception
positions. Operating in collaborative groups they can gather information
DSB Summer Study on
Defense Science and Technology ___________________________________________
210
over large areas and transmit target information along with coordinates to
a central location for compiling into a comprehensive situational
awareness system. Moving also allows robots to calculate relative target
positions. Miniature robots can be propelled into buildings, conduct
searches, and transmit real-time information back to operators.
The key to successful robot design, development, and transition to
operational use has been to focus narrowly on a specific mission where the
value-added offsets the high cost of low-volume production. The key to
future robot success will be the ability to develop small robots with
modular designs that can be inexpensively and rapidly tailored to meet
specific mission needs. In general, the demand for deployable robots that
can gather, process, and communicate information is limited by the lack of
technology and appropriate tactical thinking about the effective use of
robots for small-team operations.
Because a robot is essentially a system that operates within a system,
technical progress needs to be made in several challenging areas. These
areas are mobility, communications, navigation, sensor integration, sensor
and information processing, autonomy, collaboration, power, and human-
computer interfaces, all within the context of relevant military missions.
Mobility
Novel locomotion mechanisms for small robots to move toward an
objective and avoid or overcome obstacles are needed. Small motors offer
optimum efficiency when rotating at high speed. These speeds are orders
of magnitude higher than wheels or tracks can accommodate. Coupling
the motor speed to wheel or track speed requires either dropping the
applied voltage or employing reduction gears; both methods result in
drastically reduced efficiency.
Small robots need to be fast and at the same time have high torque to
enable them to overcome obstacles or ascend inclines. Current motors that
are optimized for high speed exhibit low torque; high torque implies low
speed. New motor concepts that offer small size, high speed, high torque,
and high efficiency at the robot platform level are needed. Wheeled
robots typically cannot overcome obstacles that are larger than
approximately one-half the diameter of a wheel. Other forms of
locomotion have similar constraints. Combinations of multiple forms of
locomotion that include, for example, wheels (good for smooth surfaces),
tracks (good for rougher terrain), jumping (good for overcoming
obstacles), slithering (good for uneven terrain), climbing (good for vertical
surfaces) or other novel ideas that allow movement in unstructured
Annex F. Unmanned
_______________________________________________Ground Vehicle Technology
211
environments are needed. A small general-purpose robot that is able to
traverse all forms of terrain is probably not feasible.
Communications
Since no fully autonomous robots will exist in the foreseeable future,
communicating to and from robots is essential to both control and receive
sensor information. Several modes of communications are needed.
Teleoperation requires signals from an operator to the robot at speeds
sufficient to obtain the desired performance. Some semi-autonomous
behavior is enabled by providing a robot waypoints and then allowing the
robot to find its own way without supervision unless it gets into trouble.
The robot must transmit sensor-derived information back to an operator.
When collaborating with other robots, networked communications
among the robots is needed. A variety of communication modes have
been employed including acoustic, radio frequency (RF), and optical.
Acoustic methods have limited operating range, multipath and noise. RF
systems have difficulty with ground coupling, multipath, and jamming.
Most robot developers seek inexpensive methods for communications and
typically use 800-to-900-MHz miniature transmitters or receivers or 1.2
GHz wireless local area networks that are commercially available and
usable without the need for Federal Communication Commission licenses.
A common problem with these radios is that potential adversaries easily
intercept their signals and they are easily jammed.
Frequency stomping has been a problem in the conduct of field
experiments with operational forces when multiple robot contractors are
all trying to operate simultaneously on the same frequency. Optical
communications have also been used. This approach suffers from
problems caused by light scattering, reflections, and transmission in rain
or fog, and it presents problems with aiming a beam between moving
robots or people. More robust communications approaches are available
but their size, weight, power requirements and available bandwidth
precludes their use in small robots. Significant investment in the
telecommunications industry (cell phones and wireless networking) is
driving the technology in the right direction, but characteristics like wide
bandwidth, security, antennas, and low cost are not yet available for use
on miniature robots in military environments.
Navigation
Most robots are teleoperated. They navigate by remote control with an
operator issuing commands while watching the robot. Cameras on robots
DSB Summer Study on
Defense Science and Technology ___________________________________________
212
can be used to provide location information to an operator when the ability
to physically see the robot is otherwise obstructed. If pre-knowledge of
the terrain exists, route planning can be used to program a robot with
waypoints. In this case, robots must know when they reach a waypoint so
they can change direction. Dead reckoning has been used by counting
wheel turns, but this approach suffers from problems of accuracy,
especially if wheels slip when traversing loose sand, or mud, or rubble.
Global positioning can be used when operating in areas that have clear
access to satellites. But GPS may only be intermittently available when
operating in urban canyons or inside buildings. Also, accuracy becomes
an issue when using commercial C/A code unless time is available to
establish differential sites. Military-accuracy GPS devices are typically
too large for use with small robots. Inertial navigation devices
(combinations of gyros and accelerometers) have been used to provide six-
degree-of-freedom positioning data. However, the state of accelerometer
technology allows navigation accuracy for only very short periods of time
(minutes). Cooperative navigation can be employed using active
mechanisms such as lasers, radio frequency, and acoustic sources to range,
triangulate or use time-difference-of-arrival methods to determine robot
positions. These mechanisms are less covert because they emit energy
into the environment. They also typically require direct line-of-sight.
Sensor Integration
Sensors are an essential component in any mission-oriented robot.
Sensors are needed to provide real-time feedback to its control system or
to an operator. Simple sensors such as proximity sensors, acoustic
sensors, and cameras have been used for this purpose. Sensors are also
needed to obtain information about targets of interest in the environment.
Military operators desire the ability to place all of their personal sensors
(eyes, ears, smell, feel) at a remote location of interest without physically
being collocated.
The most attractive sensor is the video camera, because it provides the
greatest amount of information in the shortest amount of time. A single
camera can assist in teleoperation as well as provide information about
targets. A wide variety of other sensors can be employed to detect the
presence of specific physical (temperature, pressure, light, vibration, etc.),
chemical, or biologic substances. Since none of these sensors have been
developed specifically for application on small robots, their use requires
mechanical and electrical tailoring for physically mounting and functional
operation.
Annex F. Unmanned
_______________________________________________Ground Vehicle Technology
213
Useful available sensors have a variety of deficiencies. Sensors
typically require a large aperture to achieve maximum signal collection.
Available sensors tend to be too large for small robots. Shrinking the size
of the sensor also shrinks the collection area and, while it still can perform
its intended function, its robustness declines proportional to the area
reduction. Also, there exist no standard interfaces for sensors. Every
different sensor has its own unique input/output characteristics that result
in unique system hardware and software interfaces.
A robot needs multiple sensors to determine information about its own
state, to assist in navigation, and to obtain target information. The physical
placement, interface requirements, and on-board networking approaches
present a complex design task that must also consider the minimization of
overall robot energy requirements. The positive identification of a target
of interest is dependent upon the physical, chemical or biological
signatures of the target.
There is no sensor that can automatically positively detect and identify
a human, for example. It is only through a combination of data obtained
from multiple sensors about the known characteristics of a human that a
probability can be assigned to the automatic detection or identification of
that human. Incorporating a suite of sensors with an acceptable composite
fidelity, dynamic range, and environmental ruggedness into a system with
stringent size and power constraints requires the development of novel
design philosophies, architectures, and tools that are focused on tight
integration.
Sensor and Information Processing
Since sensors react to physical, chemical or biological stimuli and
typically provide this data in the form of voltages or currents, some form
of processing is necessary to translate this data into information. Many
companies incorporate some form of signal processing in the sensor
product; others simply provide the sensing mechanism. It is therefore
possible that a robot with five different commercial off-the-shelf sensors
could also have five separate signal processors, each of which is tailored
to a sensing mechanism. Robots with multiple sensors need to address the
size, power, and cost trade-off of using a variety of sensor and signal
processing products or multiplexing raw sensors with a single signal
processor. In any case, the output of first-stage signal-processed data is
insufficient to make acceptable decisions about the object being sensed.
Sensor data needs to be fused into a more comprehensive
representation the object being sensed. While significant progress has
been made in sensor fusion technology, the processing requirements are
DSB Summer Study on
Defense Science and Technology ___________________________________________
214
complex and typically require computational power in excess of what
could be reasonably incorporated in a small robot. If this is true for a
specific mission scenario, then the sensor data could be transmitted to a
central location for processing. In addition, if multiple robots are being
used to accumulate data from a wide-area search, then it is highly likely
that each of the robots would transmit its findings to a central location.
This, however, requires communication channels with adequate bandwidth
that perform within the constraints mentioned in the communications
section above.
Autonomy
A perfect robot or system of robots would be capable of accepting a
mission level command, similar to what one person would instruct another
person to do (search that area for evidence of explosives and render
helpless any that are found), and then carry out its mission with high
probability of success and no human intervention. We are decades away
from even approaching this level of autonomy.
Intelligent-like behavior can be demonstrated through a combination
of sensors and processing. There are four distinct levels to consider. The
first level is for internal control of the robot, which involves coordinating
the actions of motors, gears, flaps, and the like for locomotion and
understanding the robots internal state. The second level involves
coordinating on-board sensors that sense and interrogate the external
environment for the purpose of navigation. The third level involves
coordinating on-board sensors to detect, identify, and/or decide about
mission-related objects (targets). The fourth level involves collaborating
with other robots to achieve the mission.
Most robot developers have a good understanding of how their robots
are internally controlled. Upper-level commands are typically translated
into voltages by a processor and digital-to-analog converter to control the
speed and direction of motors which in turn control wheels, tracks, leg
extensions, or other propulsive parts. In some cases, sensors that sense the
state of the robot (wheel rotation, leg extension, flap position, etc.) provide
feedback to a processor to adjust robot performance. Internal control also
includes the hardware and software required to achieve reconfiguration to
a different locomotion mode (e.g., track to snake).
Internal robot control and mechanical and electrical designs are tightly
coupled. During the typical development process, trade-offs are
constantly being made by adding additional software to execute better
control or to change mechanical and electrical designs. Optimization for
efficient operation is effected after first prototypes are fabricated. Care
Annex F. Unmanned
_______________________________________________Ground Vehicle Technology
215
must be taken to assure that the robot uses minimum energy for mission
success. Non-optimized hardware and software designs can consume
excess power and leave little energy available for mission-related sensors,
processing, and communications. On a small robot, the controller is
typically an 8-bit processor (PIC) operating at 20 MHz with 8k of ROM
code and 192 bytes of RAM. One processor is usually dedicated to motor
control. A second processor serves as a central manager by parsing
incoming messages and handling any local behaviors.
A second level of intelligent-like behavior involves understanding the
immediate environment to allow the robot to make decisions about where
it is currently located and to move to a different location. A third
processor is typically used to coordinate the sensor information and
provide this to the central processor. There are two distinctly different
approaches. The first includes using some form of navigation aid such as
GPS, inertial navigation, loran, or dead reckoning (counting wheel
rotations and compass readings after leaving a known location). In this
approach, the robot knows its current location and is provided with
waypoints to travel to a different location.
While progress has been made in these technologies, each have
deficiencies and none of them is adequate for operational use with mission
times greater than tens of minutes. The second approach involves
executing degrees of local behavior like obstacle avoidance, wall
following, homing and follow the leader. They derive their knowledge of
the environment from simple sensors (touch, pressure, temperature,
acoustic, semiconductor lasers, RF or LED rangefinders, LED
communications, and cameras) and progress along a pre-programmed path
or seek local maximum sensor inputs.
The third level of intelligent-like behavior involves gathering
information associated with the mission. This includes capturing images
or video or sensing the presence of a target of interest (people, vehicles,
weapons, chemicals, bio agents, explosives, etc.) by using a combination
of sensors such as acoustic, temperature, pressure, or vibration. This is a
very complex task for a small robot. It involves interrogating a sensor
suite, potentially fusing sensor data, processing raw data into information,
and making decisions about thresholds. It involves detection,
classification, and identification. While a significant amount of R&D is
being conducted in this area, the ability to incorporate an adequately
sophisticated set of of sensors and high-speed processing on a small robot
is beyond current technical capability. Small robots can, however, sense
raw data and communicate to a central processor capable of analyzing and
deciding and then communicating reaction information back to the robot.
DSB Summer Study on
Defense Science and Technology ___________________________________________
216
The use of many distributed robots (mobile sensor platforms) can add
significant value to a surveillance or reconnaissance mission.
The fourth level of intelligent-like behavior is exhibited through
collaboration and control of multiple small robots. Little progress has
been made in overcoming significant technical barriers to collaboration.
These barriers are primarily associated with communications and
networking. Small robots operate close to the ground and
communications are limited to short ranges (10 to 100 meters) because of
ground coupling, multipath, small available antenna volume, interference,
and power requirements associated with transmit and receive duty cycles.
Networking solutions that try to overcome communication limitations
result in large processing delay times. While progress is being made in
this area, the current state-of-the-art for small robots results in effective
behavior at the expense of time. So, while the overall robot group can
perform a task, the extended time consumed make the group appear to be
sluggish or slightly stupid.
The intelligent behavior of a robot and therefore its level of autonomy
is difficult to characterize quantitatively. This is primarily because the
definition of human intelligence and autonomy suffers from similar
difficulties. Intelligent-like behavior requires a systematic coordination of
multiple sensing modalities, sensor- and application-specific signal
processing, decision making in unstructured and sometimes very complex
environments, and some form of collaboration with human operators.
Incorporating locomotion, sensing, processing, and communications on a
single platform is possible, but that platform is by necessity large and
expensive and, to be effective, needs to be tailored to a specific mission.
The integration of these functions on a small robot is difficult, and
progress in all of the technology areas above needs to be developed within
the context of specific missions.
Collaboration
The integration of many smaller, less capable robots into a system of
distributed robots shows great promise. It has been postulated that many
small robots can perform some tasks (not all tasks) more effectively than a
single large robot. While simulations of large numbers of robots have
demonstrated positive results, this hypothesis has not been proven with
real robots in real environments. It is still not clear whether linearly
increasing the number of robots performing a single task results in linear,
sub-linear, or exponential effectiveness. It is, however, intuitive that for
some missions like area searches, multiple collaborating robots assigned to
individual sectors (parallel processing) can probably accomplish the task
Annex F. Unmanned
_______________________________________________Ground Vehicle Technology
217
in much less time than a single robot trying to serially search the area.
Effective collaboration requires effective communications among robots
and from robots to central processing sites. It also requires novel
networking architectures that are adaptable to mission-specific tasks.
Finally, efficient algorithms that operate within the architecture need to be
tailored to the robot capabilities, the network and the task. Much progress
has been made in the collaboration of unattended fixed ground sensors, but
using these technical approaches on a large number of mobile sensors
offers order-of-magnitude-greater challenges.
Power
Energy requirements for small robots present serious problems.
Locomotion, sensing, processing, communications and power all require
volume. It is difficult to provide the energy density needed in a small
package. There are four main approaches to address the power issue. The
first focuses upon the selection of low-power electronics and highly
efficient mechanical actuators. The second involves optimizing system
performance for specific missions. For example, the processor should not
be constantly using clock cycles to interrogate an input/output channel
when it is known that no transmission or reception is needed during
specific times. The third involves exploring novel energy supply
technologies such as using circular watch batteries for wheels, conforming
a plastic battery over the surface of the robot, exploiting new energy
sources (fuel cells, MEMS generators, etc.), and examining energy
scavenging techniques. The fourth level explores the tactical deployment
of the robots so they are used in manners that minimize energy
consumption.
Small robots can operate continuously for tens of minutes but less than
an hour depending upon the various duty cycles associated with
locomotion, sensing, processing, and communicating. For highly dynamic
missions like hostage rescue and small building takedowns, mission times
are short and matched to robots power needs. For longer duration
missions, techniques are required that shut down most of the robots
functions but continue to power a seismic sensor, for example, that can
wake the robot into full surveillance operation.
Human-Computer Interfaces
Most robotic operations require interventions by human operators for
mission planning and programming, direct teleoperation, providing
assistance to robot operations, and interpreting robot sensor information.
Human-computer interface (HCI) is typically achieved after the robot
DSB Summer Study on
Defense Science and Technology ___________________________________________
218
system has been developed. HCI provides an interesting design problem,
since human capabilities are fixed at one end and robot interfaces are fixed
at the other and the common tools for HCI such as computers, joysticks, or
displays are not typically tailored to operations in extreme environments.
Operators typically need to control the robot, interpret the information
it is providing, and make decisions in time-critical, dangerous
environments. The operator probably has other non-robot functions to
perform such as shooting a weapon, communicating to other team
members, or serving as a medic. A very sophisticated and capable robot is
likely to be completely ignored if the HCI is not simple to use, does not
provide intuitive information, or is too heavy or bulky.
Most operator control units consist of a laptop computer with a
joystick packaged in a rugged suitcase. This setup requires an operator to
place the unit on a level surface, view robot operations either on a sun
immersed LCD screen or one that projects light at night, and occasionally
enter information via key strokes while wearing protective gloves. Since
this in an unacceptable interface, other HCI hardware attempts to match
the operators body though the use of wearable computers and head
mounted displays. These solutions still require additional development to
reduce their size, bulk and heat signature.
Of significant importance is how information is presented to the
operator. Typical computer window applications require operator
attention to guide and navigate through menus and, thereby, limit the
operators ability to focus on information. Other interfaces provide a
display that shows real-time video from a robot camera. This often
requires an operator to view the information even when nothing significant
is being presented.
There is a strong need for simple-to-use hardware that is
environmentally robust, reliable, low power, and unobtrusive. The
presentation of information needs to be accomplished in a manner that is
intuitive and appeals to the perceptual rather than the cognitive abilities of
the operator. New efforts that create and present representations of the
environment and objects of interest are needed. Also needed are methods
to create representations of representations (meta-representations) to put
objects and environments in a mission context.
Many robots are developed with the presumption that their unique
design will offer discriminating mobility for a wide variety of users.
Developers often become frustrated when military operators inevitably
want to add their own sensors and radios resulting in robot modifications
that make them larger and able to carry more batteries. A mistaken
conclusion is that military operators want large robots. They dont; they
Annex F. Unmanned
_______________________________________________Ground Vehicle Technology
219
want small, lightweight capability. Robots should not be thought of
simply as platforms. They must be designed and developed as a system
that provides a capability. The original design must include unique
mobility, navigation, communications, processing, autonomy, and human-
computer interfaces. The robots must be able to operate with people, other
unmanned vehicles, and other sensors within the context of tactics and be
able to provide a capability that helps successfully perform a mission.
DSB Summer Study on
Defense Science and Technology ___________________________________________
220
Annex G.
__________________________________________________________Remoter Force
221
ANNEX G.
REMOTER FORCE ____________________________________
The U.S. Department of Defense has expanding experience with
unmanned platforms. Primarily, this experience has been with unmanned
aerial vehicles, which are increasingly an indispensable tool for our
warfighters. Indeed, it seems that the challenges of creating significant
unmanned capabilities are in the category of present-day urgent needs.
While there are important developmental improvements needed in
unmanned platform capabilities, a conclusion of this report is that all of
the important engineering problems are already being addressed to some
degree in the Departments activities with few, if any, obvious ways in
which to improve the pace of this work.
There is, however, another opportunity, that would combine the
capabilities of current and future unmanned platforms with current and
future communications capabilities. What is envisioned is a day when
unmanned platforms are remotely operated not from beyond the hill or
over the horizon, but from anywhere in the world. Such an approach
would have additional collateral benefits, namely, that it might relieve
some of the difficult technological challenges that exist today by altering
the balance between automation and human remote control. Altering this
balance would permit duty cycles that better optimize work and rest cycles
since many more individuals would be participating. As well, this
approach would relieve some of the detailed control tasks necessary for
automated systems by including, at least in the near term, a human in the
loop.
Interpretation of information requiring human processing capabilities
would be dramatically enhanced. Perhaps most importantly, however,
remote operations would change the nature of our views on military
personnel and allow DoD to effectively use people with extraordinary
skills outside the normal circle of uniformed personnel. Indeed, in the
limit, active duty personnel might consist not only of uniformed personnel
selected for the historically significant war fighting attributes, but also of a
remoter force selected for their elite capabilities in remote control
operations.
An entire Army of such remoters is in training today. (See Figure
G-1.) The revolution in massive multiplayer gaming was discussed in
DSB Summer Study on
Defense Science and Technology ___________________________________________
222
Chapter III. The idea of a remoter force seeks to extend the concept of
this continuous training and concept exploration to a concrete military
capability.
Figure G-1. In TrainingAn Army of Remoters.
This idea amounts to what was referred to in Chapter III as turning the
reality dial. MMP is a virtual environment that explores doctrine and
behaviors; the Remoter Force is a single realization of the power of such a
massively collaborative tool.
The overall objective, therefore, is to include many more people in any
conflict in which the Department engages. Not by putting more boots on
the ground, but by force-multiplication: for every one person on the
ground, a group of remoters would be responsible for controlling small
advanced surveillance systems, processing that data, managing the 24-
hour duty cycles required, and performing other support activities. This is
the concept behind Enders Game, the science-fiction novel written by
Orson Scott Card, which effectively blends the remote world with the
physical. This concept would redefine the U.S. fighting force and create a
high-fidelity, system-level capability out of massively parallel, low-
fidelity pieces that capitalize on the ability to use large, distributed, secure
world-wide networks effectively.
Annex G.
__________________________________________________________Remoter Force
223
For all the revolutionary changes and opportunities that are apparent in
this (potentially inevitable) technical concept, it is also one of the most
disquieting, particularly for military commanders. In particular, the lack of
direct control implied by this concept and the reliance on system-level
redundancy over individual element reliability causes great consternation.
Among the key questions are the following:
a. How would we determine that the right person is fighting?
b. If a 15-year-old has world-class talents in remote control,
will he or she have the discipline and emotional maturity
necessary to fight in a battle?
c. What are the implications of putting weapons systems on
these remote systems at the hands of those not directly on the
ground?
d. What are the training implications?
e. What about the limitations of current platforms in terms of
endurance, payload, and so on?
f. What is the compromise implied by reducing the fidelity of
individual platforms so as to achieve the expendability that
would be required to permit non-military personnel to
control them?
All of these are problems to be overcome, since they imply new
concepts of operations, new procedures, new training, and new enlistment
possibilities. These are also cultural and organizational issues, not purely
technical ones. Technological requirements also exist, such as the need to
ensure security of remote unmanned platform operations with high-
bandwidth and near-continuous communications. And we see the
implications for training as favorable, especially when combined with the
massive multiplayer concepts discussed in Chapter III. Indeed, these
capabilities would permit DoD to continuously train and test for the most
talented pool of remoters. It should be noted that remoters and on the
ground forces should train together in order to gain an element of
understanding and trust.
As mentioned earlier, the participation of a remoter force is likely to
be a future inevitability, especially as wireless, secure communications
and high-bandwidth infrastructures improve. Adversaries who do not draw
such strong lines of distinction between their fighting personnel and
civilians and who are more willing to create distributed command cells
will not be so reticent to embrace the power of this collaborative approach.
The events of September 11th and the role that personal-computer based
DSB Summer Study on
Defense Science and Technology ___________________________________________
224
flight simulation, as well as higher fidelity flight simulation, played in
preparing the terrorists for their role as hijackers and pilots is but a small
premonition of what may be to come.
Technological challenges involving power, sensing, mobility and
navigation, among others, are prevalent in the use of unmanned platforms.
These challenges are currently addressed in a myriad of DoD programs.
(See Annex F for a discussion of ground-platform technology.) These
programs are expansive and dedicate large dollars to the overall
improvement of such platforms. There appears to be no significant
technological breakthrough on the horizon that would significantly alter
the pace of improvement, which is largely evolutionary. This observation
should not be interpreted as an indictment of current efforts, but rather
encouragement to stay the course. The realization of a remoter force will
perhaps ease some of the demand by focusing on a largely underdeveloped
aspect of the use of such platforms: namely, achieving the desired goals
through massive parallelism and the direct intervention of humans on a
large scale.
Appendix H.
___________________________________ Glossary of Acronyms and Abbreviations
225
ANNEX H.
GLOSSARY OF ACRONYMS
AND ABBREVIATIONS__________________________________
ACTD Advanced Concept Technology Demonstrations
AEF Air Expeditionary Force
BDA Battle Damage Assessment
BW Biological Warfare
C4ISR Command, Control, Communications, and
Computers and Intelligence, Surveillance, and
Reconnaissance
C4I Command, Control, Communications, Computers
and Intelligence
CEO Chief Executive Officer
CINC Commander-in-Chief
CJCS Chairman, Joint Chiefs of Staff
CONUS Continental United States
CTO Chief Technology Officer
DARPA Defense Advanced Research Project Agency
DDR&E Director, Defense Research and Engineering
DOT&E Director, Operational Test and Evaluation
DSB Defense Science Board
DTED Digital Terrain and Elevation Data
DTRA Defense Threat Reduction Agency
EMD Engineering and Manufacturing Development
EMW Expeditionary Maneuver Warfare
FCS Future Combat System
FDA Food and Drug Administration
GMTI Ground Moving Target Indicator
GPS Global Positioning System
HCI Human-Computer Interface
DSB Summer Study on
Defense Science and Technology ___________________________________________
226
HHS Health and Human Services
ISR Intelligence, Surveillance and Reconnaissance
JFCOM Joint Forces Command
MEMS Micro-Electro-Mechanical System
MMP Massive Multiplayer
NCO Network Centric Operations
NIH National Institutes of Health
OSD(C3I) Office of the Secretary of Defense for Command,
Control, Communications and Intelligence
PPBS Planning, Programming and Budgeting System
R&D Research and Development
RAP Rapid Acquisition Program
RDT&E Research, Development, Test and Evaluation
RF Radio Frequency
S&E Scientists and Engineers
S&T Science and Technology
TEG Thermo-electric
TRAC Threat Reduction Advisory Committee
UAV Unmanned Aerial Vehicle
UHF Ultra High Frequency
USD(AT&L) Under Secretary of Defense for Acquisition,
Technology and Logistics
VC Venture Capital