A

1
A Warfighter Perspective

• Dennis K. McBride, Ph.D., CPE
• CAPT, MSC, USN (Ret.)
• Visiting Professor, Modeling Simulation
• Institute for Simulation Training
• University of Central Florida
• XMSF Strategic Opportunities
• 6 September 2002

2
The Problem

• Thats not my problem.
• I didnt say all those things I said.
• Any Navy (DoD, USG) Official
• Yogi Berra

3
The Algorithm
Select
4
Exemplar Just in time training
5
The Warfighter Redefined for the New Century

• D
• I
• E
• M

Diplomacy Intelligence Economics Military
6
Evolutionary Environment of Ancestry
Present
Pliocene
Pleistocene
4 MYA 400-500 cc Aferensis
2 MYA 600-700 cc Habilis
1 MYA 900-1100 cc Erectus
100K 1350 cc Modern
3x increase in brain volume, mass
gt3x increase in volumetric brain power
Folding provides continued geometric increase
in brain power
7
Brain Growth DataFor Hominidae
Source McHenry, 1994
8
Estimates of Human Processing Capabilities
Filtering Algorithms
Process Sensory Neural Coding Cognitive To Perm.
Store
Flow (Bits/Sec) 1 Billion 3 Million 16 0.7
Filtered 8 0.003 0.000005 0.04
Orig. Filtered -- -- .0000000016 .0000000014
Only 1.6 x 10-9 of data bombarding the operator
used in real time!
Source Steinbuck, 1962
9
Neuropsychological Sex Differences
More brain volume More asymmetry More white
matter Larger amygdala Larger inferior parietal
lobe Seasonal fluctuation in testosterone Greater
aggressive behavior Better spatial skills
Less asymmetry More gray matter, density Larger
caudate Larger hippocampus Larger pallidum Larger
cingulate sulcus More hormonal cyclicity Better
verbal skills
fMRI mapping is just beginning
Conclusion We are learning more about differences
10
Todays Function and Form
Reptilian Brain
Paleomammalian Brain
Neocortical Brain
Respiration
Memory, Discrimination, Emotion
Associative Reasoning
11
The Prefrontal Cortex
What do engineers use for birth control? Why did
the golfer wear two pairs of pants?
Problem solving preparation (Carter, et
al.) Dorsolateral prefrontal cortex Emotion
processing (Adolphs, et al.) Right hemisphere
prefrontal cortex Object categorization
(Freedman, et al.) Lateral prefrontal
cortex Theory of the mind Prefrontal lobes,
left amygdala Executive control switching
(Rubinstein, et al.), simulation Prefrontal
cortex, parietal lobes
12
Information Knowledge

• Effects of information and knowledge will be
  revolutionary. DoD has not fully exploited
  today's IT. This will happen, but the next wave
  of IT will be as difficult and expensive to adopt
  as today's
• DoD will get most IT from commercial sector. But,
  adaptation and special needs must be funded by
  DoD
• Software development and maintenance must improve
• Data validation problems must be solved
• Need to improve the science of simulation
• Must solve DoDs IT workforce problems

13
Information Knowledge

• Information and knowledge systems are vital for
• Information superiority
• Coordination of organizations and operations that
  are progressively more
• Flat
• Flexible
• Smaller
• Widely dispersed
• Joint Interagency
• Coalition-oriented
• Precision, long distance fires

14
Information Knowledge
Sensors will improve and proliferate

• Performance to increase, cost to drop (e.g.,
  uncooled IR)
• Broader spectrum coverage (e.g., Dogs Nose)
• Individual collectors to huge arrays
• In-situ processing Information, rather than
  data
• Infrared focal plane arrays obsolete much of our
  IR countermeasure capabilities
• Robotics UAVs, mini- micro-, and eventually
  nanobots
• Thousands of networks replenishing military
  databases and cross feeding a variety of
  perspectives

15
Information Knowledge
Hardware and Processing technologies will improve
in speed, power, memory, and storage capacity In
15 years chips will be about 15K more capable
than todays (24 hour task in 1 hour) without
breakthroughs

• New Silicon production techniques Vs new
  materials
• Next generation processors? Anything that
  counts!
• Optical Computing Electronics will yield to the
  Optical Age
• Quantum Computing (e.g., Spintronics)
• DNA computing
• Parallel Processing Needs new software concept!
• Storage GMR 1 CD may eventually have the
  storage capacity of 1,000 CDs, providing all
  information needed for an entire mission

16
Information Knowledge
Software technologies and approaches must change

• Problems
• Excessively long delivery times
• Failure of large, complex software codes
• Information assurance/security
• Considerable supply from foreign sources
• Challenged by large amounts of information
• Creation and maintenance

If I were to select the most critical RD need
today, it is in software tools and management
techniques. Almost every system we develop
involves the dominant use of software, and many
of the problems we face stem from software
issues. This is an area where we need long-
range Research and Development efforts to develop
new technologies for future systems and
short-range management approaches (often those
being applied in commercial industry). Jacques
Ganslers Opening Statement to the HASC,
3/1/2000.
17
Information Knowledge

• And solutions
• Better software through OOP???????
• AI High expectations, disappointing results
• Two ways to give machines common sense
• Provide background info
• Teach to learn
• Benefits include
• Data/information management Wisdom from data
• Reduce choices (Choosing your TV programs?)
• Tireless, Human-like associates
• Robotic self-reinvention

18
Information Knowledge
Communications will be ubiquitous, autonomous,
secure, broadband to distribute increasing
information

• Fiber optics and other technologies will enable
  BOD to be reached in hard-connected systems
• Wireless connections to moving stations to remain
  a challenge
• Improved smart wave-forms will be developed for
  all operations
• Internet. By 2003, Internet use will almost
  double, will be an even more complex, reliable,
  and secure system

19
Chronology of IT x Psychology
1000 1640s 1840s 1900s 2000 2010
2020 2025 2050
Mech. Clocks (Aristotle) (Egypt, map)
Pascaline Liebniz computer Babbage engine Ada
Lovelaceprogrammer Boolean logic Descartes Locke
Berkely Hume Mill Wundt Mendel Galton Darwin
Explosive growth -Electrical-Electronics-Computi
ng-Communication Gestaltism Behaviorism Ethology
Evolution Cognitivism Artificial Intel. Neural
Nets
Pentium 32-bitWWW ubiquitousPDA
ubiquitous1BIPS_at_300MHZTouch screensVoice
interactiveFace recognitionE-commerce
Wireless comm Psychobiology Learning Theory
1K1T/calc/secEmbedded comp.Wireless
wiredBiocomputingSpeech recog ubiqIntelligent
tutoringReliable transla.Orbital
sensing Sociobiology
1K10 brainsComputers ubiqui.3-D
displaysSpeechgtgtmanualVirtual presenceVisual
auditoryVirtual transactorsOrbital
personal Biology
11 brainSensory implantsDirect neural
pathsAgent learning, entrepreneuringAgentgthuma
n in- teractionAll-sensor comm.Machines
think they are consc.Orbitalpersonal Bioengin
eering
Based on multiple sources including Ray Kurzweil
20
Augmented Cognition
Moores Law
Decade of the Brain fMRI
Cognitive Revolution
21
Objective 1, 2, or 3 Order of Magnitude
Improvement in Net Human-Machine Information
Capacity a Symbiotic Marriage
This Will Improve and Enhance the Quality of
Military Decision Making
22
Computer Power and Networking

• Computers will continue to follow Moores Law
• Speed and memory double every eighteen months
  (ex terraflop, petaflop and computers run on
  light)
• Bandwidth has increased speed and sending
  capacity for all users
• Programming has completely changed in the last 15
  years and will continue to change
• Neural networking, based on brain, function may
  increase capacity of computers
• Knowledge and information may become so abundant
  it will impede our ability to learn

23
Trainability

• Income gaps follow knowledge gaps
• Americans as a whole are well-educated but many
  are still unprepared for work
• Technology has not been able to bridge the
  education gap in K-12
• Learning in the future will be reorganized to
  meet the needs of the ever-changing worker

• Predictors of success in work and school
• Socio-economic status
• Graduation rates fall along racial lines
• IQ testing, regardless of education level

24
Trainability

• Digital technology will revolutionize the
  practice of teaching and training
• There is a good possibility learning will drive
  technology
• More is spent on learning than on the movie
  industry each year in the U.S.
• Three Major types of training
• Organizational
• Learner-driven
• Technology-mediated

25
Trends in Organizational Learning

• Reduce gap in product-service time
• Need for globalization
• Ability to extend instructor impact online
• Business and education units should work together
• Minimize non-instructional cost (technology
  should reduce cost)
• Desire for different types of learning
• Need performance support in training
• Choices in pricing, assessment and class selection

26
Trends in Learner-Driven and Technology-Mediated
Learning

• Learner-Driven
• Workers have good tools for learning at home-put
  them to use
• Learning independent of time and place
• Multi-modal, multi-lingual individualized
  training
• Learnbots will help decide what to learn and
  how to learn
• Need for coaching rather than sage on the
  stage teaching

• Technology-Mediated
• Open-architecture, adaptable devices and
  authoring tools
• Standard and speedy bandwidth
• Virtual libraries
• New pedagogy
• More accessible tutors
• Substantive content