Swarm Intelligence: The Method Behind the Mobs

1
Swarm Intelligence The Method Behind the Mobs
Robert J. Marks II Distinguished Professor of
Electrical Computer Engineering, Baylor
University
NASA Office of Biological and Physical
Research Program Review California Institute of
Technology December 17-18, 2003
2
Role within BEES
(highlight)
3
What are the competing paradigms?
CONJUNCTIVE Approach Do this1 and this2 and this3
and this4 and this5 to get that.
Result Highly complex and brittle design. Loose
this4 and your system can fail.
Conjunctive statement
4
What are the competing paradigms?
DISJUNCTIVE Approach (Do this1 to get that ) or
(Do this2 to get that ) or (Do this3 to get that
) or (Do this4 to get that )
Result Highly robust and fault tolerant design.
Loose this4 and youre still in business.
Disjunctive statement
5
What are the competing paradigms?
Is DISJUNCTIVE CONJUNCTIVE? Is (Do this1 to
get that ) or (Do this2 to get that ) or (Do
this3 to get that ) or (Do this4 to get that )
(Do this1 and this2 and this3 and this4 ) to get
that. ???
In a Boolean sense,
6
Disjunctive vs. Conjunctive

• Disjunctive reasoning sometimes referred to as
  The Combs Method
• Examples of Complex Disjunctive Systems
• Swarms Insects People
• Your Body
• Animal motor functions
• Genomic symbiogenesis

William E. Combs

• J. J. Weinschenk, W. E. Combs, R. J. Marks II,
  Avoidance of rule explosion by mapping fuzzy
  systems to a disjunctive rule configuration,
  IEEE Intl Conference on Fuzzy Systems, St.
  Louis, MO, 2003, pp 43-48.
• J. J. Weinschenk, R. J. Marks II, W. E. Combs,
  Layered URC fuzzy systems a novel link between
  fuzzy systems and neural networks, Proc. IEEE
  Intl Joint Conf. on Neural Networks, Portland,
  OR, 2003, pp. 2995-3000.
• J. J. Weinschenk, W. E. Combs, R. J. Marks II,
  On the avoidance of rule explosion in fuzzy
  inference engines, Submitted to IEEE Trans.
  Fuzzy Systems, November 12, 2003.

Earl Cox, The Fuzzy Systems Handbook, Academic
Press/ Morgan Kaufman.
7
DR vs. CR Scorecard
Property Conjunctive Reasoning (CR) Disjunctive Reasoning (DR)
Scalability Exponential Linear
Plasticity Rigid Plastic
Coupling High Low
Robustness Low High
Fault Tolerance Low High
Cognitive Parallel For low order systems, CR most closely parallels human cognitive inference.. For complex systems, DR most closely parallels human cognitive inference.
Parallel Distributed Processing Ability Parallel and distributed processing increases the complexity of most properties. DR is readily applied to distributed processing as each unit has a relationship with the consequent that is independent of the other units.
8
Applied Symbiogenesis A Disjunctive Process

Disjunctively Addend
New Feature
System
Heterogeneous Disjunctive Design Genomic
Programming
Forced Symbiotic Adaptation
Evolved System
Acquiring Genomes A Theory of the Origins of
Species by Lynn Margulis and Dorion Sagan
9
Designing a Running Man
If The ball pressure is high Then Rotate joint
CW
If The heal pressure is high Then Rotate joint
CCW
OR
joint
Ball Pressure
Heal Pressure
Impose Forced symbiotic adaptation
10
Disjunctive Symbiogenetic Design
?
design of Sagittal balance.
Disjunctively Addend
New Feature
System
Forced Symbiotic Adaptation
Evolved System
11
Disjunctive Symbiogenetic Design
design of FRONTAL balance.
Disjunctively Addend
?
New Feature
System
Forced Symbiotic Adaptation
Evolved System
12
Disjunctive Symbiogenetic Design
design of BALANCED PERSON
?
13
Disjunctive Symbiogenetic Design

• CONTRAST CONJUNCTIVE BIPEDS
• ONE FOOT ALWAYS ON THE GROUND.
• Theyll never run

BALANCED PERSON WALKING Human walking is
controlled falling
14
Five Year Plan

• Formalize disjunctive paradigm as applied to
  symbiogenic processing.
• Emulate symbiogenic development of the walking
  man.
• Generate a cute name for walking man, like
  Symbio Sam or Disjunctive Dick.
• Download emulation into biped robot and force
  physical symbiotic adaptation.
• (Baylor Time Scale Robotics Lab -
  www.TimeScales.org )
• Work with JPL for NASA missions applications.

15
Homogeneous Disjunctive Systems Swarm
Intelligence
16
Applications Warfare Game Theory
Aviation Weekly , Sept 29, 2003
17
Applications Business
Swarm Intelligence A Whole New Way to Think
About Business Harvard Business Review, May
2002 Using swarm intelligence optimization,
Southwest Airlines slashed freight transfer rates
by as much as 80. Similar research into the
behavior of social insects has helped Unilever,
McGraw Hill, and Capital One, to develop more
efficient ways to schedule factory equipment,
divide tasks among workers, organize people , and
even plot strategy.
18
Applications Telecommunications
Scientific American, March 2000 Several
companies are using swarm intelligence for
handling the traffic on their networks. France
Télécom and British Telecommunications have taken
an early lead in applying antbased routing
methods to their systems The ultimate
application, though, may be on the Internet,
where traffic is particularly unpredictable.
19
Plants and Distributed Computing

• Leaves have openings called stomata that open
  wide to let CO2 in, but close up to prevent
  precious water vapor from escaping. Plants
  attempt to regulate their stomata to take in as
  much CO2 as possible while losing the least
  amount of water.
• The results are consistent with the
  proposition that a plant solves its optimal gas
  exchange problem through an emergent, distributed
  computation performed by its leaves.
• Patches of open or closed stomata sometimes move
  around a leaf at constant speed
• Under some conditions, stomatal apertures
  become synchronized into patches that exhibit
  richly complicated dynamics, similar to behaviors
  found in cellular automata that perform
  computational tasks. Our values are
  statistically indistinguishable from those of the
  same correlations found in the dynamics of
  automata that compute.

cactus leaf cocklebur

• Peak, D. A., West, J. D., Messinger, S. M Mott,
  K. A. Evidence for complex, collective dynamics
  and emergent, distributed computation in plants.
  Proceedings of the National Academy of Sciences
  USA, 101, 918 - 922, (2004).

20
Applications Optimization
Particle Swarm An (enormously effective!) multi-
agent optimization algorithm based on the
biomimetics of bird flight.
21
Application Fiction
22
What is Swarm Intelligence?Simple Rules for
Multiple Agents.

• Indy 500s Rules
• Drive Fast
• Turn Left

23
Another rule

• Drive Fast
• Turn Left
• Dont hit stuff
• Emergent Behavior
• Competition- Winning!

24
The Dumb Termite Clearing Wood

• RULES
• Run around randomly until you bump into a piece
  of wood.
• Pick it up.
• Run around randomly until you bump into a piece
  of wood.
• Put it down.
• Repeat forever.
• Q What does this do?

25
Looking for Your Lost Pet Turtle Under a Lamppost
Multi-Agent searching in the presence of sensor
range inhomogeneity.

• Agent Rule
• Diminishing Radius Momentum if the visible
  radius decreases, the momentum is increased.
• Dont tred on me.
• Emergent Behavior A parameter to tune between
  the optimization criteria.

• Tradeoffs
• Easier to look under lamppost
• Want to look uniformly in around the area.
• Pareto Optimization (Efficient Frontier)

26
A Simple Disjunctive ExtensionMulti-Agent
Criteria Uncover important search area in the
presence of sensor range inhomogeneity

• Consequents
• Velocity Components
• In direction of new discovery
• In direction of unexplored area
• Away from nearby agents
• In direction of diminished radius

• Constraints
• Information is local, or,
• Information obtained from stygmergy.

• Antecedents
• Important Parameters
• Distance from Unexplored Area
• Location of Newly Discovered area
• Distance of Nearest Agent
• Radius Diminishment

27
Five Year Plan

• Formalize disjunctive swarm paradigm.
• Applications
• NASA
• Communications
• Space Robotics
• Air Military Swarming Drones
• Navy Search Patterns
• Work with JPL for other NASA missions
  applications.