CS344M Autonomous Multiagent Systems Patrick MacAlpine Department - - PowerPoint PPT Presentation

▶
cs344m autonomous multiagent systems
SMART_READER_LITE
LIVE PREVIEW

CS344M Autonomous Multiagent Systems Patrick MacAlpine Department - - PowerPoint PPT Presentation

Sep 15, 2022 237 likes •671 views

CS344M Autonomous Multiagent Systems Patrick MacAlpine Department of Computer Science The University of Texas at Austin Good Afternoon, Colleagues Are there any questions? Patrick MacAlpine Logistics Project proposal questions? Patrick

slide-1
SLIDE 1

CS344M Autonomous Multiagent Systems

Patrick MacAlpine Department of Computer Science The University of Texas at Austin

slide-2
SLIDE 2

Good Afternoon, Colleagues

Are there any questions?

Patrick MacAlpine

slide-3
SLIDE 3

Logistics

  • Project proposal questions?

Patrick MacAlpine

slide-4
SLIDE 4

Logistics

  • Project proposal questions?

– Hand in 2 hard copies, mark 2D/3D

Patrick MacAlpine

slide-5
SLIDE 5

Logistics

  • Project proposal questions?

– Hand in 2 hard copies, mark 2D/3D – Paper on pair programming

Patrick MacAlpine

slide-6
SLIDE 6

Logistics

  • Project proposal questions?

– Hand in 2 hard copies, mark 2D/3D – Paper on pair programming

  • Next week’s readings posted

Patrick MacAlpine

slide-7
SLIDE 7

Logistics

  • Project proposal questions?

– Hand in 2 hard copies, mark 2D/3D – Paper on pair programming

  • Next week’s readings posted
  • Kim Houck RPE, Wednesday at 1, GDC 4.816

– “Evolving Structure in Deep Neural Networks”

Patrick MacAlpine

slide-8
SLIDE 8

Motivation from real insects

  • Ant colonies exhibit remarkably complex behaviors

− Food gathering − Burial − Nest building − Reproduction

Patrick MacAlpine

slide-9
SLIDE 9

Motivation from real insects

  • Ant colonies exhibit remarkably complex behaviors

− Food gathering − Burial − Nest building − Reproduction

  • Individual ants aren’t smart

− The complexity is in the environment (Simon)

Patrick MacAlpine

slide-10
SLIDE 10

Motivation from real insects

  • Ant colonies exhibit remarkably complex behaviors

− Food gathering − Burial − Nest building − Reproduction

  • Individual ants aren’t smart

− The complexity is in the environment (Simon) − They’re easily fooled out of their element (Feynman)

Patrick MacAlpine

slide-11
SLIDE 11

Motivation from real insects

  • Ant colonies exhibit remarkably complex behaviors

− Food gathering − Burial − Nest building − Reproduction

  • Individual ants aren’t smart

− The complexity is in the environment (Simon) − They’re easily fooled out of their element (Feynman) Model the ant, not the colony

Patrick MacAlpine

slide-12
SLIDE 12

Go to the Ant

  • Complex system behavior from many simple agents

Patrick MacAlpine

slide-13
SLIDE 13

Go to the Ant

  • Complex system behavior from many simple agents
  • Complexity comes from interactions, the environment

Patrick MacAlpine

slide-14
SLIDE 14

Agent Definition

Agents tied to environment

  • Agent = <State, Input, Output, Process>

Patrick MacAlpine

slide-15
SLIDE 15

Agent Definition

Agents tied to environment

  • Agent = <State, Input, Output, Process>
  • Environment = <State, Process>

Patrick MacAlpine

slide-16
SLIDE 16

Agent Definition

Agents tied to environment

  • Agent = <State, Input, Output, Process>
  • Environment = <State, Process>

Note: supports hierarchical agents

Patrick MacAlpine

slide-17
SLIDE 17

Examples from Nature

  • Ants: path planning

Patrick MacAlpine

slide-18
SLIDE 18

Examples from Nature

  • Ants: path planning
  • Ants: brood sorting

Patrick MacAlpine

slide-19
SLIDE 19

Examples from Nature

  • Ants: path planning
  • Ants: brood sorting
  • Termites: nest building

Patrick MacAlpine

slide-20
SLIDE 20

Examples from Nature

  • Ants: path planning
  • Ants: brood sorting
  • Termites: nest building
  • Wasps: task differentiation

Patrick MacAlpine

slide-21
SLIDE 21

Examples from Nature

  • Ants: path planning
  • Ants: brood sorting
  • Termites: nest building
  • Wasps: task differentiation
  • Birds and Fish: flocking

Patrick MacAlpine

slide-22
SLIDE 22

Examples from Nature

  • Ants: path planning
  • Ants: brood sorting
  • Termites: nest building
  • Wasps: task differentiation
  • Birds and Fish: flocking
  • Wolves: surrounding prey

Patrick MacAlpine

slide-23
SLIDE 23

Principles

  • Try to avoid functional decomposition

Patrick MacAlpine

slide-24
SLIDE 24

Principles

  • Try to avoid functional decomposition
  • Simple agents (small, forgetful, local)

Patrick MacAlpine

slide-25
SLIDE 25

Principles

  • Try to avoid functional decomposition
  • Simple agents (small, forgetful, local)
  • Decentralized control

Patrick MacAlpine

slide-26
SLIDE 26

Principles

  • Try to avoid functional decomposition
  • Simple agents (small, forgetful, local)
  • Decentralized control
  • System performance from interactions of many

Patrick MacAlpine

slide-27
SLIDE 27

Principles

  • Try to avoid functional decomposition
  • Simple agents (small, forgetful, local)
  • Decentralized control
  • System performance from interactions of many
  • Diversity important: randomness, repulsion

Patrick MacAlpine

slide-28
SLIDE 28

Principles

  • Try to avoid functional decomposition
  • Simple agents (small, forgetful, local)
  • Decentralized control
  • System performance from interactions of many
  • Diversity important: randomness, repulsion
  • Embrace risk (expendability) and redundancy

Patrick MacAlpine

slide-29
SLIDE 29

Principles

  • Try to avoid functional decomposition
  • Simple agents (small, forgetful, local)
  • Decentralized control
  • System performance from interactions of many
  • Diversity important: randomness, repulsion
  • Embrace risk (expendability) and redundancy
  • Agents should be able to share information

Patrick MacAlpine

slide-30
SLIDE 30

Principles

  • Try to avoid functional decomposition
  • Simple agents (small, forgetful, local)
  • Decentralized control
  • System performance from interactions of many
  • Diversity important: randomness, repulsion
  • Embrace risk (expendability) and redundancy
  • Agents should be able to share information
  • Mix planning with execution

Patrick MacAlpine

slide-31
SLIDE 31

Principles

  • Try to avoid functional decomposition
  • Simple agents (small, forgetful, local)
  • Decentralized control
  • System performance from interactions of many
  • Diversity important: randomness, repulsion
  • Embrace risk (expendability) and redundancy
  • Agents should be able to share information
  • Mix planning with execution
  • Provide an “entropy leak”

Patrick MacAlpine

slide-32
SLIDE 32

Covering of Continuous Domains

  • Simple, pheromone-based algorithm

Patrick MacAlpine

slide-33
SLIDE 33

Covering of Continuous Domains

  • Simple, pheromone-based algorithm
  • Provable properties

Patrick MacAlpine

slide-34
SLIDE 34

Covering of Continuous Domains

  • Simple, pheromone-based algorithm
  • Provable properties

− Covers whole area in a finite time

Patrick MacAlpine

slide-35
SLIDE 35

Covering of Continuous Domains

  • Simple, pheromone-based algorithm
  • Provable properties

− Covers whole area in a finite time

  • Extensions

− Repetitive coverage (continual area sweeping)

Patrick MacAlpine

slide-36
SLIDE 36

Covering of Continuous Domains

  • Simple, pheromone-based algorithm
  • Provable properties

− Covers whole area in a finite time

  • Extensions

− Repetitive coverage (continual area sweeping) − Initial pheromone profile

Patrick MacAlpine

slide-37
SLIDE 37

Covering of Continuous Domains

  • Simple, pheromone-based algorithm
  • Provable properties

− Covers whole area in a finite time

  • Extensions

− Repetitive coverage (continual area sweeping) − Initial pheromone profile − Multiple robots

Patrick MacAlpine

slide-38
SLIDE 38

Covering of Continuous Domains

  • Simple, pheromone-based algorithm
  • Provable properties

− Covers whole area in a finite time

  • Extensions

− Repetitive coverage (continual area sweeping) − Initial pheromone profile − Multiple robots − Other metrics

Patrick MacAlpine

slide-39
SLIDE 39

Covering of Continuous Domains

  • Simple, pheromone-based algorithm
  • Provable properties

− Covers whole area in a finite time

  • Extensions

− Repetitive coverage (continual area sweeping) − Initial pheromone profile − Multiple robots − Other metrics

  • Experiments

− Now multiple robots make a difference

Patrick MacAlpine

slide-40
SLIDE 40

Real Robot Applications

Trail-Laying Robots :

  • An application to real robots
  • Trails marked with a pen
  • Also use simulations (video)

Patrick MacAlpine

slide-41
SLIDE 41

Real Robot Applications

Trail-Laying Robots :

  • An application to real robots
  • Trails marked with a pen
  • Also use simulations (video)

− Future options(?): odor, fluorescence

Patrick MacAlpine

slide-42
SLIDE 42

Real Robot Applications

Trail-Laying Robots :

  • An application to real robots
  • Trails marked with a pen
  • Also use simulations (video)

− Future options(?): odor, fluorescence TERMES :

  • Termite robots
  • (video)

Patrick MacAlpine