39
Introduction to AI Robotics (MIT Pres s) Chapter 2: The Hierarchical Paradigm 1 2 The Hierarchical Paradigm Describe the Hierarchical Paradigm in terms of the 3 robot primitives and its organization of sensing Name and evaluate one representative Hierarchical architecture in terms of: support for modularity, niche targetability, ease of portability to other domains, robustness Solve a simple navigation problem using STRIPS (hint: work through Sec. 2.2.2) Understand precondition, closed world assumption, open world, frame problem List two advantages and disadvantages of the Hierarchical Paradigm rganization SPA global trips Shakey ep. Arch. evaluation NHC RCA ummary

The Hierarchical Paradigm

  • Upload
    season

  • View
    27

  • Download
    0

Embed Size (px)

DESCRIPTION

The Hierarchical Paradigm. Describe the Hierarchical Paradigm in terms of the 3 robot primitives and its organization of sensing - PowerPoint PPT Presentation

Citation preview

Page 1: The Hierarchical Paradigm

Introduction to AI Robotics (MIT Press) Chapter 2: The Hierarchical Paradigm 1

2 The Hierarchical Paradigm• Describe the Hierarchical Paradigm in terms of the 3 robot

primitives and its organization of sensing

• Name and evaluate one representative Hierarchical architecture in terms of: support for modularity, niche targetability, ease of portability to other domains, robustness

• Solve a simple navigation problem using STRIPS (hint: work through Sec. 2.2.2)

• Understand precondition, closed world assumption, open world, frame problem

• List two advantages and disadvantages of the Hierarchical Paradigm

Organization-SPA-globalStrips-ShakeyRep. Arch.-evaluation-NHC-RCASummary

Page 2: The Hierarchical Paradigm

Introduction to AI Robotics (MIT Press) Chapter 2: The Hierarchical Paradigm 2

2 Hierarchical Paradigm Organization

PLANSENSE ACT

Organization-SPA-globalStrips-ShakeyRep. Arch.-evaluation-NHC-RCASummary

World model(global):1. A priori rep2. Sensed info3. Cognitive

Page 3: The Hierarchical Paradigm

Introduction to AI Robotics (MIT Press) Chapter 2: The Hierarchical Paradigm 3

2 Shakey

• First AI mobile robot

• Built by SRI (Stanford Research Institute) for DARPA 1967-9

• Used Strips as main algorithm for controlling what to do

Organization-SPA-globalStrips-ShakeyRep. Arch.-evaluation-NHC-RCASummary

Page 4: The Hierarchical Paradigm

Introduction to AI Robotics (MIT Press) Chapter 2: The Hierarchical Paradigm 4

2 Strips: Means-ends analysis

Organization-SPA-globalStrips-ShakeyRep. Arch.-evaluation-NHC-RCASummary

 INITIAL STATE: Tampa, Florida (0,0)

GOAL STATE: Stanford, California (1000,200)

Difference: 1020 miles

“Go to Stanford AI Lab”

If can’t accomplish the task in one step, pick anAction that will reduce the difference betweenThe current state and the goal state

Page 5: The Hierarchical Paradigm

Introduction to AI Robotics (MIT Press) Chapter 2: The Hierarchical Paradigm 5

2 Difference Table

 

d<=200 miles FLY

100<d<200 TRAIN

d<=100 DRIVE

Distance(difference)

mode of transportation(OPERATOR)

d<1 WALK

Organization-SPA-globalStrips-ShakeyRep. Arch.-evaluation-NHC-RCASummary

mode=difference_table(INITIAL STATE, GOAL STATE, difference)

1. Look up what to do: FLY2. Not at SAIL, so repeat3. Look up what to do: DRIVE

Page 6: The Hierarchical Paradigm

Introduction to AI Robotics (MIT Press) Chapter 2: The Hierarchical Paradigm 6

2 Preconditions

d<=200 miles FLY

100<d<200 TRAIN

d<=100 DRIVE (rental)

  DRIVE (personal car)

difference OPERATOR

d<1 WALK

 

Organization-SPA-globalStrips-ShakeyRep. Arch.-evaluation-NHC-RCASummary

How do I know if I’m at the airport or at home?Now must keep up with the state of the world

 

 

at airport

at home

PRECONDITIONS

 

Page 7: The Hierarchical Paradigm

Introduction to AI Robotics (MIT Press) Chapter 2: The Hierarchical Paradigm 7

2 Maintaining State of the World:Add and Delete Lists

 

 

 

d<=200 miles

FLY  

100<d<200

TRAIN  

d<=100 DRIVE (rental)

at airport    

  DRIVE (personal)

at home    

distance OPERATOR PRE-CONDITIONS

d<1 WALK       

Organization-SPA-globalStrips-ShakeyRep. Arch.-evaluation-NHC-RCASummary

at city Yat airport

at city Yat train station

ADD-LIST

at city X

at city X

DELETE-LIST

Page 8: The Hierarchical Paradigm

Introduction to AI Robotics (MIT Press) Chapter 2: The Hierarchical Paradigm 8

2 Class Exercise

• Write down the world model, the operator applied, the change in world state, etc. to go from Tampa to Stanford

Organization-SPA-globalStrips-ShakeyRep. Arch.-evaluation-NHC-RCASummary

 

 

d<=200 miles

FLY  

100<d<200

TRAIN  

d<=100 DRIVE (rental)

at airport    

  DRIVE (personal)

at home    

distance OPERATOR PRE-CONDITIONS

d<1 WALK    

at city Yat airport

at city Yat train station

ADD-LIST

at city X

at city X

DELETE-LIST

Page 9: The Hierarchical Paradigm

Introduction to AI Robotics (MIT Press) Chapter 2: The Hierarchical Paradigm 9

2 Strips Summary• Designer must set up

– World model representation– Difference table with operators, preconditions, add & delete

lists– Difference evaluator

• Strips assumes closed world– Closed world: world model contains everything needed for

robot (implication is that it doesn’t change)– Open world: world is dynamic and world model may not be

complete

• Strips suffers from frame problem – specifying what is and is not changed by performing an action– Frame problem: representation grows too large to reasonably

operate over

Organization-SPA-globalStrips-ShakeyRep. Arch.-evaluation-NHC-RCASummary

Page 10: The Hierarchical Paradigm

Introduction to AI Robotics (MIT Press) Chapter 2: The Hierarchical Paradigm 10

2 Example p,. 46-47

Page 11: The Hierarchical Paradigm

Introduction to AI Robotics (MIT Press) Chapter 2: The Hierarchical Paradigm 11

2 Architecture• provides a principled way of organizing a control system.

However, in addition to providing structure, it imposes constraints on the way the control problem can be solved [Mataric]

• describes a set of architectural components and how they interact [Dean & Wellman]

• Types of architectures [Levis, George Mason University]– operational architecture: describes what the systems does, not how

it does it

– systems architecture: describes how a system works in terms on major subsystems

– technical architecture: implementation details

Organization-SPA-globalStrips-ShakeyRep. Arch.-evaluation-NHC-RCASummary

Page 12: The Hierarchical Paradigm

Introduction to AI Robotics (MIT Press) Chapter 2: The Hierarchical Paradigm 12

2 Evaluating an Architecture

• support for modularity: does it show good software engineering principles?

• niche targetability: how well does it work for the intended application?

• ease of portability to other domains: how well would it work for other applications or other robots?

• robustness: where is the system vulnerable, and how does it try to reduce that vulnerability?

Organization-SPA-globalStrips-ShakeyRep. Arch.-evaluation-NHC-RCASummary

Page 13: The Hierarchical Paradigm

Introduction to AI Robotics (MIT Press) Chapter 2: The Hierarchical Paradigm 13

2 Hierarchical Paradigm…

• Top-down: – Plan, plan, plan

• Control-theoretic: – must measure error in order to control device

• Planning means:– dependence on world models

Page 14: The Hierarchical Paradigm

Introduction to AI Robotics (MIT Press) Chapter 2: The Hierarchical Paradigm 14

2 Representative Hierarchical Architectures

• Nested Hierarchical Controller (NHC)

• NIST Realtime Control System (RCS)

– Original by Albus

– Teleoperation version for JPL called NASREM

Page 15: The Hierarchical Paradigm

Introduction to AI Robotics (MIT Press) Chapter 2: The Hierarchical Paradigm 15

2 Nested Hierarchical Controller(Meystel)

Organization-SPA-globalStrips-ShakeyRep. Arch.-evaluation-NHC-RCASummary

Locates robot, goal on map

Plans path segments

Plans actions for each path

segment

Page 16: The Hierarchical Paradigm

Introduction to AI Robotics (MIT Press) Chapter 2: The Hierarchical Paradigm 16

2 NHC Planner

Organization-SPA-globalStrips-ShakeyRep. Arch.-evaluation-NHC-RCASummary

Page 17: The Hierarchical Paradigm

Introduction to AI Robotics (MIT Press) Chapter 2: The Hierarchical Paradigm 17

2 NHC (cont)

• After action is taken, poll sensors again, update world model

• Pilot informs navigator if waypoint is reached, if robot drifts off course or obstacle is encountered

• Navigator plans new path segments based on updated world model

• Comparison with Strips– Advantage: interleaves planning and acting so that it can

replan if the world is different than expected

– Disadvantage: appropriate only for navigation

Page 18: The Hierarchical Paradigm

Introduction to AI Robotics (MIT Press) Chapter 2: The Hierarchical Paradigm 18

2 RCS (Albus)

• the hierarchy – p. 58, Fig 2.7

• how the hierarchy works for navigation

• how it is implemented– nodes and modules

– planning time periods

Page 19: The Hierarchical Paradigm

Introduction to AI Robotics (MIT Press) Chapter 2: The Hierarchical Paradigm 19

2 Examples of RCS Apps

Organization-SPA-globalStrips-ShakeyRep. Arch.-evaluation-NHC-RCASummary

Page 20: The Hierarchical Paradigm

Introduction to AI Robotics (MIT Press) Chapter 2: The Hierarchical Paradigm 20

2 RCS-4 Levels

7 Battalion 2-24h

6 Platoon 5m-2h

5 Section 1-10m

4 Individual Vehicle 5-50s

3 Subsystem Level 200-500ms

2 Primitive Level 50-500ms

1 Servo Level 5-50ms

Page 21: The Hierarchical Paradigm

Introduction to AI Robotics (MIT Press) Chapter 2: The Hierarchical Paradigm 21

2 Each Level has a RCS Node

Sensory Processing, World Modeling,Behavior Generation, Value Judgment

Sensory Processing, World Modeling,Behavior Generation, Value Judgment

Engineering of Mind, Albus & Mystel, 2001

Page 22: The Hierarchical Paradigm

Introduction to AI Robotics (MIT Press) Chapter 2: The Hierarchical Paradigm 22

2 Implementation View:Nodes are Recursive

Page 23: The Hierarchical Paradigm

Introduction to AI Robotics (MIT Press) Chapter 2: The Hierarchical Paradigm 23

2 Demo III XUV

http://museum.nist.gov/exhibits/timeline/item.cfm?itemId=38

Experimental Unmanned Vehicle in action at Ft. Indiantown Gap. Photo courtesy of the Army Research Labs. Nov. 2001

Page 24: The Hierarchical Paradigm

Introduction to AI Robotics (MIT Press) Chapter 2: The Hierarchical Paradigm 24

2 Demo III Control HierarchyPLANNER

vehicle1 vehicle2

VEHICLE PLANNERcommunications plan AM plan RSTA plan

AM PLANNERDriver Plan Gaze plan

COMMS PLANNERmessage list

RSTA PLANNERgaze plan

DRIVER PLANNERVelocity Plan

GAZE PLANNERStereo Gaze Plan LADAR Gaze Plan

VELOCITY PLANNERF Wheels R Wheels F Steer R Steer

F Wheel R Wheel F Steer F Steer

Servo 50ms

Primitive 500ms

Subsystem 5s

Vehicle 1m

Section 10m

Page 25: The Hierarchical Paradigm

Introduction to AI Robotics (MIT Press) Chapter 2: The Hierarchical Paradigm 25

2 RCS XUV Example

Vehicle Level:AM Plan(A1…A10)

Primitive Level:Driver Plan(D1…D10)

Page 26: The Hierarchical Paradigm

Introduction to AI Robotics (MIT Press) Chapter 2: The Hierarchical Paradigm 26

2 t=0.5

Primitive Level:Driver Planextends to A2

Page 27: The Hierarchical Paradigm

Introduction to AI Robotics (MIT Press) Chapter 2: The Hierarchical Paradigm 27

2 t=1 s

Obstacle Detected

Primitive Level:Driver Plannew waypoints

Vehicle Level:detects too large avariation

Page 28: The Hierarchical Paradigm

Introduction to AI Robotics (MIT Press) Chapter 2: The Hierarchical Paradigm 28

2 t=1 s Vehicle Level Planner Opt 1

Vehicle Level:new AM Plan

Primitive Level:new Driver Plan

Page 29: The Hierarchical Paradigm

Introduction to AI Robotics (MIT Press) Chapter 2: The Hierarchical Paradigm 29

2 t=3

More obstacle is seen…fail upwards again

Vehicle Level:new AM Plan

Primitive Level:new Driver Plan

Page 30: The Hierarchical Paradigm

Introduction to AI Robotics (MIT Press) Chapter 2: The Hierarchical Paradigm 30

2 t=4.5s

Vehicle Level:new AM Planskip A1, go to A2

Primitive Level:new Driver Plan

Page 31: The Hierarchical Paradigm

Introduction to AI Robotics (MIT Press) Chapter 2: The Hierarchical Paradigm 31

2 t=6 s

Vehicle Level:new AM Planskip old A2

Primitive Level:new Driver Plan

Page 32: The Hierarchical Paradigm

Introduction to AI Robotics (MIT Press) Chapter 2: The Hierarchical Paradigm 32

2 Exercise: Adapt to Rescue Robots?PLANNER

vehicle1 vehicle2

VEHICLE PLANNERcommunications plan AM plan RSTA plan

AM PLANNERDriver Plan Gaze plan

COMMS PLANNERmessage list

RSTA PLANNERgaze plan

DRIVER PLANNERVelocity Plan

GAZE PLANNERStereo Gaze Plan LADAR Gaze Plan

VELOCITY PLANNERF Wheels R Wheels F Steer R Steer

F Wheel R Wheel F Steer F Steer

Servo 50ms

Primitive 500ms

Subsystem 5s

Vehicle 1m

Section 10m

Page 33: The Hierarchical Paradigm

Introduction to AI Robotics (MIT Press) Chapter 2: The Hierarchical Paradigm 33

2 Nodes are made from Modules

uff applies rule(transition rules)

sensors actions

u=uff+G(xd-x*)u is control action

x* is predicted world statexd is desired world state

uff is the feedforward control plan

transform into x*G is feedbackxd is from “above”

Page 34: The Hierarchical Paradigm

Introduction to AI Robotics (MIT Press) Chapter 2: The Hierarchical Paradigm 34

2 Nodes are made from Modules

uff applies rule(transition rules)

sensors actions

if BALL, move toward centroidif NOT BALL, turn clockwise

(feedback determines how fast)

transform into x*G is feedbackxd is from “above”

Page 35: The Hierarchical Paradigm

Introduction to AI Robotics (MIT Press) Chapter 2: The Hierarchical Paradigm 35

2 Nodes are made from Modules

uff applies rule(transition rules)

sensors actions

if BALL, move toward centroidif NOT BALL, turn clockwise

(feedback determines how fast)

transform into x*G is feedbackxd is from “above”

when to stop?how far is far enough?what about noise/fuzzy ball?…sensor noise, actuator error,rigid models

Page 36: The Hierarchical Paradigm

Introduction to AI Robotics (MIT Press) Chapter 2: The Hierarchical Paradigm 36

2 Advantages of Hierarchies

Albus and Mystel 01:

• Natural way to organize• Not intrinsically rigid• Not intrinsically inefficient

– not the same as centralized planning

– priorities and goals are clear, therefore efficient

Page 37: The Hierarchical Paradigm

Introduction to AI Robotics (MIT Press) Chapter 2: The Hierarchical Paradigm 37

2 Summary RCS

• hierarchy with node structure at each level– have operator interface (in theory)

• nodes consist of– Sensory Processing– World Model– Behavior Generation– Value Judgment

• top-down, plan for a particular horizon– control theoretic

Page 38: The Hierarchical Paradigm

Introduction to AI Robotics (MIT Press) Chapter 2: The Hierarchical Paradigm 38

2 Evaluating the Two Architectures• support for modularity:

– decomposition by functionality

• niche targetability: – good, both have been used for apps like vehicle guidance,

mining equipment

• ease of portability to other domains: – unclear, not sure if code could be reused—lots of rewriting on

previous apps

• robustness:– RCA simulates plans in advance, but not sure what it would do

with sensor or mechanical failures, etc.

Organization-SPA-globalStrips-ShakeyRep. Arch.-evaluation-NHC-RCASummary

Page 39: The Hierarchical Paradigm

Introduction to AI Robotics (MIT Press) Chapter 2: The Hierarchical Paradigm 39

2 Hierarchical Review• Describe the hierarchical paradigm in terms of the

three robot primitives• Describe sensing in the hierarchical paradigm

• What is STRIPS?

• What is the closed world assumption?• What are preconditions?• What is the frame problem?

• What are two representative architectures?• What is the NHC decomposition?

Organization-SPA-globalStrips-ShakeyRep. Arch.-evaluation-NHC-RCASummary