introduction to ai robotics chapter 7. the hybrid deliberative/reactive paradigm 2012. 11. 14

35
Introduction to AI Robotics Chapter 7. The Hybrid Deliberative/Reactive Paradigm 2012. 11. 14 Hyeokjae Kwon

Upload: shaun

Post on 24-Feb-2016

49 views

Category:

Documents


1 download

DESCRIPTION

Introduction to AI Robotics Chapter 7. The Hybrid Deliberative/Reactive Paradigm 2012. 11. 14. Hyeokjae Kwon. Objectives. Describe the hybrid paradigm in terms of 1 ) SPA and 2 ) sensing organization. - PowerPoint PPT Presentation

TRANSCRIPT

Page 1: Introduction to AI Robotics Chapter 7. The Hybrid Deliberative/Reactive Paradigm  2012. 11. 14

Introduction to AI Robotics Chapter 7. The Hybrid Deliberative/Reactive Paradigm

2012. 11. 14

Hyeokjae Kwon

Page 2: Introduction to AI Robotics Chapter 7. The Hybrid Deliberative/Reactive Paradigm  2012. 11. 14

Objectives• Describe the hybrid paradigm in terms of 1) SPA and

2) sensing organization.• Given a list of responsibilities, be able to say whether it be-

longs in the deliberative layer or in the reactive layer.• List the five basic components of a Hybrid architecture: se-

quencer agent, resource manager, cartographer, mission planner, performance monitoring and problem solving agent.

• Be able to describe the difference between managerial, state hierarchy, and model-oriented styles of Hybrid archi-tectures.

• Be able to describe the use of state to define behaviors and deliberative responsibilities in state hierarchy styles of Hy-brid architectures.

Page 3: Introduction to AI Robotics Chapter 7. The Hybrid Deliberative/Reactive Paradigm  2012. 11. 14

Basic Important concept• Paradigm

– Paradigm is both a way of looking at the world and an implied set of tools for solving problems.

• Sense, Plan, Act.– Commonly accepted robotic primitives.– Robotics have to go through these three, or at

least two process to complete a mission.• Local Processing and Global World Model

– Local: sensor data used in specific for each func-tion.

– Global: all sensor data is processed to single model.

Page 4: Introduction to AI Robotics Chapter 7. The Hybrid Deliberative/Reactive Paradigm  2012. 11. 14

Hierarchical Paradigm• What are the two main features?

– Robot operates in a top-down fashion.– All sensor data tends to be gathered to

one global world model. A single repre-sentation that planner can use to rout the action.

SENSE PLAN ACT

Page 5: Introduction to AI Robotics Chapter 7. The Hybrid Deliberative/Reactive Paradigm  2012. 11. 14

Reactive Paradigm• What are the two main features?

– Throw out planning all together.– The inputs to an act are the direct out-

put of sensors.

SENSE ACT

Page 6: Introduction to AI Robotics Chapter 7. The Hybrid Deliberative/Reactive Paradigm  2012. 11. 14

Hybrid Paradigm• Features of Hybrid Deliberative/Reactive

Paradigm– It is reactive planning, Planning to subtask is done

at one step.– Deliberative planning take a long time comparing

to the time of reactive execution.– Sensor data go directly to each behavior but also

available to the planner for construction of task-oriented global world model.

– Model-based Architecture focuses on the creation and maintenance of a global world model.

Page 7: Introduction to AI Robotics Chapter 7. The Hybrid Deliberative/Reactive Paradigm  2012. 11. 14

Hybrid Paradigm• The basic models of Hybrid Paradigm

– Sequencer: generates a set of behaviors for subtasks.

– Resource manger: allocates resources to behavior

– Cartographer: for creating, storing, maintaining map or spatial information.

– Mission Planner: Interact with human and cre-ate a plan to achieve a goal

– Performance Monitoring: monitor the process of the executing, It’s self-awareness.

Page 8: Introduction to AI Robotics Chapter 7. The Hybrid Deliberative/Reactive Paradigm  2012. 11. 14

Hybrid Paradigm• Organization : Plan, Sense-Act

Page 9: Introduction to AI Robotics Chapter 7. The Hybrid Deliberative/Reactive Paradigm  2012. 11. 14

Motivation of Hybrids• Cohesion (object oriented programming)

– Reactivity:• Short time horizon (Present)• No global knowledge• Work with sensors and actuators

– Deliberation:• Long time horizon (Past, Future)• Global knowledge• Work with symbols

• Multi-tasking– Deliberative functions execute in parallel with

reactive functions.

Page 10: Introduction to AI Robotics Chapter 7. The Hybrid Deliberative/Reactive Paradigm  2012. 11. 14

Sensing OrganizationThe Map (World Model)

– Can have its own sensors– Can “eavesdrop” on other sensors– Can act as “virtual” sensor

World Map/ Knowledge

Rep

Behavior

Behavior

Behavior

Sensor 3

Sensor 1Sensor

2

Virtual sensor

Behavior control only

Feed-back

Planning only

Eaves-drop

Page 11: Introduction to AI Robotics Chapter 7. The Hybrid Deliberative/Reactive Paradigm  2012. 11. 14

Connotations of Global• “Global” isn’t always truly global in

Hybrids.• Behavioral Management

– Planning which behaviors to use requires knowledge about current and future world state

• Performance monitoring– Detecting task progress and sensor confliction require

knowledge about the robot hardware and the overall goals.

Page 12: Introduction to AI Robotics Chapter 7. The Hybrid Deliberative/Reactive Paradigm  2012. 11. 14

Architecture Styles• Managerial (division of responsibility as

in business)– AuRA : Autonomous Robot Architecture– SFX : Sensor Fusion Effects

• State Hierarchies (strictly by time scope)– 3T : 3-Tiered

• Model-Oriented (Model serve as virtual sensors)– Saphira – TCA : Task Control Architecture

Page 13: Introduction to AI Robotics Chapter 7. The Hybrid Deliberative/Reactive Paradigm  2012. 11. 14

Styles of hybrid architec-tures

● Managerial styles

● State hierarchies styles

● Model-oriented styles

Page 14: Introduction to AI Robotics Chapter 7. The Hybrid Deliberative/Reactive Paradigm  2012. 11. 14

Managerial Architectures• Description -- top agents – high level planning

↓ subordinate agents – refine plan, gather resources ↓ lowest level agents

▲ AuRA Architectures

▲ SFX Architectures

Page 15: Introduction to AI Robotics Chapter 7. The Hybrid Deliberative/Reactive Paradigm  2012. 11. 14

Autonomous Robot Architecture (AuRA)

• It consists of five subsystems -- planner : responsible for mission and task plan-ning

-- cartographer : all map making, reading functions

-- motor : motor schema

-- sensor

-- homeostatic control : modify the relationship between behaviors by changing the gain as a function of robot or other constraints

Page 16: Introduction to AI Robotics Chapter 7. The Hybrid Deliberative/Reactive Paradigm  2012. 11. 14

AuRA Architectural LayoutCartographer

Sequencer

MissionPlanner

Behavioral manager(mgr+schemas)

PerformanceMonitoring

Emergent behavior

reac

tive

deli b

erat

ive

Page 17: Introduction to AI Robotics Chapter 7. The Hybrid Deliberative/Reactive Paradigm  2012. 11. 14

The table below summarizes AuRA in term of the common components and style of emergent behavior

AuRA SummarySequencer Agent   Navigator, PilotResource Manager   Motor Schema ManagerCartographer   CartographerMission Planner   Mission Planner

Performance Monitoring Agent   Pilot, Navigator, Mission Planner

Emergent Behavior

Vector summation, spreading activation of behaviors, homeostatic control

Page 18: Introduction to AI Robotics Chapter 7. The Hybrid Deliberative/Reactive Paradigm  2012. 11. 14

Sensor Fusion Effects (SFX)• description – It is an extension to

AuRA. The extension was to add modules to specify how sensing and handling sensor failure.

Page 19: Introduction to AI Robotics Chapter 7. The Hybrid Deliberative/Reactive Paradigm  2012. 11. 14

Sensor Fusion Effects (SFX)• Deliberative layers

-- Mission planner : acts as a CEO giving a directions

-- effector

-- Task

-- Sensor All of three of above determine the best alloca-tion of effect, sensing resource and perceptual schema. -- Cartographer : map making, path planning

Page 20: Introduction to AI Robotics Chapter 7. The Hybrid Deliberative/Reactive Paradigm  2012. 11. 14

SFX (Sensor Fusion Effects)

Behaviors(using direct

perception, fusion)

SenseSenseSenseSenseMuscleMuscleMuscleActuators

Deliberative Layer Managers

SenseSenseSenseSensor

SenseSenseSenseReceptiveField

Choice of behaviors, resourceallocation, motivation, context

Focus of attention,recalibration

SensorWhiteboard

BehavioralWhiteboard

Del

iber

ativ

e La

yer

Rea c

tive

L ay e

r Parameters to behaviors,sensor failures, task progress

actions

SuperiorColliculus-likefunctions

CerebralCortex-likefunctions

Cartographer(model/map

making)Recognitionperception

Page 21: Introduction to AI Robotics Chapter 7. The Hybrid Deliberative/Reactive Paradigm  2012. 11. 14

Sensor Fusion Effects (SFX)• Reactive layers

All these layers reflect to - strategic behaviors and tactical behaviors

Tactical behavior serves as filter on strategic commands to ensure to robot acts in a safe manner in as close accordance with the strate-gic intent as possible

The interaction of strategic and tactical be-haviors is still considered emergent behavior.

Page 22: Introduction to AI Robotics Chapter 7. The Hybrid Deliberative/Reactive Paradigm  2012. 11. 14

Tactical Behaviorssensors strategic behaviors tactical behaviors actuators

follow-path speed-controlcamera drivemotor

avoidsonarsteermotor

center-cameracamerapanmotor

inclino-meter

slope

clutter

obstacleshow much vehicle turns

direction to path safe direction

safe velocity

swivel camera

strategicvelocity

Page 23: Introduction to AI Robotics Chapter 7. The Hybrid Deliberative/Reactive Paradigm  2012. 11. 14

The table below summarizes SFX in term of the common components and style of emergent behavior

SFX Summary

Sequencer Agent   Task Manager

Resource Manager   Sensing and Task Manager

Cartographer   Cartographer

Mission Planner   Mission PlannerPerformance Monitoring

Agent   Performance Monitor, Habitat Monitor

 Emergent Behavior Strategic behaviors grouped into abstract

  behaviors or scripts, then filtered by

  tactical behaviors

Page 24: Introduction to AI Robotics Chapter 7. The Hybrid Deliberative/Reactive Paradigm  2012. 11. 14

State-hierarchy Architectures• 3 – tiered (3T)

Page 25: Introduction to AI Robotics Chapter 7. The Hybrid Deliberative/Reactive Paradigm  2012. 11. 14

3 – tiered (3T)• Structure

-- planner : setting goal and strategic plans -- sequencer : select a set of primitive behaviors develop a task network -- skill manager : in this layer the skills have associated events to verify explicitly that an action has had to correct effect

Page 26: Introduction to AI Robotics Chapter 7. The Hybrid Deliberative/Reactive Paradigm  2012. 11. 14

3T Architecture

Page 27: Introduction to AI Robotics Chapter 7. The Hybrid Deliberative/Reactive Paradigm  2012. 11. 14

The table below summarizes 3T in term of the common components and style of emergent behavior

3T

Sequencer Agent   Sequencer

Resource Manager   Sequencer (Agenda)

Cartographer   Planner

Mission Planner   PlannerPerformance Monitoring

Agent   Planner

Emergent Behavior   Behaviors grouped into skills,

 skills grouped into task

network

Page 28: Introduction to AI Robotics Chapter 7. The Hybrid Deliberative/Reactive Paradigm  2012. 11. 14

Model-oriented Architec-tures

• Two of best-known model-oriented architecture▲Saphira Architecture▲Task Control Architecture

Page 29: Introduction to AI Robotics Chapter 7. The Hybrid Deliberative/Reactive Paradigm  2012. 11. 14

Saphira Architecture -- PRS-Lite it is capable of taking natural language voice com-mands from humans and then operationalizing that into navigation tasks and perceptual recognition routines.

-- virtual sensor

-- navigation tasks manage the behaviors

-- LPS (Local Perceptual Space) determine the planning and execution improve the quality of the robot’s overall behavior

Page 30: Introduction to AI Robotics Chapter 7. The Hybrid Deliberative/Reactive Paradigm  2012. 11. 14

Saphira Architecture

Page 31: Introduction to AI Robotics Chapter 7. The Hybrid Deliberative/Reactive Paradigm  2012. 11. 14

The table below summarizes Saphira in term of the common components and style

of emergent behavior

Saphira

Sequencer Agent  Topological planner, Navigation

Tasks

Resource Manager   PRS-Lite

Cartographer   LPS

Mission Planner   PRS-LitePerformance Monitoring

Agent   PRS-Lite

Emergent Behavior   Behaviors fused with fuzzy logic

Page 32: Introduction to AI Robotics Chapter 7. The Hybrid Deliberative/Reactive Paradigm  2012. 11. 14

Task Control Architecture (TCA)

-- Task Scheduling (Mission Planner) determine the goal and order of execution

-- Path Planning (Cartographer)

-- Navigation (Sequencer) to determine what the robot should be looking for, where it is, where it has been.

-- Obstacle Avoidance To factor in not only obstacle but how to respond with a smooth trajectory for the robot’s current velocity.

Page 33: Introduction to AI Robotics Chapter 7. The Hybrid Deliberative/Reactive Paradigm  2012. 11. 14

Task Control Architecture (TCA)

Page 34: Introduction to AI Robotics Chapter 7. The Hybrid Deliberative/Reactive Paradigm  2012. 11. 14

The table below summarizes TCA in term of the common components and style of emer-gent behavior

TCASequencer Agent   Navigation Layer

Resource Manager   Navigation Layer

Cartographer   Path-Planning Layer

Mission Planner   Task Scheduling LayerPerformance

Monitoring Agent  Navigation, Path-Planning,

Task-Scheduling

Emergent Behavior   Filtering

Page 35: Introduction to AI Robotics Chapter 7. The Hybrid Deliberative/Reactive Paradigm  2012. 11. 14

Thank you!