Transcript Robot Control Paradigms Intelligent Mobile Robotics CS 490 Fall 2002

Robotics
“In which agents are endowed
with physical effectors with
which to do mischief”1
1Russell
and Norvig, Artificial Intelligence A Modern Approach, 2003, 901.
Derived from slides by Jerry Weinberg: http://www.cs.siue.edu/classes/Fall%202002/CS/CS490-CIS588/Weinberg/Lectures/
Why Study Robotics?
• Industrial Robots
Derived from slides by Jerry Weinberg: http://www.cs.siue.edu/classes/Fall%202002/CS/CS490-CIS588/Weinberg/Lectures/
Why Study Robotics?
• Service Robots
Derived from slides by Jerry Weinberg: http://www.cs.siue.edu/classes/Fall%202002/CS/CS490-CIS588/Weinberg/Lectures/
Why Study Robotics?
• Exploration
Derived from slides by Jerry Weinberg: http://www.cs.siue.edu/classes/Fall%202002/CS/CS490-CIS588/Weinberg/Lectures/
Why Study Robotics?
• Consumer Robots
Derived from slides by Jerry Weinberg: http://www.cs.siue.edu/classes/Fall%202002/CS/CS490-CIS588/Weinberg/Lectures/
Why Study Robotics?
• Embedded Systems
Programming
– Cars, microwave
ovens, mobile phones
• Integrated Systems
Engineering
– Mechanical
Engineering
– Electrical Engineering
– Computer Science
Derived from slides by Jerry Weinberg: http://www.cs.siue.edu/classes/Fall%202002/CS/CS490-CIS588/Weinberg/Lectures/
Why Study Robotics?
• Multitasking
Derived from slides by Jerry Weinberg: http://www.cs.siue.edu/classes/Fall%202002/CS/CS490-CIS588/Weinberg/Lectures/
Why Study Robotics?
• Its Fun!
Derived from slides by Jerry Weinberg: http://www.cs.siue.edu/classes/Fall%202002/CS/CS490-CIS588/Weinberg/Lectures/
What is An Intelligent Robot?
• A machine able to extract information from
its environment and use knowledge about its
world to move safely in a meaningful manner
Derived from slides by Jerry Weinberg: http://www.cs.siue.edu/classes/Fall%202002/CS/CS490-CIS588/Weinberg/Lectures/
Humans vs. Robots
People
Robots
Bones
Mechanical Structure
Muscles
Effectors
Senses
Sensors
Digestion/Respiration
AC/DC Power
Brain
Computer
Knowledge
Program
Derived from slides by Jerry Weinberg: http://www.cs.siue.edu/classes/Fall%202002/CS/CS490-CIS588/Weinberg/Lectures/
Robots vs. Softbots
• A softbot is a pure software agent whose
environment consists of computer file
systems, databases, and networks
– Microsoft Office Helper, Game Agents, Web
Crawlers, Expert Systems
• Robot is an active, artificial agent whose
environment is the physical world
Derived from slides by Jerry Weinberg: http://www.cs.siue.edu/classes/Fall%202002/CS/CS490-CIS588/Weinberg/Lectures/
The Real World is A Harsh Place
• Inaccessible
– nearby stimuli, limited attention, imperfect
sensors
• Non-deterministic
– Robot structure and dynamics, environment
• Dynamic
– Changes happening as decisions are made
• Continuous
– The world is not a set of discrete events
Derived from slides by Jerry Weinberg: http://www.cs.siue.edu/classes/Fall%202002/CS/CS490-CIS588/Weinberg/Lectures/
Dealing with the Physical World
A robot needs to be able to handle its environment
or the environment must be altered and
controlled.
• Closed World Assumption
– The robot knows everything relevant
– no surprises
– Reasonable only in very restricted domains
• Open World Assumption
– The robot must be able to handle unexpected events.
– The usual state of affairs
Derived from slides by Jerry Weinberg: http://www.cs.siue.edu/classes/Fall%202002/CS/CS490-CIS588/Weinberg/Lectures/
What does it take to get an intelligent
robot to do a simple task?
Robot Parts: Two Arms, Vision, and Brain
The Brain can communicate with all parts
Arms can take commands as left, right, up, down, forward, and backward
Arms can answer yes/no about whether they are touching something but
cannot distinguish what they are touching
The vision system can answer any question the brain asks, but cannot
volunteer information.
The vision system can move around to get a better view.
Derived from slides by Jerry Weinberg: http://www.cs.siue.edu/classes/Fall%202002/CS/CS490-CIS588/Weinberg/Lectures/
Why is this simple task so difficult?
•
•
•
•
Coordination is difficult
Indirect feedback
Updating world knowledge
Unexpected events
– Need to re-plan
• Different coordinate systems need to be
resolved
– Box-centered and arm-centered
Derived from slides by Jerry Weinberg: http://www.cs.siue.edu/classes/Fall%202002/CS/CS490-CIS588/Weinberg/Lectures/
Robot Control
• Two general approaches to controlling robot
behavior:
– Deliberative: reason about world, plan behaviors, act
• Human example: vacuum a room.
– Reactive: sense world, take action
• Human example: pull hand away from a hot surface
• Hybrid approaches: combine both
• Human example: ride a bike
Derived from slides by Jerry Weinberg: http://www.cs.siue.edu/classes/Fall%202002/CS/CS490-CIS588/Weinberg/Lectures/
Deliberative/Hierarchical Robot
Control
• Classic Robot Control, emphasizes planning
• Basic paradigm is Sense -- Plan --Act
• World knowledge must be represented in a form that
the robot can reason about.
Robot senses the world,
constructs a model
representation of the
world, “shuts its eyes”,
creates a plan of action,
makes the action, then
senses the results of the
action.
Derived from slides by Jerry Weinberg: http://www.cs.siue.edu/classes/Fall%202002/CS/CS490-CIS588/Weinberg/Lectures/
Deliberative: Good & Bad
• Goal Oriented
– Solve problems that need cognitive abilities
– Ability to optimize solution
• Predictable
• Dependence on a world model
– Requires a closed world assumption
– Symbol Grounding Problem
– Frame Problem
– Qualification Problem
Derived from slides by Jerry Weinberg: http://www.cs.siue.edu/classes/Fall%202002/CS/CS490-CIS588/Weinberg/Lectures/
Reactive/Behavior-Based
Paradigm
•
•
•
•
•
Rodney Brooks 1987
Ignores world models
“The world is its own best model”
Sense -- act
Reactive Paradigm tightly couples perceptions
to actions
– No intervening abstract representations or time
history
• Individual Behaviors are used as building
blocks
Derived from slides by Jerry Weinberg: http://www.cs.siue.edu/classes/Fall%202002/CS/CS490-CIS588/Weinberg/Lectures/
Where does the overall robot
behavior come from?
• No overall goal, no planning
• Emergent Behavior
– Emergence is the appearance of a novel property of a
whole system that cannot be explained by examining
the individual components, for example the wetness
of water.
– Overall behavior is a result of robots interaction with
its surroundings and the coordination between the
individual behaviors.
Derived from slides by Jerry Weinberg: http://www.cs.siue.edu/classes/Fall%202002/CS/CS490-CIS588/Weinberg/Lectures/
Reactive: Good & Bad
• Works with the Open World Assumption
– Provides a timely response in a dynamic
environment where the environment is difficult
to characterize and contains a lot of
uncertainty.
• Unpredictable
• Low level intelligence
– Cannot manage tasks that require memory
and higher level cognition
• Tasks requiring localization and order dependent
steps
Derived from slides by Jerry Weinberg: http://www.cs.siue.edu/classes/Fall%202002/CS/CS490-CIS588/Weinberg/Lectures/
Hybrid: Planning – Reactive
Interaction
• Reactive in primary control and Planner
provides advice
– Planner configures the Reactive system
• Planner is primary and Reactive provides
actions to avoid uncertain situations
– Layered approach
– Requires re-planning
• Planner and Reactive work concurrently
Derived from slides by Jerry Weinberg: http://www.cs.siue.edu/classes/Fall%202002/CS/CS490-CIS588/Weinberg/Lectures/
Spectrum of Robot Control
Derived from slides by Jerry Weinberg: http://www.cs.siue.edu/classes/Fall%202002/CS/CS490-CIS588/Weinberg/Lectures/
Where are we with robotics?
• AAAI robot competitions
– Robot Rescue
– Robot Host
– Robot Challenge
• Grace
•
•
•
•
Lots of autonomous commercial applications
Some fairly impressive research
Beginning to see consumer applications
Barely at the beginning of applications which
involve interacting with humans
Derived from slides by Jerry Weinberg: http://www.cs.siue.edu/classes/Fall%202002/CS/CS490-CIS588/Weinberg/Lectures/
TODAY!
Webcast: Artificial
Intelligence for Autonomous
Control in Space. Thurs, April
15, “PST”
http://www.jpl.nasa.gov/events
/lectures/apr04.cfm
NASA Mars Rovers
Status 24 Mar 2004
A balloonshaped robot
explorer during
a 70-kilometer
wind-driven trek
across
Antarctica.
DARPA Desert race too tough for robots
Derived from slides by Jerry Weinberg: http://www.cs.siue.edu/classes/Fall%202002/CS/CS490-CIS588/Weinberg/Lectures/