Robin R. Murphy - Introduction to AI Robotics (Intelligent Robotics and Autonomous Agents series)

List of Figures

Two views of robots: a) the humanoid robot from the 1926 movie Metropolis (image courtesy Fr. Doug Quinn and the Metropolis Home Page), and b) a High Mobility Multipurpose Wheeled Vehicle (HMMWV), a military vehicle capable of driving on roads and open terrains (photograph courtesy of the National Institute for Standards and Technology).

Two examples of UGVs being used in a hazardous materials incident exercise: a) a man-packable QinetiQ Dragonrunner with the operator control unit and antennas mounted to the backpack and b) a man-portable iRobot Packbot 510.

Two examples of rotor-craft UAVs in a hazardous materials incident exercise: a) a Leptron Avenger and b) a AirRobot 100B quadrotor.

Two types of unmanned marine vehicles used at the 2011 Tohoku tsunami response: a box-like SeaBotix ROV on the bottom and a torpedo shaped YSI Oceanmapper AUV on the top.

Timeline of major milestones in intelligent robotics.

A Model 8 Telemanipulator or waldo. The upper portion of the device is placed in the ceiling, and the portion on the right extends into the hot cell. (Photograph courtesy Central Research Laboratories.)

An RT3300 industrial manipulator. (Photograph courtesy of Seiko Instruments.)

A MOVEMASTER robot: a.) the robot arm and b.) the associated joints.

Motivation for intelligent planetary rovers: a.) Astronaut John Young awkwardly collecting lunar samples on Apollo 16, and b.) Astronaut Jim Irwin stopping the lunar rover as it slides down a hill during the Apollo 15 mission. (Photographs courtesy of the National Aeronautics and Space Administration.)

Shakey, the first AI robot. (Photograph courtesy of SRI.)

Sojourner Mars rover. (Photograph courtesy of the National Aeronautics and Space Administration.)

The iRobot Roomba vacuum cleaner that uses animal-like intelligence to adapt to rooms.

Teledyne Ryan Firebee UAV (target variant, IDF designation Shadmit) at Muzeyon Heyl ha-Avir, Hatzerim airbase, Israel. 2006. (Photograph courtesy of WikiCommons under the Creative Commons Attribution-Share Alike 3.0 Unported license.)

Visualizing how the four aspects of automation and autonomy combine to create the DNA for a particular intelligent system.

Central nervous system represented as an operational architecture.

Organization of the primitives in the Reactive Layer.

Primitives in the Deliberative Layer.

Sublayers in the Deliberative Layer with the four key deliberative functions.

Canonical AI robotics operational architecture.

The five common subsystems in intelligent robotics, adapted from Technology Development for Army Unmanned Ground Vehicles.

The three styles of interaction: a) SENSE , PLAN , ACT , b) SENSE-ACT , and c) PLAN , SENSE-ACT .

Example of the three sensing routes for a driverless car: local, global, and hybrid.

Example of a Hierarchical Systems Architecture for a driverless car.

Example of a Reactive Systems Architecture for a driverless car.

Example of a Hierarchical Systems Architecture for a driverless car.

SENSE-PLAN-APPROVE-ACT cycle.

Organization of a telesystem. (Photographs courtesy of Oak Ridge National Laboratory.)

Manipulator interface used in nuclear hot cell work. (Photograph courtesy of Central Research Laboratories.)

The types of supervisory control represented as quadrants.

Human supervisory control matrix expanded for telesystems.

Manual control in teleoperation, showing that the remote may have some onboard control.

Local display from an iRobot Packbot showing perception.

Traded control between a teleoperator and a telefactor.

Shared control between a teleoperator and a telefactor.

DarkStar unmanned aerial vehicle. (Photograph courtesy of DefenseLink Office of the Assistant Secretary of Defense-Public Affairs.)

A teleoperated Bobcat and Talon robots working together to enable the Bobcat to clear debris at Fukushima. (Photograph courtesy of QinetiQ.)

Marrs Computational Theory.

Robots built at the Bio-Bot Laboratory at Case Western Reserve University imitate cockroaches at Level 3: a.) Robot I, an earlier version, and b.) Robot III. (Photographs courtesy of Roger Quinn.)

Genghis, a legged robot built by Colin Angle, IS Robotics, which imitates an insect at Levels 1 and 2. (Photograph courtesy of the National Aeronautics and Space Administration.)

Schema theory of a frog snapping at a fly.

Graphical illustration of the feed behavior emphasizing the computational theory Level 3 specificity needed for implementing in a computer.

Schema theory representation of a toad snapping at a fly when presented with two equidistant flies.

Unified model language representation of behavior, motor, and perceptual schemas.

Feed behavior decomposed into perceptual and motor schemas.

A graphical representation of a behavior.

Action-Perception Cycle.

A collection of artificial bait, possibly the first example of humans exploiting affordances. Notice that the lures exaggerate one or more attributes of what a fish might eat.

The GRUFF system: a.) input, and b.) different types of chairs recognized by GRUFF. (Figures courtesy of Louise Stark.)

Innate Releasing Mechanism as a process for activating behaviors.

The hiding behavior expressed as actigrams and internal state for the set of component behaviors.

A graphical illustration of the coordination function.

A taxonomy of algorithms that can serve as the coordination function.

Example of an obstacle exerting a repulsive potential field over the radius of one meter.

Five primitive potential fields: a.) uniform, b.) perpendicular, c.) attraction, d.) repulsion, and e.) tangential.

Plots of magnitude profiles for a field of radius 5 units: a.) constant magnitude, b.) linear drop off with a slope of -1, and c.) exponential drop off.

A birds-eye view of a world with a goal and obstacle and the two active behaviors for the robot who will inhabit this world.