Cybernetics Principles captured in the design of ... - University of Haifa

doubleperidotΤεχνίτη Νοημοσύνη και Ρομποτική

30 Νοε 2013 (πριν από 3 χρόνια και 9 μήνες)

69 εμφανίσεις

1
1
Cybernetics
 Pioneered by Norbert Wiener (1940s) (From Greek
"steersman" of steam engine).
 Marriage of control theory (feedback control),
information science and biology.
 Seeks principles common to animals and machines,
especially for control and communication.
 Ashby (1952) and Wiener furthered this view of an
organism as a machine by using the mathematics
developed for feedback control systems to express
natural behavior.
 A strong two-way coupling between an organism and
its environment (situatedness).
2
Principles captured in the design of the turtle.
 Parsimony: simple is better (e.g., clever recharging
strategy).
 Exploration/speculation: keeps moving (except when
charging).
 Attraction (positive tropism): motivation to approach
light.
 Aversion (negative tropism): motivation to avoid
obstacles, slopes.
 Discernment (showing good judgement): ability to
distinguish and make choices, i.e., to adapt.
2
3
Braitenberg Vehicles
 Valentino Braitenberg (early 1980s) extended Walter s
model in a series of thought experiments.
 Also based on analog circuits and same drawbacks.
 Direct connections (excitatory or inhibitory) between
light sensors and motors.
 Complex behaviors from very simple mechanisms.
Vehicle 1
(Single motor/single sensor):Motion is always
forward in the direction of the sensor stalk, with
the speed controlled by the sensor.
Environmental perturbations (slippage, rough
terrain) produce charges in direction.
4
Braitenberg Vehicles
This results in a net motion away from the light. The photophobe
on the right is attracted to light when the wires connecting sensors
and motors are merely reversed from the photophobe (exhibiting
"aggression" by charging into the attractor).
Vehicle 2
(Two sensors/two motors): The
photophobe on the left is aversive to
light (exhibiting "fear" by fleeing)
since the motor closest co the light
source moves faster than the one
farther away.
3
5
Braitenberg Vehicles
 By varying the connections and their strengths,
numerous behaviors result, e.g.:
"fear/cowardice" - flees light
"aggression" - charges into light
"love" - following/hugging
 many others, up to memory and learning!
 Reactive control
 Later implemented on real robots
6
Braitenberg Vehicles
Vehicle 3: Same wiring as for
Vehicle 2 but now with inhibitory
connections. The vehicles slow down in
the presence of a strong light source
and go fast in the presence of weak
light. In both cases, the vehicle
approaches and stops by the light
source (with one facing the light and
one with the light source to the rear).
The vehicle on the left is said to "love" the light source
since it will stay there indefinitely, while the vehicle on the
right explores the world, liking to be near its current
attractor, but always on the lookout for something else.
4
7
Braitenberg Vehicles
Vehicle 4: By adding various nonlinear
speed dependencies to Vehicle 3, where
the speed peekspeeks somewhere between the
maximum and minimum intensities, other
interesting motor behaviors can bebe
observed.
This can result in oscillatory navigation between two
different light sources (top) or by circular or other
unusual patterns traced around a single source (bottom).
8
Asimov's Three Laws of Robotics
1. Safety:A robot may not injure a human being, or,
through inaction, allow a human being to come to
harm.
2. Service:A robot must obey the orders given it by
human beings except where such orders would conflict
with the First Law.
3. Prudence:A robot must protect its own existence, as
long as such protection does not conflict with
the First or Second Laws.
Isaac Asimov, Runaround, Astounding Science Fiction, March 1942
5
9
Artificial Intelligence.
Early:
Intelligent machine" would use internal models to search for
solutions and then try them out (M. Minsky) => deliberative
model!
Planning became the tradition
Explicit symbolic representations
Hierarchical system organization
Sequential execution
Early AI had a strong impact on early robotics
Focused on knowledge, internal models, and
reasoning/planning.
10
Flakey
SRI's mobile robot,Flakey, built in 1984
reflected the dramatic improvements 14
years makes in the state-of-the-art. It was
approximately 3 feet high and 2 feet in
diameter. Like Shakey, it had 2 drive wheels
with encoders, but their maximum velocity
was 2 feet per second instead of inches. The
robot had 12 sonar range finders for obstacle
avoidance and navigation in real spaces, not
just a laboratory like Shakey's.
Again a video camera and laser provided range-finding information over a small
area in front of Flakey. And Flakey's computers included one of the then new
personal workstations, giving it far more intelligence than its predecessor, in far
less space.
Flakey was the first robot server to use Saphira, the high-level robotics
development environment. With Saphira,Flakey could navigate and recognize
corridors, find doorways and plan a path from one place to another.
www.activrobots.com/HISTORY
6
11
ERRATIC
Both Shakey and Flakey were large, one-of-a-
kind robots costing many tens or hundreds of
thousands of dollars to build.
ERRATIC was developed in 1993 by Kurt
Konolige of SRI's Artificial Intelligence Center
to fill a need for a low-cost robot platform in
his classroom at Stanford University where he
teaches part-time.
ERRATIC participated in the NCAI94 robotics contest, placing 2nd in a field of
more than a dozen robots.
The design provided
motion and sensing services
to clients running on a
host machine, similar to Flakey's and using the same Saphira software. By
then, the host computer could be either a desktop PC connected via radio
modem, or one of the new little notebook computers, sitting atop ERRATIC.
12
Pioneer 1
In 1994,Konolige teamed with ActivMedia for
integration, software and support and with RWI for
manufacturing a commercial version of the
ERRATIC robot, called the Pioneer 1. It was a
breakthrough action. For the first time, a robot was
available for under $2,500 when other off-the-shelf
robots cost ten times as much.
Suddenly, universities everywhere could afford to begin teaching
robotics. Students at leading universities could actually use a robot
more than a few times a semester.
In addition, because hundreds were sold,ActivMedia could afford to begin
doing some of the integration work. Prior to Pioneer, every time someone
wanted to run a pan-tilt-zoom camera on their robot, they had to write a
driver to operate the camera and perhaps even build the cables
themselves.ActivMedia was the first company to begin integrating a wide
range of accessories with a commercial robot, a move that make their
robots the versatile plug-n-play systems they are today.
7
13
HILARE and CART/Rover
 HILARE
 LAAS in Toulouse, France (late 1970s)
 Video, ultrasound, laser range-finder
 Still in use!
 Multi-level spatial representations
 Deliberative -> Hybrid Control
CART/Rover
 Hans Moravec
 Stanford Cart (1977) /CMU rover (1983)
 Sonar and vision
 Deliberative control
14
HILARE
is a modular, triangular, and computer-controlled mobile
cart
equipped with
three wheels (two of them motor-driven), an onboard
microcomputer,
a sophisticated sensor bank (vision, infrared, ultrasonic
sonar/proximity, and telemetry laser), and
a manipulator arm was added in 1980.
HILARE's control systems include "expert modules" for
object identification, navigation, exploration, itinerary
planning, and sensory planning.
8
15
HILARE
Characteristics:
Wheels:2 driving wheels and a free wheel
Batteries:24V
Bus: Multibus
Processors: 4 x Intel 80286
Operating system: none
Communication: serial radio modem
(9600 bauds)
Sensors: Odometer,
16 UltraSonic sensors,
a Laser Range Finder
Dimensions (LxWxH): 80cm x 80cm x 60 cm
Weight: 400kg
Built in 1977 at LAAS
16
A new platform (at LAAS), it's a XR4000, built by Nomandic
Technologies. All the control architecture from LAAS has been ported to
Linux and currently runs on the robot.
Characteristics:
Wheels:4 driving wheels providing
holonomic motion
Batteries:4x12V
Bus:2 PCI Busses
Processors:2 x Intel Pentium II
Operating system: Linux
Communication: Ethernet radio modem (3 Mbit/s)
Dimensions (Diameter x H):62cm x 95cm
Sensors:odometry,
48 US,
48 Infrared sensors,
2D Laser Range Finder,
2B&W cameras on a pan & tilt
platform
9
17
Robotics Today
 Assembly and manufacturing (most numbers of
robots, least autonomous)
 Materials handling
 Guards (hospitals, security)
 Hazardous environments (Chernobyl)
 Remote environments (Pathfinder)
 Surgery (brain, orthopedic: hips)
 Tele-presence and virtual reality
 Entertainment
18
IS Robotics
Ariel:Autonomous Legged
Underwater Vehicle
Modeled after a crab, Ariel is designed to remove
mines and obstacles on land and underwater
The All-Terrain Robot Line
10
19
20
Robotic Student Competitions
www.robocup.org/
8
th
INTELLIGENT
GROUND VEHICLE
COMPETITION
July 8-10, 2000 Walt Disney
World's Coronado Springs Resort,
Orlando, Florida
LEGO KIT
RoboSoccer
Problem solving in complex domains often
involves multiple agents, dynamic
environments, and the need for learning from
feedback and previous experience. Robotic
soccer is an example of such complex tasks
for which multiple agents need to collaborate
in an adversarial environment to achieve
specific objectives. Robotic soccer offers a
challenging research domain to investigate a
large spectrum of issues of relevance to the
development of complete autonomous agents.
11
21
CMU Pioneer
Semi Autonomous random
patrol and surveillance vehicle
22
RIMCC
Robotics and Intelligent Machines Cooperative Council
chartered by Robotic Industries Association and the IEEE
Robotics and Automation Society.
developed the recommendations of the National Science
Foundation/Department of Energy Workshop on Research
Needs in Robotics and Intelligent Machines for Emerging
Industrial and Service Applications.
for simplicity's sake the meeting is called the Needs Workshop.
The workshop was held at Sandia National Lab. in Albuquerque,
New Mexico in October, 1996.
12
23
RIMCC Workshop
WHY ARE INTELLIGENT MACHINES
IMPORTANT TO THE U.S.?
"The coupling of machines with sensors, software
and computers - intelligent machines - promises a
revolution as profound as the computer
revolution itself.
Prof.Bekey from University of Southern California
24
Revolution Features
Transfer space/ defense oriented expensive
robotic and telerobotic technologies to the
consumer products.
We have to keep the system inexpensive
and easy to use in order to enable the
widely use in consumer market products
,
which will bring the "revolution".
13
25
Human Operator in the loop
Fully autonomous systems are not always needed
even for hazardous environments like military,
waste disposal, mining or space.
Manufacturing may sometime be hazardous, but
automation is justified for enhanced quality or
higher production rates.
The most one may really need is to have the
Human Operator in the loop to serve as the highest
control level.
26
Mixed Initiative Systems
Complex AI systems typically cannot tolerate
human intervention in details of their solutions.
Human users, should be able to impose constraints
on the automated task's actions or propose a partial
solution.
The joint efforts of humans and automation are to
be organized around a shared task model that is
understandable and performable by both users and
automation.
14
27
Concept
Penetrate denied areas and project operational
influence in ways that humans cannot by using
reliable semi-autonomous robotic platforms.
Tactical Mobile Robotics - DARPA
28
Approach
Integrate sensors, locomotion, power, communications,
and sufficient smarts on a compact, man-portable
platform to provide a semi-autonomous system capable
of penetrating into denied areas and serving as an
extension of the human soldier.
Tactical Mobile Robotics - DARPA
Top Technical Challenges
Robotic mobility in cluttered and complex terrain
Machine perception for obstacle negotiation
Autonomous operation and fault recovery
15
29
Tactical Mobile Robotics - DARPA
Users
Special-operations
forces
Early-entry forces
Marine Corps
Goals
FY99: Mature enabling technologies for
machine perception and mobility
FY00: Demonstrate autonomous fault
recovery
FY01: Conduct operational demonstrations
with integrated systems
30
Braitenberg Vehicles
Asimov
AI)
Flakey28
Erratic 29
Pioneer 1 31
HILARE and CART/Rover31-33
XR4000, built by Nomandic34
IS Robotics & WRII37-38
Robotics and Intelligent Machines Cooperative Council
44-45
TMRDARPA