Download here - KB e-learning Site for IB ITGS and IGCSE ICT

Sandford International School – IT Dept – IB ITGS Notes 2013 ©
3.11: Robotics, Artificial intelligence & Expert systems (Chap 15 pg 534-560
Tomorrows Technology & You 8th ed)
Artificial Intelligence
By definition, AI is the Behavior by a machine that, if performed by a human being, would be called
intelligent. It can also be defined as “The study of how to make computers do things at which, at the
moment, people are better at”.
AI also involves studying the thought processes of humans (to understand what intelligence is); secondly,
it deals with representing those processes via machines (computers, robots, etc.).
The emergence of AI as a viable commercial technology can be attributed to the development of
semiconductor devices and enhanced computer architectures. AI takes a lot of memory, but memory is now
inexpensive.
AI differs from traditional computer software engineering and simple automation in several ways.
The latter produces predefined actions according to preprogrammed instructions. In contrast, the actions of
an AI system may differ from situation to situation in a way not predefined but determined from a collection
of facts, assumptions, and general procedures for decision-making – much as a human would act.
Turing Test
Turing's Prediction disapproved
Alan Turing predicted in 1950 that in about fifty years "an average interrogator will not have more than a 70
percent chance of making the right identification after five minutes of questioning" in a version of the
imitation game (the Turing Test) in which a computer imitates a human. – for more of this and CAPTCHA
(Completely Automated Public Turing test to tell Computers and Humans Apart) see :


http://www.wisegeek.com/what-is-the-turing-test.htm
http://www.howstuffworks.com/captcha1.htm
Artificial Vs Natural (Human) Intelligence
The potential value of artificial intelligence can be better understood by contrasting it with natural, or
human, intelligence. AI has several commercial advantages:
1. AI is permanent. Natural intelligence is perishable from a commercial standpoint in that workers
can change their place of employment or forget information. AI, however, is as permanent as
computer systems and programs.
2. AI offers case duplication. Transferring a body of knowledge from one person to another usually
requires a lengthy process of apprenticeship; even so, expertise can never be completely
duplicated. However, when knowledge is embodied in a computer system, it can be copied from
that computer and easily moved to another computer, sometimes across the globe.
3. AI can be less expensive than natural intelligence. There are many circumstances in which buying
a computer of any size costs less than having human power carry out the same tasks.
4. AI being a computer technology is consistent and thorough. Natural intelligence is erratic
because people are erratic; they do not perform consistently.
5. AI can be documented unlike natural intelligence.
1
Sandford International School – IT Dept – IB ITGS Notes 2013 ©
Some advantages of Natural Intelligence are as follows:
1. Natural intelligence is creative, whereas AI is rather uninspired. The ability to acquire knowledge is
inherent in human beings, but with AI, tailored knowledge must be built into a carefully constructed
system.
2. Natural intelligence enables people to benefit and use sensory experience directly, whereas AI
systems must work with symbolic input.
3. Human reasoning is able to make use at all times of a very wide context of experience and bring
that to bear on individual problems, whereas AI systems typically gain their power by having a very
narrow focus.
Categories of Artificial Intelligence
Research in this field can be divided into the following categories:
1. Knowledge based and Expert systems,
2. Neural Networks (Simulation of Human Sensory capabilities)
3. Robotics.
3.11.1. Knowledge based and Expert Systems
A knowledge-based system relies on a knowledge base that is filled with “rules” of thumb (intuition,
judgment, and inferences) about a specific application area to solve a particular problem.
Expert systems are the most sophisticated implementation of a knowledge –based system. Once
the knowledge of one or more human experts has been entered to an expert system’s knowledge base,
users can tap this knowledge by interacting with the system in much the same way they would interact with
a human expert in the field. Many expert systems are based on fuzzy logic. Fuzzy logic allows conclusions
to be stated as probabilities (e.g there is a 70% chance …..) rather than certainties.
Exercise 3.11.1: AI - HOME WORK 1:
A situation in which fuzzy logic can be used is demonstrated at the following
website.
Visit http://people.clarkson.edu/~esazonov/neural_fuzzy/loadsway/LoadSway.htm
(Your computer must be flash and Java enabled for this to work!)
Try both manual and fuzzy control of getting the load from the ship to the dock. These
controls can be selected by clicking on the buttons underneath the simulation of the ship and
dock.
Manual controls: Arrow Right - increase the power; Arrow Left - decrease the power;
Space - stop the crane; Arrow Up - lift the load; Arrow Down - lower the load;
Enter - lock the load.
2
Sandford International School – IT Dept – IB ITGS Notes 2013 ©
1. Explain which method causes the least sway (movement forwards and
backwards) at the fastest speed of the box being loaded/unloaded from the
ship. (2 marks)
2. What kind of knowledge (linguistic variables) is needed here? Hint: have a
look at the boxes and graph underneath the simulation. (3 marks)
3. Explain the kind of inference rules might be used here? Hint: look at the
text box. (2 marks)
4.
Explain why fuzzy logic may be used for this situation. (3 marks)
3.11.2. Neural Networks (Simulation of Human Sensory capabilities)
Neural networks is a section of AI that deals with specialized software that allows a system to learn to
recognize features or characteristics of situations which are input in it. The technique is based on a model
of logical properties of interconnected ‘nerve cells’ in a network. The network, called neural net, is made up
of very many junctions and nodes. Each of these nodes continuously learn features according to input.
Once the learning phase is completed, the neural net can be used to recognize features of the same type
that were originally presented to it.
Neural nets have been used for visual recognition and also for financial prediction.
3
Sandford International School – IT Dept – IB ITGS Notes 2013 ©
3.11.3. Robotics
Robotics is the study and design of robots.
Def 1: A robot is a computer controlled mechanical device
which is sufficiently flexible to be able to do a variety of
tasks.
Def 2:A robot is a machine that looks like a human being
and has the capacity to perform human tasks. Robotics is
one of the fastest growing areas of artificial intelligence.
Honda Asimo robot
Not all robots need to look human... See
the images below
The 'Dalek' circa 1964
'Aibo' robot pet.
"Blue Max" BBC Robot Wars
All robots don’t look the same, but they share one thing in common:
They are able to take in information from their surroundings and make decisions based on that information.
Robots need information as well as programming in order to function. This information is provided by
sensors. The types of sensors required by robots include:
•
•
•
•
•
•
•
•
•
Mercury tilt switches – these are switches that sense titling movement.
Light sensors – these are sensors that detect different levels of light and dark.
Push switches – these are switches that operate when something is opened or closed (e.g. the switch that
controls the light in a fridge is a push switch).
Temperature sensors – these detect changes in the temperature.
Sound sensors – these detect different levels and frequencies of sound.
Proximity sensors – these detect how close something is or if something is being moved away.
Position sensors – these detect the angle between one part of the robot arm from another.
pH sensors – these detect how acidic something is.
Humidity sensors – these detect how much moisture is present in the atmosphere or in a sample.
4
Sandford International School – IT Dept – IB ITGS Notes 2013 ©
Robots combine sensory systems with mechanical motion to produce machines of widely varying
intelligence and abilities. The difference between the automatic machine and an intelligent robot is that the
robot senses its environment and modifies its behavior as a result of the information gained. The intelligent
robot is thought to have humanlike capabilities and attributes. For example, the robots are distinguished
from regular automation by their ability to deal with uncertainty. The intelligent robot flexibly connects
perception to action. Humans can be viewed as super intelligent robots because they can see and feel,
have graceful hands for fast, accurate motion, etc.
WHERE USED
Robots are used day after day in factories, while others
are highly experimental and use artificial intelligence to
behave more and more like living creatures.
Some robots are being designed to perform surgery that
requires great precision, explore space, the ocean, other
dangerous areas.
What are the advantages of Robots over Humans ?
1. More reliable work, 95 percent uptime, works three
shifts, no vacations, and no coffee breaks.
2. Work at lower cost, no workman's compensation, no
retirement benefits, and no insurance plans.
3. Environmentally insensitive i.e can work in poor
climatic conditions, in high noise environs, in hazardous location e.t.c.
4. More consistent, accurate work, improve quality, reduce scrap, production predictable, efficient use of
capital, self-diagnosis of errors, improves work methods.
5. Work only for the organization, no moonlighting, no espionage for competition, no Sabotage, no
theft, no feather-bedding.
Rudimentary Robot
This is a single mechanical arm controlled by a computer. The arm is called manipulator, and has a
shoulder, forearm, and wrist and is capable of performing motions of the human arm. An industrial robot is
best at repetitive work and especially tasks that require precision movements, moving heavy loads and
working in hazardous areas. Such tasks include
manufacturing, painting, welding, circuit testing,
connecting chips to circuit boards; surgeons are
using
robots to help in brain surgery. They can be set up
to
manipulate the surgical drill and biopsy needle
with
great accuracy, thereby making brain surgery
faster,
more accurate, and safer.
A computer program is written to control
the
robot just as one is written to print payroll cheque.
It
includes such commands as when to reach, in
which
direction, how far to reach when to grasp, and so
on.
Robots don’t need too much attention once
programmed.
Exercise 3.11.2: AI - HOME WORK 2:
Look up the following definition in wikipedia or any other useful website:
1. robot,
2. android,
3. cyborg,
4. sensors
5
Sandford International School – IT Dept – IB ITGS Notes 2013 ©
Exercise 3.11.3: AI- CLASSWORK
The chances of having an obedient robot to do unwelcome or dangerous jobs have increased
tremendously, with orders for industrial robots rising to a record 18 per cent in the first half of 2009. In
homes about 4 million autonomous vacuum cleaners and lawn-mowers are already in operation. Right now
Japan uses about 320 robots of all sorts per 10,000 employees, while Germany uses 148 industrial robots
per 10,000 employees. In the car industry alone, there is one robot per 10 workers in Japan, Italy and
Germany.
1. Humans have used tools and machines to assist them since the dawn of time. Distinguish between
a traditional machine and a robot. [4 marks]
2. Examine two benefits robots offer for the car industry. [6 marks]
3. Evaluate one social issue and one ethical issue of the increased use of robots in the home.
[8 marks]
4. Define A.I. [2 marks]
5. 2. Explain the Turning test [4 marks]
6. 3. Describe two extra features that would be needed for a robot used in surgery to be considered as
a medical expert system. [4 marks]
7. 4. Evaluate one social or ethical issue that the possible development of AI has for us. [6 marks]
References:
1. Beekman G. and Quin M, Tomorrows Technology & You 8th edition, Pearson Education
2. www.teach-ict.com (Accessed November 18th 2008)
3. http://senior.british.edu.uy/it_dept/itgs/The Future of the Turing Test.htm , (Accessed 25th April
2009).
4. http://www.clarkson.edu/~esazonov/neural_fuzzy/loadsway/LoadSway.htm , (Accessed 25th April
2009).
5. http://www.senri.ed.jp/site/ , (Accessed 25th April 2009).
6