SBIT JOURNAL OF SCIENCES AND TECHNOLOGY ISSN 2277-8764

VOL-2, ISSUE 1, 2013.

AN OVERVIEW OF ARTIFICIAL
INTELLIGENCE

1 2 3
Gyanendra Singh Ajitanshu Mishra Dheeraj Sagar

1,2,3
Assistant Professor, Mechanical Engineering Department,
Invertis University Bareilly
E-mail:[email protected]

Abstract: The recent growth in computer architecture has changed the face of science and engineering. This growth is so
fundamental that it is dramatically reshaping relationships among people and organizations and providing a
foundation for understanding and learning of intelligent behavior in living and engineered systems. Is this growth
beneficial to our society and how the medical facilities, police departments, and manufacturing plants have all been
changed by this, are such questions of the general public which are due to the lack of education concerning rapidly
advancing technologies.
This paper attempts to present an overview of Artificial Intelligence (AI). A generally accepted theory that
“machine will do and think like humans more in the future” is the concept behind AI. Brief literature of different
aspects by which AI is achieved like expert system, knowledge based systems (knowledge engineering), neural networks,
fuzzy logic, neuro fuzzy logic and fuzzy expert system, is included in order to have a clear understanding of AI. Along with
the different applications of AI, application of expert system to solve the design problem of mechanical spring is also
included in this paper.
It is concluded that extensive ongoing research in the field of AI gives an idea that in near future a day will
come when human beings and machines will merge into cyborgs or cybernetic organisms that are more capable and
powerful than either. This idea is called transhumanism

Key words: Artificial Intelligence, Expert System, Neural Network, Fuzzy Logic

1. INTRODUCTION i.e. assertions whose truth-value is independent of the

algorithmic context. They can induce, deduct and
Artificial Intelligence term was coined by sometimes guess data. They can reconsider decisions
John McCarthy in 1956. He defined it as "the science by employing back tracking for solutions.
and engineering of making intelligent machines.” AI is
the branch of computer science which deals with the 3. COMPONENTS OF AI
study and design of intelligent agents that perceives its
environment and takes actions which maximize its AI has four main components
chances of success. AI may be defined as: “The ability 1. Expert systems:
to hold two different ideas in mind at the same time 2. Heuristic problem solving
and still remain the ability to fucntion”. But AI must 3. Natural Language Processing
include the learning from past experience, reasoning 4. Vision
for the decision making, inference power and quick Expert system handles the situation as an expert and
response. Also it must be able to take decisions on the gives performance. Heuristic problem solving is meant
basis of priorities and tackle complexity and ambiguity. to evaluate small range of solutions, may involve some
Machines programmed to carry out tasks, when carried guesswork to find near optimal solution. Natural
out by humans would require intelligence, are said to language processing provides communication between
possess artificial intelligence. AI's scientific goal is to human and machine in natural language. Vision is the
understand intelligence by building computer programs ability to recognize shapes and features etc.
that exhibit intelligent behavior by using symbolic automatically.
inference, or reasoning inide the machine. AI definition
is not time-independent. It gives the judgment of any
4. EXPERT SYSTEM
system by keeping time in mind.
An expert system is a machine system in
2. FEATURES OF AI PROGRAMMING which useful human knowledge is added in machine
memory in order to give intelligent advice and offer
According to Brook ordinary programming explanations and justifications of its decisions or
languages don’t have the abilities to deal with demand. Expert systems relies on a large database of
qualitative information. So, the AI machines, are well defined specialized knowledge about a particular
programmed to work with their own developed area. Construction of such programs is refered to as
programming language to manipulate knowledge more Knowledge Engineering. All such AI programs that
effectively. AI programs are different from ordinary achieve expert-level competence in solving problems
programming languages. They are written to in task areas by using knowledge about specific tasks
manipulate predominantly qualitative rather than are called knowledge-based Systems or expert systems.
numeric information. They use declarative knowledge, These programs contains the knowledge used by
1
human experts, in contrast to knowledge gathered from b) Rule Base consists of constraints like length, size limits on
textbooks. Because of this expert systems are like human outer diameter, inner diameter, cost of material, etc which
experts e.g. doctors, engineers, analysts, teachers, are provided by user. Inference engine is used to test
geologists etc which encapsulate the skills of an expert and to the design under these constraints.
dispense advice to less knowledgeable users. This transfer of c) Data Base database consists of around 700
knowledge depends upon the task and will take place compression designs. When the design completes, the
gradually through many interactions between expert and the inference engine compares this design with these standard
system. compression designs to find the optimal solution. The
It is easier to build expert system than ones with results are reported to user.
common sense. They represent task domain. Task means some d) User Interface is there throughout the designing process.
goal-oriented, problem-solving activity and domain refers to The expert system asks the user for any additional input like
the area within which the task is being performed. One of any space limitations, any material specification, any load
the earliest expert system MACSYMA which performed a requirements, any solid height limitations etc.
variety of symbolic mathematical tasks, was composed of a After the completion of process by expert system final
set of fairly unstructured LISP functions. There are many selection is left to the user on the basis of its application.
expert systems exists which have been designed for giving
expertise training, designing and trouble-shooting etc. like 5. FUZZY LOGIC
MYCIN, TURNX, PROSPECTOR. The expert systems are
still in their infancy. It was introduced by Dr. Lotfi Zadeh of
UC/Berkeley known as father of fuzzy set theory, in the
4.1 Benefits of Expert Systems 1960's as a means to model the uncertainty of natural
Following are the benefits of expert system language. Lofti A.Zadeh. Fuzzy Logic are used as a profitable
a) Expert systems have proved to do a better job than tool to control the subway systems, complex industrial
humans. They make fewer mistakes and more processes, entertainment electronics,
consistent in their recommendations. diagnosis systems and household
b) Artificial Expertise is usually cheaper compared to appliances e.g. in washing machines fuzzy logic sense load size
human expertise. and detergent concentration and adjust their wash cycles
c) They achieved a notable success in the field of automatically. Fuzzy logic is very useful in manufacturing
training, to train non-experts and even to improve processes as it can handle situations that can not be
expertise of expert. adequately handled by traditional true/false logic. Fuzzy
d) They can handle the mechanical type of repetitive Logic was invented in the United States and rapid growth of
tasks of experts, so that experts can well concentrate this technology has started from Japan.
on their unique skills in given domain. Fuzzy logic is a mathematical approach to problem
e) They are compatible with many manager’s solving. It is very powerful method of reasoning when there
decision styles. is no simple mathematical model and input data are imprecise
f) They can enable operations in environment not for very complex processes and highly nonlinear processes. It
suitable for humans. produces exact results from imprecise data, and is especially
g) They improve the productivity of industry. useful in computers and electronic applications. Fuzzy logic
differs from classical logic in that statements are no longer true
4.2 Expert System Applied to Design of Spring or false, on or off. In traditional logic an object takes on a
An expert systems are able to make flexible value of either zero or one but in fuzzy logic, an object can
programs by using qualitative and quantitative data to solve a assume any real value between 0 and 1. The human brain can
wide range of design problems. An expert system can be reason with uncertainties and judgments. Computers can only
used to design the spring. It requires minimum input from manipulate precise valuations. Fuzzy logic is an attempt to
user like one load and length of spring. Any additional combine the two techniques." Fuzzy logic performs better
information like second load at different length of spring rate, when compared to conventional control mechanism like PID.
spring inner diameter and outer diameter limits can also be According to Nagrath and Gopal Fuzzy Logic is only a
specified if required because this additional information small part of the logics available to AI. Fuzzy Logic is
reduces the guesswork of expert system. Knowledge Base, basically a multivalued logic derived from fuzzy set theory
Rule Base and Data Base are used by the user while to deal with reasoning that is approximate rather than precise.
designing the spring. Intermediate values between conventional evaluations like
a) The Knowledge Base consists of following two yes/no, true/false, black/white, etc. can be formulated
equations: mathematically and processed by computers. In this
K= Gd4/ (8D3N) and T= 8Ks FD/ (πd3),
Here F represents load on the spring, T represents shear stress,
K s represents shear stress concentration factor, D represents
mean coil data, D represents wire diameter, G represents
shear modulus and N represents number of active coils.
2
way an attempt is made to apply a more human-like way of following:
thinking in the programming of computers. if x is low and y is high then z = medium
5.1 Fuzzy Expert System where x and y are input variables (names for know data values), z
A fuzzy expert system uses fuzzy logic instead is an output variable (a name for a data value to be computed),
of Boolean logic. In other words, a fuzzy expert system is a low is a membership function (fuzzy subset) defined on x, high
collection of membership functions and rules that are used to is a membership function defined on y, and medium is a
reason about data. Fuzzy expert systems are oriented toward membership function defined on z. The part of the rule between
numerical processing whereas conventional expert systems the "if" and "then" is the rule's _premise_ or _antecedent_. The
are mainly symbolic reasoning engines. The rules in a part of the rule following the "then" is the rule's
fuzzy expert system are usually of a form similar to the _conclusion_ or _consequent_. This part of the rule assigns a
membership function to each of one or more output variables. first proposed in 1943 by Warren Mc Culloch, a
Fuzzy expert systems can have more than one conclusion per neurophysiologist, and Walter Pitts, an MIT logician.
rule and can also have more than one rule. The group of rules is Computers can be operated in nanoseconds, and work
collectively known as knowledge base. In a fuzzy expert without error. But it can’t do walking, talking and reasoning like
system, the inference process is a combination of human being. ANN are an attempt to emulate (very roughly)
four subprocesses: fuzzification, inference, composition, and the basic functions of the human brain to perform complex
defuzzification. The defuzzification subprocess is optional. In functions that everyday computer systems are incapable of
fuzzification subprocess, the membership functions defined on doing. The human brain is a naturally occurring model of
the input variables are applied to their actual values, to Neural network. So, idea is to simulate functioning of the brain
determine the degree of truth for each rule premise. In the directly on a computer and thus develop artificial neurons.
inference subprocess, the truth value for the premise of each Neural Networks are not meant to duplicate the human brain,
rule is computed, and applied to the conclusion part of each but to receive information about how the brain works.
rule. This results in one fuzzy subset to be assigned to each An ANN involves a network of simple processing elements
output variable for each rule. In the composition sub process, (artificial neurons) which can exhibit complex global behavior,
all of the fuzzy subsets assigned to each output variable are determined by the connections between the processing elements and
combined together to form a single fuzzy subset for each element parameters whereas Biological neural networks are
output variable. In defuzzification subprocess fuzzy value is made up of real biological neurons that are connected or
converted to a single number crisp value. functionally related in the peripheral nervous system or the
central nervous system.
5.2 Neuro-fuzzy Logic In a neural network model a large number of very simple
They use combination of fuzzy logic and neural neurons like nodes of processing elements are connected
networks. Fuzzy logic uses approximate human reasoning in together with a large number of weighted connections
knowledge-based systems while the neural networks aim at between these elements which are highly parallel and
pattern recognition, optimization and decision making. distributed. Neurons are extremely slow, operate in
According to Arbib the explicit knowledge representation of milliseconds and yet humans can perform extremely complex
fuzzy logic is augmented by the learning power of simulated tasks in just a tenth of a second because brain contains a
neural networks. This combination of two technological huge number of processing elements that act in parallel.
innovations delivers the best results and has led to a new According to Brook ANN are used to solve artificial
science called neuro-fuzzy logic. intelligence problems by using algorithms designed to alter
the strength of the connections in the network to produce a
6. ARTIFICIAL NEURAL NETWORKS (ANN) desired signal flow.
ANN can recognize something it has never seen before
The concept of a neural network appears to have and predict the future, by extracting patterns in the past.
first been proposed by Alan Turing in his 1948 paper Application areas of ANN also include the system
"Intelligent Machinery". Artificial neurons were identification and control (vehicle control, process control),
function approximation or regrassion analysis, time series
prediction, modeling game- playing, sequential decision
making (chess, racing), pattern recognition (radar systems, face
identification, object recognition, etc.), sequence recognition
(gesture, speech, handwritten text recognition), medical
diagnosis, financial applications, data processing
(including filtering, clustering), knowledge discovery in
databases (KDD), visualization and e-mail spam filtering.

7. APPLICATION OF AI

a) To design and analyze the mechanical elements

on basis of size limitations.
b) To diagnose electronics locomotion systems.
c) Can be used in electronics and
electrochemical systems.
d) To diagnose the software development process.
e) To identify chemical compound structures and
chemical compounds.
f) Can be used in medical diagnosis.

3
g) To plan experiments in biology, chemistry and common people life and is limited to some areas like military,
molecular genetics. space, industry, medical, neutral networks and geological.
h) To make stock and bond portfolio for selection It may be expected that at the end of 2035 with the
and management. extensive research and advancement in the field of AI, we will
i) Trouble shooting systems be able to move away from today’s machinery that necessarily
j) To plan and explore the space. come with weighty manuals regarding machine languages
k) To forecast crop damage and develop the machinery which will be able to
l) To develop completely automated plants and understand human completely. We will have robot as doctor in
industries. hospitals, professor in class room, driver in bus. According to

8. CONCLUSION

Till now AI has not such a great effect directly on

Bostrom that will be the era of transhumanism where human Theory and Neural Networks”, p. 666
beings and machines will merge into cyborgs or cybernetic Bostrom, Nick (2005) "A history of transhumanist thought" ,
organisms that are more capable and powerful than Journal of Evolution and Technology, p. 2-21
either. Brook R.A. “Intelligence without representation- Artificial
Intelligence” 1991,p.139-159
Nagrath I.D. and Gopal M. “Control Systems
Engineering” New Age International Publications.
9. REFERENCES PC "Fuzzy Logic and Neural Networks - Practical Tools
for Process Management", May/June, 1994, p.17
Arbib, Michael A. (Ed.) (1995), “The Handbook of Brain Surinder Kumar and Jha A.K. “CAD/CAM”