National Conference on Recent Advances in Computer Science and IT (NCRACIT)

International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2018 IJSRCSEIT | Volume 4 | Issue 1 | ISSN : 2456-3307

Artificial Intelligence in Mechanical Engineering

Saquib Rouf1, Moonis Ali2 , Ali Hussain3
1 Department of Mechanical Engineering, SSM College of Engineering and Technology, Parihaspora, Pattan, Jammu & Kashmir, India
2,3Department of Computer Science & Engineering, University of Kashmir, Delina, Baramulla, Jammu & Kashmir, India

[email protected], [email protected], [email protected]

ABSTRACT

This paper is a survey based analysis about the applications of Artificial Intelligence (AI) in various domains
of mechanical engineering as well as its use in industry and fault diagnostic systems. This paper presents the
case for the use of AI instead of conventional techniques in industrial processes and fault diagnosis of
mechanical systems as they are time consuming and in some cases ineffective. The use of AI can result in
early detection and diagnosis of faults in mechanical systems which can increase the plant safety and can be
useful in optimising operational costs. This paper also lists the various uses of an AI driven Knowledge Based
System. At the end some diverse applications of AI in mechanical systems have been identified while
discussing their positive impact on the field.
Keywords: Artificial Intelligence, Mechanical Engineering, Fault Diagnosis, Knowledge Based System

I. INTRODUCTION initiating intelligent reasoning and problem-solving

power in machines is a hard and intricate task.
Artificial intelligence is an area of computer
science that deals with the simulation of human This is where a sub-domain of AI known as
intelligence processes by machines, especially Machine learning plays an important role. Learning
computer systems. An AI process consists of and acquisition of information without human
acquiring information (learning), deducing supervision requires a tremendous capability to
inferences (by applying logic to reach estimated or identify/classify and cluster patterns across diverse
definite conclusions), and implementing methods of streams of inputs, on the other hand the process of
self-correction over a period of time. Owing to the learning with adequate supervision involves
increasingly vast use of AI across various fields and classification and numerical regressions.
disciplines applications of AI may range from Classification recognises the category that an object
character recognition and machine vision to natural belongs to and regression deals with procuring a set
language processing (NLP) and robotics. Artificial of numerical input or output examples, thus enabling
intelligence based algorithms and systems must be the causation of suitable outputs by their respective
given access to various kinds of features, objects, inputs.
characteristics and relations between them so as to
make effective use information and knowledge Another sub-domain of AI, Machine perception
engineering techniques. However, the process of deals with the capability of a smart AI based system

CSEIT411836 | Published - 25 April 2018 | March-April-2018 [ (4 ) 1 : 217-220 ] 217

 in making use of sensory inputs to deduce diverse g) Case base.
inferences from different subjects of the real-world. h) Rule library.
Also in an AI system, the discipline of computer i) Data base.
vision is responsible to analyse visual-inputs such as j) Learning system.
facial, object and gesture recognition. Robotics is also Now the working of this system is explained as
a major field attributed to the advancements in AI. below:
Robots require intelligence to handle tasks such as
object manipulation and navigation, along with sub- When the user gives data through user - machine
problems of localization, motion planning and interface, It is acquired by the reasoning machine .
mapping. [1] This part of the system is programmed by efficient
algorithm and hence it provides an expert advice.
The main role of artificial intelligence in mechanical This input data is saved in the database as a history.
engineering is that it makes systems autonomous / Whenever the data is entered by the user, it is
automatic. The machines can be made automatic checked for a similarity in database in order to take
when they are designed to have ability to detect and decisions quickly.
diagnose errors in manufacturing system [2]. When
we have a combined system of mechanical and A. Intelligent Diagnosis System For Rotating
electronic components, in order to achieve efficiency Machinery
it is necessary for a system to transfer data quickly Consider an example of fan. Since fan is rotating
from one part of a system to another. This timely device therefore it needs a motor. The monitoring
transfer of data is done by artificial intelligence. At parameters are
present, AI is used for diagnosis of mechanical
engineering failure [3]. 1. Temperature.
2. Vibrations.
II. APPLICATION OF AI ACROSS VARIOUS 3. Noise.
METHODS OF DIAGNOSIS IN MECHANICAL 4. Oil.
SYSTEMS 5. Performance.

The main methods for fault diagnosis in Here universal integrated neural network
mechanical systems by artificial intelligence are: diagnosis system is used. The core of whole
intelligent system is fault diagnosis and decision
1. CBR (case based reasoning). system.
2. RBR (rule based reasoning).
3. FBTD (fault based tree diagnosis). B. Intelligent Diagnosis System For Reciprocating
Systems
Based on CBR and RBR an expert fault diagnosis Diesel engine is the typical reciprocating system.
system is prepared whose key members are as follows: The fault diagnosis can be done by integrated neural
network diagnosis system.
a) User.
b) Man machine interface. The performance faults of diesel engine can be
c) Knowledge acquisition system. achieved using a sub – neural network. The input
d) Reasoning machine. parameters are speed, power, pressure etc . The
e) Interpreter. mechanical fault diagnosis can be achieved by two
f) Knowledge process. sub neural networks. Here the integrated neural

Volume 4 | Issue 1 | March-April-2018 | www.ijsrcseit.com 218

 network diagnosis system is formed by vibro acoustic B. AI IN MANUFACTURING INDUSTRY
signal with oil analysis information as input. To maintain profitability and high productivity AI is
used in manufacturing industries. Machines working
III. AI PRACTICES IN MANUFACTURING with AI can meet the needs of industry and market
INDUSTRY AND APPLIED MECHANICAL with better efficiency and reduce the operational
ENGINEERING costs drastically.
Figure 2, shown below presents an overview to
A. ROLE OF AI IN MECHATRONICS understand the role of AI in manufacturing industry.
Mechatronics is a specialised discipline, which is the The interface engine interacts with the knowledge
combination of mechanical, electrical and electronics. base where the information is pooled along with
In Mechatronics, electrical engineering is associated relevant rules. Information from the knowledge base
with study of motors and power system while is applied to real world problems based on the expert
electronics is related with the study of boards and its system which plays an important role in executing
control engineering; the mechanical is associated relevant rules after scanning the information from
with the hardware apparatus and their controlled the human interface. Therefore, the AI-Mechanical
movement. The focus of Mechatronics is on robot system can be hugely beneficial in manufacturing
hardware and its working while AI takes care of industry as it has the power to learn by itself and
software at all level to make it operational and make amends to the existing rules with the passage
autonomous. It is a field of engineering focused on of time and human interaction.
the design and manufacturing of robots. Robots are
often used to perform tasks that are difficult for
humans to perform or perform consistently. They are
used in assembly lines for car production or by
NASA to move large objects in space. More recently,
researchers are using machine learning to build
robots that can interact in social settings [4].
Nowadays research is done in the field of
mechatronics which is specified just on safe and
intelligent robots. The methodologies in industrial Figure 2. Overview of application of AI in
robotics are based on Automatic control theory [5]. manufacturing industry.

C. AI Driven Decision Making In Knowledge Based

Mechanical Processes
An AI driven expert system incorporates the use of
Artificial Intelligence in smart decision making
which not only improves its performance with time
but also makes the system versatile as new rules can
be added and modified easily. Since the system is
quite dynamic, the languages used to program such
systems like LISP or PROLOG can consolidate new
changes as the system progresses with continuous
Figure 1. An Overview of Automatic Control Theory human interaction. The Knowledge Based Systems
in Mechatronics. engulf the events from the past and provide

Volume 4 | Issue 1 | March-April-2018 | www.ijsrcseit.com 219

 analytical, efficient solutions to real life problems. efficiency of these systems. With introduction of
Some advantages of KBS based AI system are: targeted AI approaches for specific sector based
1. It reduces risk in a mechanical system i. e it acts requirements in manufacturing and robotics the
as risk manager. scope for this field in the future seems ever more
2. It reduces cost of mechanical process by using promising.
optimisation techniques.
3. It increases the overall life of an industrial plant V. REFERENCES
by automation of safety systems.
It can be used for fault analysis and failure [1] Production technology by R. k Jain .
diagnostics. [2] https://www.techopedia.com/definition/190/artificial-
4. Assessment of surface quality during machining intelligence-ai
of metals can be easily achieved using AI [3] H Yang, J Mathew, L Ma. Intelligent diagnosis of
rotating machinery faults - A review [M]. Pattern
recognition techniques.
Recognition & Data Mining, 2002.
5. It can used for sensor monitoring of tools in
[4] http://searchcio.techtarget.com/definition/AI
unmanned machining process [6].
[5] L. Sciavicco and B. Siciliano. Modelling and Control
6. It can be also used in checking the thermal error of Robot Manipulators, 2nd ed. Springer-Verlag,
of machines by using Supervised machine Berlin Heidelberg, 2000
learning techniques. [7]. [6] R. Teti, K. Jemielniak, G. O. Donnell , D. Dornfeld
7. Optimization /planning of metal cutting process ,Advanced monitoring of machining ,CIRP Annals-
can be achieved by AI [8] Manufacturing Technoperationsology, 59 (2010) pp.
717-739
[7] J. Mayr,J. Jedrzejewski,etal, Thermal issues in
machine Tool, CIRP Annals-Manufacturing
Technology,61(2012),771-791.
[8] I. Mukherjee, P. K. Ray, A review of optimization
techniques in metal cutting processes, Computers &
Industrial Engineering 50 (2006) 15–34

Figure 3. Block diagram of a Knowledge Based

System

IV.CONCLUSION

The use of various AI applications and techniques for

automation of manual tasks, use of smart methods for
diagnostics, pattern recognition and intelligent
decision making processes has hugely benefited
manufacturing, industrial robotics, as well as fault
diagnosis systems amid other allied branches of
mechanical engineering. Use of AI not only speeds
up conventional mechanical processes but also
results in drastic reduction of costs and a surge in

Volume 4 | Issue 1 | March-April-2018 | www.ijsrcseit.com 220