World Academy of Science, Engineering and Technology

International Journal of Mechanical, Aerospace, Industrial, Mechatronic and Manufacturing Engineering Vol:1, No:12, 2007

An Expert System for Car Failure Diagnosis

Ahmad T. Al-Taani

International Science Index, Mechanical and Mechatronics Engineering Vol:1, No:12, 2007 waset.org/Publication/161

AbstractCar failure detection is a complicated process and

requires high level of expertise. Any attempt of developing an expert

system dealing with car failure detection has to overcome various
difficulties. This paper describes a proposed knowledge-based
system for car failure detection. The paper explains the need for an
expert system and the some issues on developing knowledge-based
systems, the car failure detection process and the difficulties involved
in developing the system. The system structure and its components
and their functions are described. The system has about 150 rules for
different types of failures and causes. It can detect over 100 types of
failures. The system has been tested and gave promising results.

KeywordsExpert system, car failure diagnosis, knowledgebased system, CLIPS.

I. INTRODUCTION

XPERT systems (ES) are a branch of artificial intelligence

(AI), and were developed by the AI community in the
mid-1960s. An expert system can be defined as "an intelligent
computer program that uses knowledge and inference
procedures to solve problems that are difficult enough to
require significant human expertise for their solutions [1]".
We can infer from this definition that expertise can be
transferred from a human to a computer and then stored in the
computer in a suitable form that users can call upon the
computer for specific advice as needed. Then the system can
make inferences and arrive at a specific conclusion to give
advices and explains, if necessary, the logic behind the advice.
ES provide powerful and flexible means for obtaining
solutions to a variety of problems that often cannot be dealt
with by other, more traditional and orthodox methods [2]. The
terms expert system and knowledge-based system (KBS) are
often used synonymously. The four main components of KBS
are: a knowledge base, an inference engine, a knowledge
engineering tool, and a specific user interface. Some of KBS
important applications include the following: medical
treatment, engineering failure analysis, decision support,
knowledge representation, climate forecasting, decision
making and learning, and chemical process controlling [2].
Previous work has shown that systems concerned with car
fault detection were very limited. Jeff Pepper [3] has
described a proposed expert system for car fault diagnosis
called SBDS, the Service Bay Diagnostic System. SBDS is
being developed by a joint project team at Ford Motor
Company, the Carnegie Group, and Hewlett Packard. SBDS's
Ahmad T. Al-Taani is with Faculty of Information Technology,
Department of Computer Sciences, Yarmouk University, Irbid, Jordan.
(phone:
+962-7-77438520;
fax:
+962-2-7211128;
e-mail:
[email protected]).

International Scholarly and Scientific Research & Innovation 1(12) 2007

knowledge base will contain the expertise of Ford's top

diagnosticians, and it will make their diagnostic skills
available to mechanics in every Ford dealership in North
America. This system will guide a human technician through
the entire service process, from the initial customer interview
at the service desk to the diagnosis and repair of the car in the
garage [3].
There are a lot of related expert systems in the literature
concerned with diagnostic problems. Daoliang et al. [4]
presents a web-based expert system for fish disease diagnosis.
The system is now is use by fish farmers in the North China
region. Yu Qian et al. [5] proposed an expert system for realtime failure diagnosis of complex chemical processes. Other
diagnosis systems are described in [6-9].
A knowledge-based system for car failure diagnosis is
presented in this paper. Section 2 gives a brief description of
the problem domain. Section 3 presents the proposed system.
In section 4, I will give some discussion and finally I will
draw some conclusions.
II. PROBLEM IDENTIFICATION
The proposed system divides car failures into three major
types:
1. Start-up state, problems that may occur when a person try
to start up the car, for example; engine does not work,
some sounds noticed, engine works ones and stops. These
problems could be due to one or more failures; will
happen, the battery needs to be recharged, the dynamo is
dead, or the battery is dead.
2. Run-stable state, problems that may occur after starting
the car, for example; unburned fuel, cycle on-off, blue gas
emitted, avance is very high.
3. Movement-state, problems that may occur while the car is
moving; this state also includes problems related to the
brake system. Most of movement problems that might
occur appears on the cars tableau the proposed system
takes advantage of these facilities and use them to
diagnose the problem and to gives advice of the solution
to the driver. Some of these problems are: oil pressure,
water temperature, voltage, and fuel pointer.
If the car is in the start-up state and doesn't start, then the
cause could be one of three main reasons: a bad fuel mix, lack
of compression or lack of spark. In addition, thousands of
minor things can create problems, but these are the main three.
Bad fuel mix: A bad fuel mix can occur in several ways:
1. The car ran out of gas, so the engine is getting air but no
fuel.

740

scholar.waset.org/1999.8/161

World Academy of Science, Engineering and Technology

International Journal of Mechanical, Aerospace, Industrial, Mechatronic and Manufacturing Engineering Vol:1, No:12, 2007

1. If the spark plug or the wire leading to it is worn out, the

spark will be weak.
2. If the wire is cut or missing, or if the system that sends a
spark down the wire is not working properly, there will be
no spark.
3. If the spark occurs either too early or too late in the cycle
(i.e. if the ignition timing is off), the fuel will not ignite at
the right time, and this can cause all sorts of problems.

Agenda

Working Memory

Fig. 1 Structure of the car failure diagnosis system

A. User Interface
Communication between the user and the system is done
through the user interface which implemented in both Arabic
and English languages. The user interface is represented as a
menu which displays the questions to the user and the user
answers with Yes or No. When the system is started a main
menu is displayed on the screen which asks the user to choose
one of the three car states (Fig. 2).
Car Failure Diagnosis Expert System
1. Start-Up state.

Other Problems: The following problems are also taken

into consideration in the system.

2. Run-Stable state.

1.
2.

4. Exit the program.

3.
4.
5.

If the battery is dead, the engine cannot turn over.

If the bearings that allow the crankshaft to turn freely are
worn out, the crankshaft cannot turn so the engine cannot
run.
If the valves do not open and close at the right time or at
all, air cannot get in and exhaust cannot get out, so the
engine cannot run.
If someone sticks a potato up your tailpipe, exhaust
cannot exit the cylinder so the engine will not run.
If you run out of oil, the piston cannot move up and down
freely in the cylinder, and the engine will seize.

In addition to the above problems, the proposed method

also has dealt with problems that may occur in the following
systems: cooling system, air intake system, starting system,
lubrication system, fuel system, exhaust system, emission
control, and electrical system.

3. Movement state.

Enter your selection according to the car state: ---Fig. 2 Start up menu of then system

B. Explanation Facility
Illustrates to the user how and why the system gave a
certain cause for the failure, i.e. explains the reasoning of the
system to the user.
C. Knowledge Base
The knowledge of the system is collected from mechanic
experts, specialized books, and from different car websites
[13-26]. The knowledge base contains about 150 production
rules for different types of car failures and causes.
As a rule-based shell, CLIPS stores the knowledge in rules,
which are logic-based structures, as shown in Fig. 3.

III. DESIGN AND IMPLEMENTATION OF THE KBS

The KBS developed in this work consists of the user
interface, the explanation facility, the knowledge base, and the
inference engine. The structure of KBS is shown in Fig. 1.

International Scholarly and Scientific Research & Innovation 1(12) 2007

Interface

Lack of spark: The spark might be nonexistent or weak for

a number of reasons:

Inference
Engine

User

Lack of compression: If the charge of air and fuel cannot be

compressed properly, the combustion process will not work
like it should. Lack of compression might occur for these
reasons:
1. Piston rings are worn (allowing air/fuel to leak past the
piston during compression).
2. The intake or exhaust valves are not sealing properly,
again allowing a leak during compression.
3. There is a hole in the cylinder.

Knowledge Base
Explanation Facility

International Science Index, Mechanical and Mechatronics Engineering Vol:1, No:12, 2007 waset.org/Publication/161

2. The air intake might be clogged, so there is fuel but not

enough air.
3. The fuel system might be supplying too much or too little
fuel to the mix, meaning that combustion does not occur
properly.
4. There might be an impurity in the fuel (like water in your
gas tank) that makes the fuel not burn.

741

scholar.waset.org/1999.8/161

World Academy of Science, Engineering and Technology

International Journal of Mechanical, Aerospace, Industrial, Mechatronic and Manufacturing Engineering Vol:1, No:12, 2007

____________________________________________
;Rule No 1 Run-Stable State
(Defrule Rule1 "gas smile and high avance"
(Selection 2)
(Fuel NotBurn)
(Cycle Ok)
(BlueGas No)
(Avance Bad)
=>
(printout t "Dirt in the
injections/carburetor or the
adjustment of ear and gasoline is not good")
(printout t "Advise: clear
injections/carburetor and adjust the ear"))

Rule
Rule
Rule
Rule

The engine will

turn over

1
2
3
4

The problem is
spark plugs

______________________________________

Fig. 6 The production system after Rule 1 is fired

Fig. 3 CLIPS representation of a rule

International Science Index, Mechanical and Mechatronics Engineering Vol:1, No:12, 2007 waset.org/Publication/161

Production
rules

Working Memory
The engine is
getting gas

An English counterpart of the above rule in represented in Fig.

4.
_________________________________________

The problem is X

Rule No 1: gas smile and high avance

IF the selection is 2 "Run-Stable State"
AND the fuel is not burning well
AND the engine running cycle is ok
AND there is no blue gas
AND the avance is bad
THEN
There is a Dirt in the injections/carburetor
or The adjustment of ear and gasoline is not
good, clear injections/carburetor and adjust
the ear.
____________________________________________________

Rule 2

the engine is
getting gas

Fig. 4 English representation of the rule in Fig. 3

D. Inference Engine
When abnormal situation arises in the car, the KBS makes
inferences by deciding which rules are satisfied by facts stored
in the working memory and executes the rule with highest
priority and propose proper correcting solution. The rules
whose patterns are satisfied by facts in the working memory
are stored in the agenda part of the inference engine. Figure 6
explains the inference process of the system using the rules
listed in Figure 5. Figure 7 shows the relationships between
the four rules used in this example.
____________________________________________________
Rule 1: IF the engine is getting gas,
AND the engine will turn over,
THEN the problem is spark plugs.
Rule 2:

IF the engine does not turn over,

AND the lights do not come on,
THEN the problem is battery or cables.

Rule 3:

IF the engine does not turn over,

AND the lights do not come on,
THEN the problem is the starter motor.

Rule 4:

IF there is gas in the fuel tank,

AND there is gas in the carburetor,
THEN the engine is getting gas.

Rule 1

Rule 3

the engine will

turn over

Rule 4
gas in fuel
tank

gas in
carburetor

Fig. 7 Example shows relationships among rules

E. Language selection
The proposed knowledge-based system is implemented
using the CLIPS expert system tool [12]. CLIPS is a forwardchaining rule-based language that resembles OPS5 and ART,
two other widely known expert system tools [1]. CLIPS was
developed by NASA and it has shown an increasing
popularity and acceptance by end users [13]We decided to use
CLIPS to implement the KBS of car failure diagnosis system
for four reasons. First, the data-driven nature of the domain
suggests the use of a forward-chaining inference engine.
Second, CLIPS runs in Windows environment, which is the
platform of choice among our end users. Third is the
availability of CLIPS source code. Finally, CLIPS provides
the ability to construct production rules which rely on patternmatching.
IV. RESULTS AND CONCLUSIONS

Fig. 5 Four assumed rules in the agenda

We presented in this paper a KBS for car failure diagnosis.

The system consists of four main stages. We implemented the
KBS using the CLIPS expert system language. During the test
phase of system it never gave wrong diagnosis according to

International Scholarly and Scientific Research & Innovation 1(12) 2007

742

scholar.waset.org/1999.8/161

International Science Index, Mechanical and Mechatronics Engineering Vol:1, No:12, 2007 waset.org/Publication/161

World Academy of Science, Engineering and Technology

International Journal of Mechanical, Aerospace, Industrial, Mechatronic and Manufacturing Engineering Vol:1, No:12, 2007

the rules used. The system indicated that a full expert system
will be practical and can be extremely useful in providing
consistent car failure detection. Further work is needed to
improve the system by adding sufficient domain knowledge
that represents domain knowledge thoroughly. Plans are
underway to convene experts to use the system to assist them
in their jobs of car failure detection. The first advantage of
using CLIPS is it allowed us to keep the system small, while
maintaining speed and ease of programming. The second
important advantage of using CLIPS is the suitability of the
forward reasoning and matching to the application and
representation of the knowledge. No meta-level reasoning was
necessary in the system. In summary, the system has the
characteristics of good expert systems, such as high
performance, adequate response time, understandability, and
understandability.

http:// www.mercedescarpart.com/
[21] U Kiencke and L Nielsen, Automotive Control Systems: For Engine,
Driveline, and Vehicle. Berlin: Springer, 2000.
[22] Brown Franklin Lee, Lebeck A. O., 1997, Cars, Cans and Dumps, New
Mexico.
[23] R Smith, H Aderson, and J E Morris (2004), Cylinders and Cycles,
Physics Department, Victoria University, Wellington, New Zealand.
[24] A R Jones (2004), Combustion Physics. Imperial Collage, London, UK.
[25] http://www.chroma.med.miami.edu/~idickers/m2-engine.html
[26] AutoWord,
http://www.autoword.com/automaintainance/improve-yourengin.asp
[27] http://www.car.com/applications/engine/builging.html

REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]

Joseph Giarratano, Gary Riley (2004). Expert Systems: Principles and

Programming, Fourth Edition.
Shu-Hsien Liao (2005). Expert system methodologies and applications a decade review from 1995 to 2004, Expert Systems with Applications,
28, 93-103.
Jeff Pepper (1990). An Expert System for Automotive Diagnosis in Ray
Kurzweil's book, The Age of Intelligent Machines.
Daoliang Lia, Zetian Fua, Yanqing Duanb (2002). Fish-Expert: a webbased expert system for fish disease diagnosis, Expert Systems with
Applications, 23, 311-320.
Yu Qian*, Xiuxi Li, Yanrong Jiang, Yanqin Wen (2003). An expert
system for real-time failure diagnosis of complex chemical processes,
Expert Systems with Applications, 24, 425-432.
Deschamps, D., & Fernandes, A. M. (2000). An expert system to
diagnosis periodontal disease. Proceedings of Sixth Internet World
Congress for Biomedical Sciences in Ciudad Real, Spain.
Guvenir, H. A., & Emeksiz, N. (2000). An expert system for the
differential diagnosis of erythemato-seuamous diseases. Expert Systems
with Applications, 18, 4349.
Huang, Q. M., Li, X. X., Jiang, Y. R., & Qian, Y. (2001). A fault
diagnosis expert system based on fault tree analysis for lubricating dewaxing process. Computers and Applied Chemistry, 18, 129133.
Cho, H. J., & Park, J. K. (1997). An expert system for fault section
diagnosis of power systems using fuzzy relations. IEEE Transactions on
Power Systems, 12, 342348.
Wiig, K. M. (1994). Knowledge management, the central management
focus for intelligent-acting organization. Arlington: Schema Press.
Dhaliwal, J. S., & Benbasat, I. (1996). The use and effects of knowledge
based system explanations: theoretical foundations and a framework for
empirical evaluation. Information Systems Research, 7, 342362.
NASA, Lyndon B. Johnson Space Center, CLIPS Basic Programming
Guide, 1991.
Martin, Linda and Taylor, Wendy, A Booklet About CLIPS Applications,
NASA, Lyndon B. Johnson Space Center, 1991.
How stuff works, http://auto.howstuffworks.com/engin.htm,
http://science.howstuffworks.com/fire-engin.htm
Rebuilt Auto Car N' Truck Engines,
http://www.rebuilt-auto-engines.com,
Car Engine Products in The UK's Leading Online Car Accessory Store,
http://www.speeding.co.uk/acatalog/engine.html
car: engine: combustion,
http://www.halfbakery.com/category/Car_3a_20Engine_3a_20Combusti
on/
Stone Age Corporation and Car.com,
http://www.car.com/applications/engine/work.html
Showcars, OPEL Catalogue, http://www.showcarsbodyparts.com/opel.html
Mercedes Car Parts, Mercedes Catalogues, A26, A27, A28, A28.6,

International Scholarly and Scientific Research & Innovation 1(12) 2007

743

scholar.waset.org/1999.8/161