Global Journal of Pure and Applied Mathematics.

ISSN 0973-1768 Volume 13, Number 9 (2017), pp. 5265-5273

© Research India Publications
http://www.ripublication.com

Analysis Of System Reliability Using Markov

Technique

S.Kalaiarasi1, A.Merceline Anita2, R.Geethanjalii3

1,2,3
Department of Mathematics, Sacred Heart College, Tirupattur, India.

Abstract
The Reliability rate of Human Error and other equipment error have been
studied over the past years. There are many models that have been developed
on the Reliability Analysis using Markov Technique. Markov models are very
much useful in finding the System Reliability in various life situations. In this
Paper, we present the Markov Model for a system comprising 4-Elements to
determine the Reliability R(t), Failure Probability F(t) and Mean Time to
Failure MTTF of the system using human error rate. Numerical illustration is
given and Conclusion is derived.
Key Words: Reliability, Markov Model, States, Mean Time to Failure,
Failure Rate.

1. INTRODUCTION
Reliability is the probability of a device performing its purpose adequately for the
period intended under the given operating conditions. In Reliability models, the states
usually represent the various working and failed conditions of the system.[9]
Markov Processes are widely used in Engineering, Science and Business Modeling.
Markov Modeling is a widely used technique in the study of Reliability analysis of
system. They are used to model systems that have a limited memory of their past. In a
Markov Process, if the present state of the process is given, the future state is
independent of the past. This property is usually referred to as the Markov Process.[1]
5266 S.Kalaiarasi, A.Merceline Anita, R.Geethanjalii

Andrei.A.Markov, a great Mathematician, after whom the term 'Markov Model have
been named.[5]
Markov modeling is a modeling technique that is widely useful for the Reliability
analysis of complex systems. It is very flexible in the type of systems and system
behavior it can model. This modeling technique is very helpful in most of the
situations .The model is quite useful to modeling operation system with dependent
failure and repair models. In fact it is widely used to perform Reliability and
Availability analysis of responsible system with constant failure and repair rates.
From time to time the Markov method is also used to perform human Reliability
Analysis.[7]
The Paper is organized as follows. In Section 1, we give the Introduction, In Section
2, we give the definitions and basic Concepts, In section 3, we present the Markov
Model related to our study, In Section 4, we discuss with a Numerical Example in
Medical Field, In Section 5, we draw the Conclusion.

2. BASIC DEFINITIONS AND RELATED CONCEPTS

Reliability
Reliability is the Probability that an item will carry out its function satisfactorily for
the stated period when used according to the specified conditions.
Reliability is defined as
∞
𝑅(𝑡) = ∫ 𝑓(𝑥)𝑑𝑥
𝑡

Failure Rate
Failure Rate is the frequency with which an engineered system or component fails,
expressed in failures per unit of time. It is often denoted by the Greek Letter λ and is
highly used in Reliability Engineering.
Human Error
This is the failure to carry out a required task that could result in disruption of
scheduled operations or damage to equipment or property.

3. MARKOV MODEL FOR A SYSTEM WITH 4 ELEMENTS

We discuss the Markov Model with 4-Elements. The Reliability of Markov Model for
a System with 3-Elements have been already derived.[8]
Analysis Of System Reliability Using Markov Technique 5267

In the 4-Elements Markov Model, each element has two states - good and failed state.
The states of the Model are generated based on the elements being in one of these two
states. An element with constant failure rate has a transition Probability that is
approximated by λΔt . The Probability for more than one element failure in time t is
considered to be negligible.
The transition diagram for four element Markov Model Process is given below.

Fig 1. Transition diagram for 4 element Markov Model

The Markov differential equation is developed by giving the probability of each state
at time 𝑡 + ∆𝑡 as a function of that state at time 𝑡. The Probability of being in state 1
at some time 𝑡 + ∆𝑡 is equal to the probability of being in state 1 at time 𝑡 and not
transitioning out during time ∆𝑡. Thus, we get the equation,
5268 S.Kalaiarasi, A.Merceline Anita, R.Geethanjalii

𝑃1 (𝑡 + ∆𝑡) = 𝑃1 (𝑡)[1 − (𝜆1 + 𝜆2 + 𝜆3 + 𝜆4 )∆𝑡]

The probability of being in state 2 at time 𝑡 + ∆𝑡 is equal to the probability of being in
state 1 at time t and transitioning to state 2 at time ∆𝑡 with the probability of being in
state 2 at time t and not transitioning out during time ∆𝑡. Thus, we get the equation,
𝑃2 (𝑡 + ∆𝑡) = 𝑃1 (𝑡)𝜆1 ∆𝑡 + [1 − (𝜆2 + 𝜆3 + 𝜆4 )∆𝑡]𝑃2 (𝑡)
The other state equations are as follows,
𝑃3 (𝑡 + ∆𝑡) = 𝑃1 (𝑡)𝜆2 ∆𝑡 + [1 − (𝜆1 + 𝜆3 + 𝜆4 )∆𝑡]𝑃3 (𝑡)
𝑃4 (𝑡 + ∆𝑡) = 𝑃1 (𝑡)𝜆3 ∆𝑡 + [1 − (𝜆1 + 𝜆2 + 𝜆4 )∆𝑡]𝑃4 (𝑡)
𝑃5 (𝑡 + ∆𝑡) = 𝑃1 (𝑡)𝜆4 ∆𝑡 + [1 − (𝜆1 + 𝜆2 + 𝜆3 )∆𝑡]𝑃5 (𝑡)
𝑃6 (𝑡 + ∆𝑡) = 𝑃2 (𝑡)𝜆2 ∆𝑡 + [1 − (𝜆3 + 𝜆4 )∆𝑡]𝑃6 (𝑡)
𝑃7 (𝑡 + ∆𝑡) = 𝑃3 (𝑡)𝜆3 ∆𝑡 + [1 − (𝜆1 + 𝜆4 )∆𝑡]𝑃7 (𝑡)
𝑃8 (𝑡 + ∆𝑡) = 𝑃4 (𝑡)𝜆4 ∆𝑡 + [1 − (𝜆1 + 𝜆2 )∆𝑡]𝑃8 (𝑡)
𝑃9 (𝑡 + ∆𝑡) = 𝑃5 (𝑡)𝜆1 ∆𝑡 + [1 − (𝜆2 + 𝜆3 )∆𝑡]𝑃9 (𝑡)
𝑃10 (𝑡 + ∆𝑡) = 𝑃6 (𝑡)𝜆3 ∆𝑡 + [1 − 𝜆4 ∆𝑡]𝑃10 (𝑡)
𝑃11 (𝑡 + ∆𝑡) = 𝑃7 (𝑡)𝜆4 ∆𝑡 + [1 − 𝜆1 ∆𝑡]𝑃11 (𝑡)
𝑃12 (𝑡 + ∆𝑡) = 𝑃8 (𝑡)𝜆1 ∆𝑡 + [1 − 𝜆2 ∆𝑡]𝑃12 (𝑡)
𝑃13 (𝑡 + ∆𝑡) = 𝑃9 (𝑡)𝜆2 ∆𝑡 + [1 − 𝜆3 ∆𝑡]𝑃13 (𝑡)
𝑃14 (𝑡 + ∆𝑡) = 𝑃10 (𝑡)𝜆4 ∆𝑡 + 𝑃11 (𝑡)𝜆1 ∆𝑡 + 𝑃12 (𝑡)𝜆2 ∆𝑡 + 𝑃13 (𝑡)𝜆3 ∆𝑡 + 1. 𝑃14 (𝑡)

Solving the above equations with the initial conditions,

P1(0)=1, P2(0)=0,…P14(0)=0
We get the following set of equations
𝑃1 (𝑡) = 𝑒 −(𝜆1 +𝜆2 +𝜆3 +𝜆4 )𝑡
𝑃2 (𝑡) = −𝑒 −(𝜆1 +𝜆2 +𝜆3 +𝜆4 )𝑡 + 𝑒 −(𝜆2 +𝜆3 +𝜆4 )𝑡
𝑃3 (𝑡) = −𝑒 −(𝜆1 +𝜆2 +𝜆3 +𝜆4 )𝑡 + 𝑒 −(𝜆1 +𝜆3 +𝜆4 )𝑡
𝑃4 (𝑡) = −𝑒 −(𝜆1 +𝜆2 +𝜆3 +𝜆4 )𝑡 + 𝑒 −(𝜆1 +𝜆2 +𝜆4)𝑡
𝑃5 (𝑡) = −𝑒 −(𝜆1 +𝜆2 +𝜆3 +𝜆4 )𝑡 + 𝑒 −(𝜆1 +𝜆2 +𝜆3 )𝑡
𝜆2 𝜆1
𝑃6 (𝑡) = 𝑒 −(𝜆1 +𝜆2 +𝜆3 +𝜆4 )𝑡 − 𝑒 −(𝜆2 +𝜆3 +𝜆4 )𝑡 + 𝑒 −(𝜆3 +𝜆4 )𝑡
𝜆1 +𝜆2 𝜆1 +𝜆2
Analysis Of System Reliability Using Markov Technique 5269

𝜆3 𝜆2
𝑃7 (𝑡) = 𝑒 −(𝜆1 +𝜆2 +𝜆3 +𝜆4 )𝑡 − 𝑒 −(𝜆1 +𝜆3 +𝜆4 )𝑡 + 𝑒 −(𝜆1 +𝜆4 )𝑡
𝜆2 +𝜆3 𝜆2 +𝜆3
𝜆4 𝜆3
𝑃8 (𝑡) = 𝑒 −(𝜆1 +𝜆2 +𝜆3 +𝜆4 )𝑡 − 𝑒 −(𝜆1 +𝜆2 +𝜆4 )𝑡 + 𝑒 −(𝜆1 +𝜆2 )𝑡
𝜆3 +𝜆4 𝜆3 +𝜆4
𝜆1 𝜆4
𝑃9 (𝑡) = 𝑒 −(𝜆1 +𝜆2 +𝜆3 +𝜆4 )𝑡 − 𝑒 −(𝜆1 +𝜆2 +𝜆3 )𝑡 + 𝑒 −(𝜆2 +𝜆3 )𝑡
𝜆1 +𝜆4 𝜆1 +𝜆4

𝜆2 𝜆3 𝜆3
𝑃10 (𝑡) = − 𝑒 −(𝜆1 +𝜆2 +𝜆3 +𝜆4 )𝑡 + 𝑒 −(𝜆2 +𝜆3 +𝜆4 )𝑡
(𝜆1 + 𝜆2 )(𝜆1 + 𝜆2 + 𝜆3 ) 𝜆2 + 𝜆3
𝜆1 𝜆4
− 𝑒 −(𝜆1 +𝜆4 )𝑡
𝜆4 (𝜆2 + 𝜆3 )
𝜆2 𝜆3 𝜆3 𝜆1 𝜆4
+[ − + ] 𝑒 −𝜆4 𝑡
(𝜆1 + 𝜆2 )(𝜆1 + 𝜆2 + 𝜆3 ) 𝜆2 + 𝜆3 𝜆4 (𝜆2 + 𝜆3 )

𝜆3 𝜆4 𝜆4
𝑃11 (𝑡) = − 𝑒 −(𝜆1 +𝜆2 +𝜆3 +𝜆4 )𝑡 + 𝑒 −(𝜆1 +𝜆3 +𝜆4 )𝑡
(𝜆2 + 𝜆3 )(𝜆2 + 𝜆3 + 𝜆4 ) 𝜆3 + 𝜆4
𝜆2 𝜆4
− 𝑒 −(𝜆1 +𝜆4 )𝑡
𝜆4 (𝜆2 + 𝜆3 )
𝜆3 𝜆4 𝜆4 𝜆2 𝜆4
+[ − + ] 𝑒 −𝜆1 𝑡
(𝜆2 + 𝜆3 )(𝜆2 + 𝜆3 + 𝜆4 ) 𝜆3 + 𝜆4 𝜆4 (𝜆2 + 𝜆3 )

𝜆1 𝜆4 𝜆1
𝑃12 (𝑡) = − 𝑒 −(𝜆1 +𝜆2 +𝜆3 +𝜆4 )𝑡 + 𝑒 −(𝜆1 +𝜆2 +𝜆4 )𝑡
(𝜆3 + 𝜆4 )(𝜆1 + 𝜆3 + 𝜆4 ) 𝜆1 + 𝜆4
𝜆1 𝜆3 𝜆1 𝜆4 𝜆1
− 𝑒 −(𝜆1 +𝜆2 )𝑡 + [ −
𝜆1 (𝜆3 + 𝜆4 ) (𝜆3 + 𝜆4 )(𝜆1 + 𝜆3 + 𝜆4 ) 𝜆1 + 𝜆4
𝜆1 𝜆3
+ ]𝑒 −𝜆2 𝑡
𝜆1 (𝜆3 + 𝜆4 )
𝜆1 𝜆2 𝜆2
𝑃13 (𝑡) = − 𝑒 −(𝜆1 +𝜆2 +𝜆3 +𝜆4 )𝑡 + 𝑒 −(𝜆1 +𝜆2 +𝜆3 )𝑡
(𝜆1 + 𝜆4 )(𝜆1 + 𝜆2 + 𝜆4 ) 𝜆1 + 𝜆2
𝜆2 𝜆4 𝜆1 𝜆2 𝜆2
− 𝑒 −(𝜆2 +𝜆3 )𝑡 + [ −
𝜆2 (𝜆1 + 𝜆4 ) (𝜆1 + 𝜆4 )(𝜆1 + 𝜆2 + 𝜆4 ) 𝜆1 + 𝜆2
𝜆2 𝜆4
+ ]𝑒 −𝜆3 𝑡
𝜆2 (𝜆1 + 𝜆4 )
𝑃14 (𝑡) = 1 − 𝑃1 (𝑡) − 𝑃2 (𝑡) − ⋯ − 𝑃13 (𝑡)
5270 S.Kalaiarasi, A.Merceline Anita, R.Geethanjalii

The Reliability of the System at time t is the sum of all the probabilities,
𝑛−1
𝑅(𝑡) = ∑ 𝑃𝑖 (𝑡), 𝑛 = 14
𝑖=1

𝑅(𝑡) = −3𝑒 −(𝜆1 +𝜆2 +𝜆3 +𝜆4 )𝑡

𝜆2 𝜆3 𝜆4
+ 𝑒 −(𝜆1 +𝜆2 +𝜆3 +𝜆4 )𝑡 [ + +
𝜆1 + 𝜆2 + 𝜆3 𝜆2 + 𝜆3 + 𝜆4 𝜆1 + 𝜆3 + 𝜆4
𝜆1 𝜆3 𝜆1
+ ] + 𝑒 −(𝜆2 +𝜆3 +𝜆4)𝑡 ( ) + 𝑒 −(𝜆1 +𝜆2 +𝜆4 )𝑡 ( )
𝜆1 + 𝜆2 + 𝜆4 𝜆2 + 𝜆3 𝜆1 + 𝜆4
𝜆2 𝜆4
+ 𝑒 −(𝜆1 +𝜆2 +𝜆3 )𝑡 ( ) + 𝑒 −(𝜆1 +𝜆3 +𝜆4 )𝑡 ( )
𝜆1 + 𝜆2 𝜆3 + 𝜆4
𝜆1 𝜆2 𝜆2 𝜆3
+[ ] 𝑒 −𝜆4 𝑡 + [ ] 𝑒 −𝜆1 𝑡
(𝜆2 + 𝜆3 )(𝜆1 + 𝜆2 + 𝜆3 ) (𝜆3 + 𝜆4 )(𝜆2 + 𝜆3 + 𝜆4 )
𝜆3 𝜆4 𝜆1 𝜆4
+[ ] 𝑒 −𝜆2 𝑡 + [ ] 𝑒 −𝜆3 𝑡
(𝜆1 + 𝜆4 )(𝜆1 + 𝜆3 + 𝜆4 ) (𝜆1 + 𝜆2 )(𝜆1 + 𝜆2 + 𝜆4 )
The Failure Rate is given by
𝐹(𝑡) = 1 − 𝑅(𝑡)
The Mean Time to Failure is given by
∞
𝑀𝑇𝑇𝐹 = ∫ 𝑅(𝑡)𝑑𝑡
0

3
𝑀𝑇𝑇𝐹 = −
𝜆1 + 𝜆2 + 𝜆3 + 𝜆4
1 𝜆2 𝜆3 𝜆4
+[ ][ + +
𝜆1 + 𝜆2 + 𝜆3 + 𝜆4 𝜆1 + 𝜆2 + 𝜆3 𝜆2 + 𝜆3 + 𝜆4 𝜆1 + 𝜆3 + 𝜆4
𝜆1 𝜆3
+ ]+
𝜆1 + 𝜆2 + 𝜆4 (𝜆2 + 𝜆3 )(𝜆2 + 𝜆3 + 𝜆4 )
𝜆1 𝜆2
+ +
(𝜆1 + 𝜆4 )(𝜆1 + 𝜆2 + 𝜆4 ) (𝜆1 + 𝜆2 )(𝜆1 + 𝜆2 + 𝜆3 )
𝜆4 𝜆1 𝜆2
+ +
(𝜆3 + 𝜆4 )(𝜆1 + 𝜆3 + 𝜆4 ) 𝜆4 (𝜆2 + 𝜆3 )(𝜆1 + 𝜆2 + 𝜆3 )
𝜆2 𝜆3 𝜆3 𝜆4
+ +
𝜆1 (𝜆3 + 𝜆4 )(𝜆2 + 𝜆3 + 𝜆4 ) 𝜆2 (𝜆1 + 𝜆4 )(𝜆1 + 𝜆3 + 𝜆4 )
𝜆1 𝜆4
+
𝜆3 (𝜆1 + 𝜆2 )(𝜆1 + 𝜆2 + 𝜆4 )
Analysis Of System Reliability Using Markov Technique 5271

4. NUMERICAL EXAMPLE
A Medical Laboratory consists of 4 equipments maintained by four Lab Technicians.
We assume that each of the four Technicians have some error rate in doing their jobs
successfully and their error rates are found based on their Past Records. Using their
error rates we calculate the Reliability of the System and Failure Probability of the
Medical Laboratory at any particular time period t. At time t=2,3,4,5 years the
Probabilities are tabulated and given below.
Assume that 𝜆1 = 0.10, 𝜆2 = 0.15, 𝜆3 = 0.20, 𝜆4 = 0.25

Table 1. Probabilities of different states at time t

Probabilities t=2 3 4 5
𝑃1 (𝑡) 0.2465 0.1224 0.0608 0.0301
𝑃2 (𝑡) 0.0546 0.0428 0.0299 0.0196
𝑃3 (𝑡) 0.0863 0.0696 0.0500 0.0338
𝑃4 (𝑡) 0.1213 0.1007 0.0745 0.0519
𝑃5 (𝑡) 0.1600 0.1368 0.1044 0.0752
𝑃6 (𝑡) 0.0094 0.0118 0.0117 0.0104
𝑃7 (𝑡) 0.0207 0.0278 0.0295 0.0276
𝑃8 (𝑡) 0.0345 0.052 0.0599 0.0605
𝑃9 (𝑡) 0.0185 0.0255 0.0821 0.0272
𝑃10 (𝑡) 0.0014 0.003 0.0046 0.0054
𝑃11 (𝑡) 0.0043 0.0098 0.0159 0.0313
𝑃12 (𝑡) 0.003 0.0071 0.0116 0.0157
𝑃13 (𝑡) 0.0021 0.005 0.0077 0.0101
5272 S.Kalaiarasi, A.Merceline Anita, R.Geethanjalii

Fig 2. Reliability of System with Human Error

Table 2. Reliability and Failure Probability at time t

Probability t=2 3 4 5
R(t) 0.7626 0.6143 0.5426 0.4588
F(t) 0.2374 0.3857 0.4574 0.5412

Fig 3. Success and Failure Probabilities at time t

Analysis Of System Reliability Using Markov Technique 5273

5. CONCLUSION
Reliability of the system can be calculated by many ways. Markov Modeling
technique is the one which is more effective in calculating the System Reliability. In
this Paper, we analyze the Reliability of the System based on the human error rate.
We determine the Reliability, Failure Probability and Mean Time to Failure of the
System, which would enable us to improve the effectiveness of the Health Care
System and to take care of Patients Safety. We observe that when the value of t
increases, the Reliability Rate R(t) decreases and when t increases, Failure Rate F(t)
also increases.

REFERENCES
[1] Balagurusamy.E, Reliability Engineering, Tata Macgraw Hill, New Delhi
1984.
[2] Dhillon.B.S, Reliability Engineering Applications Areas, Beta Publishers,
Gloucester, Ontario, 1992.
[3] Dhillon.B.S, Medical Device Reliability and Associated Areas, C.R.C Press,
Washington.D.C, 2000.
[4] Medhi.J, Stochastic Processes, New Age International (P) Limited Publishers,
New Delhi, 2006.
[5] Norman.B.Fuqua, The Applicability of Markov Analysis methods to
Reliabilty, Maintainability and Safety, Selected Topics in Assurance Related
Technologies (START), Volume 10 Number 2.
[6] Reni Sagayaraj.M, Merceline Anita.A, Chandrababu.A, Markov Models in
System Reliability with Applications.
[7] Reni Sagayaraj.M, Merceline Anita.A, Chandrababu.A, Gowtam Prakash.S,
Markov Models in System Reliability with Applictions, International Journal
of Innovative Research and Development, Volume 3 Issue 11, Nov- 2014.
[8] Reni Sagayaraj.M, Merceline Anita.A, Chandrababu.A, Shanmuga Priya.C,
Reliability Analysis of Medical System Using Markov Approach.
[9] Srinath.L.S, Reliability Engineering, East West Press, New Delhi, 2013.
[10] Wai-ki-ching, Michael.K.Ng, Markov Chains-Models, Algorithms and
Applications, Springer International, New Delhi, 2008.
5274 S.Kalaiarasi, A.Merceline Anita, R.Geethanjalii