International Journal of Automation and Computing

7(2), May 2010, 146-152

DOI: 10.1007/s11633-010-0146-z

Statistical Safety Analysis of Maintenance Management

Process of Excavator Units
Ljubisa Papic1
1

Milorad Pantelic2

Ajit Kumar Verma4

Research Center of Dependability and Quality Management, P.O. Box 132, Cacak 32102, Serbia
2

Joseph Aronov3

Kolubara Metal Company, Dise Djurdjevica 32, Vreoci 11560, Serbia

Federal Scientific Research Institute of Standardization, Elektricheskiy Pereulok 3/10, Moscow 123557, Russia
4

Indian Institute of Technology Bombay, Powai, Mumbai 400076, India

Abstract: Within this paper, the process of statistical safety analysis has been presented, which involves the following steps: formulation of basic principles of statistical safety analysis, initial events analysis, accident sceneries progress analysis, risk calculation, and
risk calculation results analysis. On this basis, it has been concluded that the bucket wheel excavator SRs 120024/40(400 kW)+VR
safety criteria is the mechanism for the hoist of rotor0 s arrow failure modes, because in that case whole bucket wheel excavator
failure would necessarily happen (excavator falling down on counterweight). Therefore, excavator units statistical safety analysis is
accomplished preventively to obtain its effective maintenance management.
Keywords:

Safety, maintenance management, statistical analysis, event tree, accident scenarios, risk analysis, excavator units.

Introduction

The basic problems when carrying out statistical safety

analysis involve selecting the most hazardous scenarios
which have the biggest impact on risk assessment. Statistical safety analysis is carried out by a team of experts
which, as a rule, consists of designers, process engineers,
mechanics and experts for engineering statistics and statistical safety analysis. Their role is working out certain calculations which enable establishing possible scenarios for the
development of accidents, as well as assessment of the system accident effects. Any kind of accident condition results
in damaging personnel health and life and great economic
losses resulting from the cost of reengineering and restarting of the system. Well carried out statistical safety analysis
and risk assessments of excavator units, which are considered in this paper, are preconditions for choosing a suitable
maintenance management system.

Basic principles of statistical safety

analysis

When carrying out statistical safety analysis in the earliest stage possible its results are more effective as system safety assurance will involve a lower cost. This is explained by the rule of ten times increase of costs for removing/elimination of defects (nonconformities) when transitioning to the next system life cycle phase[1] . This simple
rule of ten times cost increase shows the importance of the
early detection of potential problems in the field of safety.
The base of statistical safety analysis is event tree construction, i.e., carrying out system analysis of what occurs
after the initial event.
The procedure of carrying out of statistical safety analysis includes the following stages:
1) Choice and classification of initial events, as well as
Manuscript received December 12, 2009; revised February 27, 2010

assessment of their frequency;

2) Using real data about systems0 items reliability in
studied scenarios of what occurs after initial event;
3) Analysis of the sequence of accident occurrence;
4) Probabilities calculation of realization of accident occurrence sequence;
5) Classification of final stages and risk calculation.
It should be pointed out that carrying out statistical
safety analysis presents a very complex and difficult problem whose solution often demands the engagement of other
experts except for the ones mentioned above. As statistical safety analysis is carried out in the phase of designing
and maintenance, when some data and pieces of information about certain processes and phenomena are missing
or incomplete, risk calculation is followed by a high degree
of uncertainty[2] . As a rule, possible uncertainties in risk
calculations do not influence solutions to other problems
of statistical safety analysis, as it is supposed that they
equally influence calculation risk values, for example when
comparing different variants of system design.
In many cases, carrying out statistical safety analysis
fully can be more difficult. Therefore only qualitative and
quantitative analysis of system reliability is carried out[3] .
In order to increase safety analysis, objectively verified (approved) databases should be used which contain
knowledge about reliability indicators of similar items and
regimes and their work conditions, as well as specialized
databases about personnel reliability.
Results of statistical safety analysis are formed in reports
and submitted to the archive for later checking and expert
opinions. These results are to be inspected by detailed engineering analysis so that suitable corrective actions can
be taken. A very important demand when carrying out
statistical safety analysis is simplicity of their results interpretation as in the opposite case it may happen that it
is not understood by engineering personnel. Furthermore,
the basic stages of carrying out statistical safety analysis
are considered step by step.

147

L. Papic et al. / Statistical Safety Analysis of Maintenance Management Process of Excavator Units

Initial events analysis

At this stage a list of possible events potentially hazardous from the aspect of damage occurrence which exceeds
an allowed level is made and a selection from the list is made
of initial events group which is later used for modeling by
means of event tree construction. Carrying out of this stage
is necessary in order to reduce selected scenarios of possible
accidents.
When making the whole list of initial events, internal and
external initial events should be separated. Internal initial
events are caused by system items failures, operator0 s incorrect activities or mechanic0 s errors, while the external are
caused by influences connected with natural phenomena or
human activities in the territory (region) where the system
is located (earthquakes, winds, floods, terrorist attacks).
Classification of initial events is shown in Fig. 1.

2) Winds
Winds are atmosphere whirlwinds (whirlpools). Winds
of great dimensions (surfaces) speed of up to 120 km/h are
hurricanes. Annual hurricanes quantity possible to describe via Poisson0 s law with density function:
f () =

Classification of initial events

As the starting data for carrying out this stage, accident

analysis of similar systems is used. The importance of work
in this stage is conditioned by the need for safety assurance
not only in the period of normal exploitation, but when
initial event occurs[4] .
A short consideration of some kinds of initial events is
given.
1) Earthquakes
Earthquakes are oscillations of the earth0 s crust due to
sudden movements and fractures in the earth0 s crust. Movement of the ground during an earthquake has a wave character.
Classification of earthquakes according to the dimension
and strength is done according to the dimensionless scale
of magnitude M which characterizes total energy of elastic
oscillation, caused by an earthquake. The scale M is in
the interval between 0 and 9. The rate of an earthquake on
the ground surface is assessed according to the international
scale UNESCO MSK 1964. Earthquakes classification according to the magnitudes, degrees, and average frequency
is shown in Table 1.
Table 1

Classification of earthquakes

Earthquake

Magnitude

Degree

characteristic

Average frequency
of earthquake
(during the year)

World proportion
Strong, regional

12

78

910

1520

67

78

100150

significance
Strong, local
significance
Mean

56

67

7501000

Weak, local

45

56

50007000

(1)

where is mean annual hurricane frequency (for hurricanes

of Atlantic Ocean coast in USA, = 2).
3) Floods
Floods are sinking of regions (grounds) due to rising of
water level in a river, lake or sea. Between different natural
catastrophes, according to frequency and material losses, in
many countries floods are on first place.
4) System items failures
System items failures have a main role in accidents occurrence. Thus, in the last ten years in the coal mines of
the Serbian Electric-Power Industry, the causes of excavator units accidents can be grouped in the way as it is given
in Table 2.
Table 2

Fig. 1

exp()
!

Causes of excavator units accidents

Cause of accident

Accident share (%)

Difficult exploitation

27

conditions
Error in manufacture

22

and assembly
Operator0 s error

18

Mechanic0 s error

13

Fatigue of materials,

wear of equipment and

corrosion processes
Inadequacy design

Other miscellaneous factors

In other potentially hazardous systems, this accident

share is changeable, but even in these cases system items
failures have the main role in accidents occurrence.
As a rule, failure frequency or failure rate of system items
are determined according to the results obtained from specially organized testing of these items for reliability assessment or when testing within the system structure. However,
the most accurate assessment of failure rate is obtained by
data processing about failures from the exploitation of the
similar systems. As a result, a database about items reliability is formed and it can be used for initial events analysis. By taking out information about items reliability from
the corresponding database, frequency assessment of initial
events can be done for the factor of system items failure.
5) Personnel errors
Personnel errors (operator, mechanic) also play an essential role in accident occurrence which is proved by the
data from Table 2. Analysis of these data show that human
(personnel) errors in more than 30 % of cases present causes
of initial events of accidents with bucket wheel excavators.
Here, error is meant a human (personnel) failure mode,
which is not connected with a strike or sabotage. According
to previous research data[5] , incorrect, i.e., wrong activities
of personnel when operating complex systems caused up to

148

International Journal of Automation and Computing 7(2), May 2010

40% of unwanted results when rocket testing, up to 30%

radio electronics equipment failures.
For initial events analysis connected with personnel errors, it is important to have data about human reliability.
Today, this information about human errors is found in special databases which are formed according to the results of
special laboratory experiments or according to results of exploitation of one system type. Detailed information about
personnel reliability analysis is given in the literature[6] .
Taking into account the human factor in risk analysis
presents significant value of statistical safety analysis.
6) Screening
Screening presents a procedure of excluding those initial
events from the starting list whose frequency (rate) value
is very low and whose consequences (on the basis of results of analysis of similar systems accidents) are worthless
(minimal) in comparison with other initial events. In that
way screening enables shortening of the list of initial events
up to a reasonable level. As a result, a final list of initial
events with suitable frequency values of their occurrences
is formed, which further enables carrying out quantitative
risk calculation[7] . The procedure of screening can be carried out by applying Pareto diagram method[8] or method
of failure modes, effects and criticality analysis[9] .

Analysis of accidents occurrence scenarios

As previously explained, system safety is defined by order

of events, for example by personnel errors, external events
and others, which cannot be treated as failures. Furthermore, analysis of a great number of accidents of different
kinds of systems (see Fig. 2) show that they usually result
(cumulate) from an order of events, whose order (and connections) is suitable to be presented as an event tree[10] .
Event tree presents a continuous non periodical graph where
an initial event and intermediate event are distinguished,
caused by the occurrence of initial and final states. Initial
and intermediate events which come later really describe
possible paths of unwanted event flow (accident).

simplicity, visualization of risk calculation and possibility

of taking into account of the operator (mechanic) by means
of including real elements into the event tree which characterize the work of the operator (mechanic) or by means
of studying initial events of accidents which are connected
with personnel errors. Event trees are oriented towards
taking into account cause-effect dependence among system
items condition in certain instants of time among which
critical conditions may be found.
An event tree is constructed with an aim of effects analysis of some initial event I0 (item failure, personnel error, or
external event), which is drawn at the foot of a tree base.
This initial event may (or may not) lead to later events,
directly caused by the initial event that are called events
of the first level: I11 , I12 , , I1k . Each event of the first
level may (or may not) cause later events which are directly
connected with it. Putting it in another way, an event tree
by itself presents a logical diagram which defines (shows) a
set of system final states, out of which each represents realization of certain intermediate events combination, which
can influence accident development process in initial event.
The following events can be used as intermediate:
1) Successful or unsuccessful activation of system items,
which include insurance and blockade,
2) Correct or incorrect activity of personnel (operator,
mechanic).
Development of an event tree is carried out by a certain
schedule:
Step 1. Choice of a certain initial event from the final
list of initial events and its description.
Step 2. Determination of functions, which should be carried out by certain system items when certain initial event
chosen from the list occurs.
Step 3. Modeling event tree (construction of accident
development scenario).
Step 4. Classification of final states set.
The first step is clear and is not necessary to be explained.
As a result of carrying out the first step it is possible to construct an event tree base.
The system for lifting of a bucket wheel excavator SRs
120024/40(400 kW)+VR rotor0 s arrow which consists of
two driving items parallel connected (in reliability block diagram) whose task is lifting and lowering rotor0 s arrow during the digging process in the coal mines is shown in Fig. 3.

Fig. 2 Accident of bucket wheel excavator SchRs 1760/632

No. IX at coal mine field D, Mining Basin Kolubara (Serbia)

The most important and bright sides of the methodology

of event tree construction for risk calculation are analysis

Fig. 3 System for lifting of bucket wheel excavator SRs

120024/40(400 kW)+VR rotor0 s arrow (fragment)

149

L. Papic et al. / Statistical Safety Analysis of Maintenance Management Process of Excavator Units

From the final list of initial events, which have been

considered in the previous step, the event I0 is chosen
mechanical defect of a small gear on the outgoing spindle
of driving item No. 1. After carrying out the first step,
columns of Fig. 4 are filled and the event tree base is constructed. In Fig. 4, the following marks are introduced:
1) Items 1 and 2 denote driving item No. 1 and driving
item No. 2, respectively;
2) I denotes initial event (mechanical defect of small gear
on outgoing spindle of driving item No. 1).

probability is determined by using results of initial events

analysis.
The last expression presents total probability formula
which characterizes unconditional (full) accident occurrence
probability, i.e., accident risk R [5] .
In practice, as initial events are very rare, for probability
distribution of their occurrence for the time T a Poisson0 s
distribution can be taken:
P ( = m) = m e /m!,

m = 0, 1, 2, , , > 0 (4)

which characterizes occurrence probability of exact m initial events in a time unit. Here, is intensity of initial event
occurrence which is measured by the number in a unit time.
Supposing that m = 1 and T 0 (which is justified
for high reliable potential dangerous systems) it is obvious
that:
P ( = 1) = P (I0 ) .

Fig. 4 Example of event tree construction table fill out after

first step of the analysis

It is useful to give additional clarifications which should

completely explain modeling of the event tree. Modeling
the event tree enables, as it has already been said, taking
into account personnel role assessment after occurrence of
initial event if its participation is predicted. This can be
achieved by introducing a fictions item in second column of
Fig. 4 and correspondingly showing of the point of branching which show the action of personnel: stair up correct
reaction of the operator and stair down incorrect work
of the operator.

Risk calculation

If for a particular initial event I0 we can select n scenarios

of accidents occurrence which are marked as: E1 , E2 , ,
En , in this case accident may occur before realization of
n non-simultaneous (random) scenarios of accident occurrence. Thus, accident is an event (in statistical sense) which
represents a collection of non simultaneous (random) events
E1 , E2 , , En . Thus, accident probability (conditional)
is shown with the formula:
Q(I0 ) =

n
X

Qi (Ei /I0 ),

i = 1, 2, , n

(2)

i=1

where Qi (Ei /I0 ) denotes probability of realization of the

scenario of accident occurrence for a particular initial event.
For calculating total probability R(I0 ) of accident occurrence (unconditional) it is necessary to take into account
probability P (I0 ) of initial event occurrence. In that case,
according to the total probability formula, accident probability R(I0 ) can be calculated when initial event I0 occurs:
R(I0 ) = P (I0 )

n
X
i=1

Qi (Ei /I0 ) =

n
X

P (I0 ) Qi (Ei /I0 ) (3)

i=1

where P (I0 ) denotes probability of initial event occurrence

I0 for a certain period of time T , e.g., for one year. This

Thus, in (3) for calculating risk instead of initial event

occurrence probability it is useful to change the rate (frequency) of its occurrence:
R(I0 ) =

n
X

Qi (Ei /I0 ).

(5)

i=1

This substitution is connected with simpler risk defining

as accident frequency in a time unit. Majority of quantitative safety analysis includes risk assessment exactly in
this form. Apart from this, very often analysis of initial
events relies on the information about frequency and not
on probability of their occurrence.
On the other hand, values Qi (Ei /I0 ), i = 1, 2, , n is
calculated according to the formula of simultaneous occurrence of independent events probability (in a set) which
form a particular scenario of accident occurrence Ei . In
other words, if Ei is a scenario of accident occurrence caused
by ki independent, in a set of events (items failures, personnel errors, items operation without failures) whose probabilities are equal to ij then
Qi (Ei /I0 ) =

ki
Y

ij

(6)

j=1

where j = 1, 2, , ki , ij = pij which denotes probability

of operation without failures, or ij = qij which denotes
failure probability.
It should be emphasized that assumption of independence within a group of events, which enter an accident
occurrence scenario, is rather disputable. However, taking
into account dependence of events can make the calculation
of probability Qi (Ei /I0 ) much more difficult that is why it
is not considered here.
Calculated values Qi (Ei /I0 ) are entered in the fifth column of Fig. 5. Apart from that, sometimes, it is useful to
enter values of all events scenarios realization probability in
this column. As an example, in Fig. 5, probability values
of all possible scenarios previously classified in appropriate
groups are given. In Fig. 5, SCO denotes state of capability
to operate, and SAC denotes state of accident.

150

International Journal of Automation and Computing 7(2), May 2010

Fig. 5 Example event tree with presentation final states probabilities

Analysis of the fourth column in Fig. 5 shows that the

number of accident scenarios equals unit (i = 1). In that
case,
Q(I0 ) = Q1 (E1 /I0 ).
(7)
On the other hand, conditional probability Q1 (E1 /I0 )
of accident scenario realization (failure probability of both
items) is determined as
Q1 (E1 /I0 ) = (1 P1 ) (1 P2 ).

(8)

Here, when calculating value Q factor of time is not taken

into account (determined operation time), which has an important role when calculating probability of operation without failure. It is obvious that if a set of final states matches
with the full set of elementary events (within the limits of
elementary probability theory), in that case the sum of all
final states probability equals to one.
The risk accident value is calculated according to (8) taking into account condition (7):
R(I0 ) = P (I0 ) Q(E1 /I0 ) = P (I0 ) (1 P1 ) (1 P2 ). (9)
In complex cases, the event tree can be extended, thus
the analysis of risk calculation results becomes complicated
accordingly.

Analysis of calculation risk results

As the initial events in the event tree analysis, besides

functional failure modes in the mechanism for the hoist of
rotor0 s arrow, observed modes are of personnel errors (operators and mechanics). Based on data from the bucket wheel
excavator SRs 120024/40 (400 kW)+VR, failure map[11]
finds the list of modes of operator error, modes of mechanic
error and failure modes of mechanism for the hoist of rotor0 s
arrow.
1) List of modes of operator errors, n = 1, 2:
i) The operator often turns on mechanism for the hoist
of rotor0 s arrow;
ii) The operator often turns on mechanism for the hoist
of rotor0 s arrow when the excavator is on ground level.
2) List of modes of mechanic error, m = 1, 2, 3:
i) Mechanic has not properly performed assembly of the
coupling at the small group generator;
ii) Mechanic has not made centering of electric motors
precisely;
iii) Mechanic has not adjusted arrester for car interlocking.
3) List of failure modes of mechanism for the hoist of
rotor0 s arrow, k = 1, 2, 3:

i) Breaking at the back gearbox shaft (front-end) for the

hoist of rotor0 s arrow;
ii) Outage of electric-hydraulic lifter (releaser) at operating brake;
iii) Mechanical defect of ropes for the hoist of rotor0 s arrow.
Event tree for the initial event failure mode of mechanism for the hoist of rotor0 s arrow, k = 3: mechanical defect
of ropes for the hoist of rotor0 s arrow, is shown in Fig. 6,
where No denotes that it happened without unwanted
event, Yes denotes that it happened with unwanted event,
SNO denotes state of noncapability to operate.
The probability of occurrence of a state of accident scenario realization is
P (E2 /I0 ) = (1 Pi ) (1 P2 ) (1 P3 ) (1 P4 )
(1 P5 ) = 0.025 0.115 0.125 0.035 0.045 =
0.566 106 .
This result analysis presents the final stage of statistical
safety analysis. Its content depends (to a great extent) on
overall aims of statistical safety analysis. For example, risk
calculation results enable solving problems:
1) Comparison of several system variants (in the safety
section),
2) Showing of principal realization of required safety,
3) Choice of effective maintenance management process
or system.
To solve this problem, it is necessary to compare risk values R(I0 ), calculated for several system variants and choose
the one where the risk value is minimal. Solving of the
second problem is connected with comparison of calculated
risk value R(I0 ) with criterion risk value. To solve the third
problem, the special importance is that with inadequate operation maintenance safety must not be endangered, which
causing accident states of excavator units. Any state of
accident results in endangered health and life of personnel
and great economic losses expressed through cost of reengineering and repeated starting of excavator units.

Conclusions

The role of statistical safety analysis in the maintenance

management process of excavator units in the exploitation
phase is different from their design phase. This is explained
by the fact that excavator units are already built by that
time and are in the exploitation phase, so the safety analysis
is directed towards clarification whether it meets requirements in the safety field. If the excavator units do not meet
standard safety requirements it is very important to detect
that fact soon enough before any accident happens. In that
case, it is of the highest priority to develop maintenance
operation activities to increase safety.
In condition of limited resources, maintenance operation
should be clearly and precisely explained, so there will not
be wastage of resources. Subjectivity in making decisions
about maintenance operation concerning the safety of excavator units makes great economic losses. Therefore, development of scientific approaches in maintenance management using statistical safety analysis is of great importance.

L. Papic et al. / Statistical Safety Analysis of Maintenance Management Process of Excavator Units

Fig. 6

151

Event tree for initial event mechanical defect of ropes for the hoist of rotor0 s arrow

References

[8] B. G. Dale. Managing Quality, Oxford, USA: Blackwell

Publishing, pp. 495, 1999.

[1] G. Yang. Life Cycle Reliability Engineering, Hoboken, USA:

John Wiley & Sons Inc., pp. 531, 2007.
[2] J. Knezevic, L. Papic, B. Vasic. Sources of fuzziness in vehicle maintenance management. Journal of Quality in Maintenance Engineering, vol. 3, no. 4, pp. 281284, 1997.
[3] J. A. Nachlas. Reliability Engineering, Probabilistic Models
and Maintenance Methods, Boca Raton, USA: Taylor and
Francis Group, pp. 403, 2005.
[4] T. A. Kletz. Learning from Accidents, Oxford, USA:
Butterworth-Heinemann Ltd., pp. 279, 1994.
[5] J. Aronov. Methodology of Operational Safety Management

[9] D. H. Stamatis. Failure Mode and Effects Analysis, FMEA

from Theory to Execution, Milwaukee, USA: ASQ Quality
Press, pp. 487, 2003.
[10] Zio E. An Introduction to the Basis of Reliability and Risk
Analysis, Singapore: World Scientific Publishing Co. Pte.
Ltd., pp. 234, 2007.
[11] L. Papic, M. Pantelic. Implementation Methodology for
Risk Minimization into Maintenance Process of Production System at Coal Mines, Report of Contract No. 4617,
DQM Research Center Kolubara Metal, Prijevor Vreoci, pp. 468, 2009. (in Serbian)

on the Base Disturbances Statistical Analysis during Systems Operation and Assessment Methods Standardization,
Ph. D. dissertation, VNIIS, Moscow, pp. 254, 1998. (in Russian)
[6] T. Kletz. An Engineer0 s View of Human Errors, Rugby,
USA: Institution of Chemical Engineers, pp. 209, 1991.
[7] E. Zio. Computational Methods for Reliability and Risk
Analysis, Singapore: World Scientific Publishing Co. Pte.
Ltd., pp. 362, 2007.

Ljubisa Papic
received the Ph. D. degree in reliability engineering from University of Novi Sad, Serbia. Currently, he is
a professor and the head of the Department of Industrial and Systems Engineering in the Faculty of Engineering Cacak at
the University of Kragujevac, Serbia. He
is a member of Russian Quality Problems
Academy and Corresponding Member of Serbian Engineering
Sciences Academy. He is director and founder of the Research

152
Center of Dependability and Quality Management (DQM Research Center) and he is serving as editor-in-chief of Communications in Dependability and Quality Management, an international journal. Also, he is serving as an editorial board member
of Metody Menedzhmenta Kachestva and Problemy Mashinostroeniya i Avtomatizatsii, journals in Moscow, Russia. He was a
visiting professor at Ben-Gurion University of the Negev, Israel
and at Valencia Polytechnic University, Spain. He is a member of the IEEE Reliability Society and senior member of the
American Society for Quality (ASQ). He is a recipient of Education and Research Leadership Award by Society for Reliability
Engineering, Quality and Operations Management.
His research interests include reliability testing, failure analysis, safety analysis, and concurrent engineering
E-mail: papicdqm@ptt.rs (Corresponding author)

Milorad Pantelic
received the Ph. D.
degree in maintainability engineering from
University of Kragujevac, Serbia. He is the
deputy director of Kolubara Metal Company, Vreoci, Serbia. Also, he is an assistant professor at the Department of Industrial and Systems Engineering in the Faculty of Engineering Cacak at the University of Kragujevac, Serbia. He is a recipient Best Technological
Innovation Award by Ministry of Science and Technological Development of Republic Serbia.
His research interests include maintenance technology, maintenance concepts, and diagnostics.
E-mail: milorad.pantelic@kolubarametal.com

International Journal of Automation and Computing 7(2), May 2010

Joseph Aronov received the Ph. D. degree in safety engineering from Federal Scientific Research Institute of Standardization (VNIIS), Moscow, Russia. Currently,
he is a professor and the department head
at VNIIS in Moscow, Russia. He is serving as scientific editor of Metody Menedzhmenta Kachestva, journal in Moscow, Russia. Also, he is serving as an editorial board member of Communications in Dependability and Quality Management, an international journal, published by DQM Research Center, Serbia.
His research interests include reliability analysis, safety analysis, statistical methods, and conformity assessment.
E-mail: aiz@gost.ru
Ajit Kumar Verma
received
the
B. Tech. (Hons) and Ph. D. (Engg.) degrees
from Department of Electrical Engineering, IIT Kharagpur, India. He is currently
working as a director of the International
Institute of Information Technology, Pune
on lien from IIT Bombay, India, where he is
a professor in the Department of Electrical
Engineering. He has supervised twenty three Ph. D. dissertations and eighty four master dissertations at IIT Bombay. He
has been a guest editor of special issues of various international
journals. He is a senior member of IEEE and life fellow of IETE.
His research interests on reliability engineering include interdisciplinary applications in software engineering, computing,
maintenance, quality management, and power systems.
E-mail: akvmanas@gmail.com