CORE Metadata, citation and similar papers at core.ac.

uk
Provided by Repository of the Academy's Library

POLLACK PERIODICA
An International Journal for Engineering and Information Sciences
DOI: 10.1556/606.2018.13.1.9
Vol. 13, No. 1, pp. 99–112 (2018)
www.akademiai.com

RELIABILITY BASED MAINTENANCE

METHODOLOGY FOR SUSTAINABLE TRANSPORT
ASSET MANAGEMENT
1
Ejiroghene Onome EKPIWHRE, 2 Kong Fah TEE

Department of Engineering Science, University of Greenwich, Kent, ME4 4TB

United Kingdom, e-mail: [email protected], [email protected]

Received 13 February 2017; accepted 4 September 2017

Abstract: Functional failures and structural deterioration defects are among the leading
causes of growing failure probabilities of the road systems and networks. Thus, asset maintenance
intervention is an essential task for the unified management of road assets and systems. The
objective of this research is to consider reliability based probabilistic approach established on
functional failure mode, effects and criticality analysis parametric reliability analysis, risk-based
assessment and strategic asset deterioration decisions. The combined progressive assessment
frameworks and algorithms based procedures utilize service inspection, safety inspection and
survey inputs. The illustrated reliability maintenance based methodologies offer sustainable asset
management for highway transport infrastructure and systems with emphasis on resolutions to
their functional failures, defect related risk and appropriate deterioration treatment.

Keywords: Highway assets, Reliability centered maintenance, Reliability growth curves,

Risk based inspection, Deterioration treatment decisions

1. Introduction
The highway infrastructures are vital for any nation’s economic wellbeing and are
often the most valuable asset for local authorities but still do not often receive the
anticipated attention that commensurates an optimal state of its operations and
maintenance [1]. The code of practice for the maintenance of highway structures [2]
portrays some key importance of a well maintained and available highway
infrastructure, which brings better economic, social and environmental wellbeing of any
country. The arrangement for the management of highway maintenance is best effective
when set within the framework of an overall asset management regime. The

HU ISSN 1788–1994 © 2018 Akadémiai Kiadó, Budapest

100 E. O. EKPIWHRE, K. F. TEE

catastrophic effect and consequences of the lack of reliable based asset management
solutions often lead to reduced system safety, decreased asset availability, short asset
lifetime resulting in an increased whole life cycle cost [3] or total renewal. Fig. 1 shows
the failed (a) and the renewed (b) condition of a road junction network.

a) b)
Fig. 1. Drone view of a road junction network, a) failed and b) renewed state

Frameworks provide the means of understanding the value and liability of the asset,
giving authorities and asset owners a better understanding of the asset condition in
taking a right strategic decision [4]. The choices of highway asset conception, operation,
inspection, maintenance, renewal and disposal require an asset management framework
to achieve a safe and cost-effective infrastructure system. The likelihood of major
appalling asset failures and budget challenges can be decreased by the application of
asset management tools. Highway assets and elements are multifaceted physical assets
that consist of different road infrastructure (e.g. pavement, safety systems, traffic
system, lighting systems, road surface systems). The road authority does not own some
others highway elements that make up the road network (e.g. post office box, utility
services). In consideration of the complexity of the various disparate highway assets, a
mere technique is not feasible to carry out an effective maintenance. Instead, a more
strategic procedure is required with consideration of various highway asset types, their
diverse, unique functions, intermittent functional failures and random deterioration
pattern.
The advancement in computer technology has aided understanding prognosis
associated with ageing of many critical structures (e.g. buildings, bridges, highway
infrastructures, underground pipelines, offshore structures, mechanical structures).
However, the need for renewed reliable methods and models are still required in
highway asset functional Failure Mode, Effects and Criticality Analysis (FMECA),
parametric reliability analysis, risk-based inspection and assessment, strategic
deterioration models, and treatment cost optimisation. The contents of this paper are
organized as follows. In Section 2, reliability based maintenance process is projected.
2.1, Reliability Centered Maintenance (RCM) strategy is presented. Reliability analysis
and growth curves modeling are discussed in Section 2.2. In Section 2.3, a risk based
inspection and assessment of category 2 defects are discussed, and in conclusion of the

Pollack Periodica 13, 2018, 1

 RELIABILITY BASED MAINTENANCE METHODOLOGY 101

models, section 2.4, deliberated on the strategic treatment of deterioration prediction.

Finally, an application applied to carriageway is used to validate the proposed models in
Section 3 and the conclusion and discussion are presented in Section 3.

2. Reliability based maintenance process

Amongst the drive of a reliability based maintenance practice for a sustainable
transport asset management is the developmental support highway infrastructure brings
providing the base for essential services a vibrant economy requires. However,
managing highway infrastructures that are reliable, safe to user and cost effective to
infrastructure owners is a challenging task. The functional failure and gradual
deterioration of road assets are very precarious to an economy surface transportation
with its effect diffusing throughout its entire networks. The results of highway assets in
their poor condition with continuous functional failures lead to increased operating cost,
longer travel time and damage to the vehicle to road users [5]. The unavailability of
highway assets not only account for high accident rates but also are considered
originators of most traffic congestion and accidents [6]. The highway assets of the road
network often handle more traffic than they are often designed for, therefore requiring
an adequate road traffic management systems. This section defines the approaches
proposed to achieve a reliable based maintenance asset management with consideration
of highway infrastructure.
Maintenance, which is a core function for sustaining long-term profitability of assets
by organizations, is defined as a combination of all actions with intent to restore the
asset to its original state or a state where it can perform its required function. This report
devices a novel approach using systematic knowledge of the basics of reliability [7], [8],
[9], risk [10], [11], deterioration [12], [13] and cost strategies [3], [14] and models
embedded in qualitative and quantitative assessment methods as it is shown in Fig. 2
recognizing that algorithms support effective optimizations of systems [15], [16].

Fig. 2. Reliability-based maintenance methods for asset sustainability

2.1. Reliability, centered maintenance

The Reliability Centered Maintenance (RCM) is a technique that determines the
maintenance requirement of a system and the interims at which these are to be carried

Pollack Periodica 13, 2018, 1

102 E. O. EKPIWHRE, K. F. TEE

out in its operating context through a failure mode, effects and criticality analysis. It
focuses on the functions and failures of the assets as well as identifies their
consequences. It uses a standardized logical resolution procedure to implement
preventive measures from these identified consequences. RCM techniques decide the
required maintenance of a system while in its operating environment. The asset
management institute classifies RCM as a control that supports optimized management
of physical assets. RCM combines different approaches to aid the development of a
systematic maintenance program to manage risks as a basis for maintenance decisions
[17]. RCM framework focuses on preserving system functions, rather than preserving
physical asset since it offers asset availability, reliability and maintainability. RCM
process ensures seven key questions are answered reasonably order as follows:
I. What are the functions and associated desired standards of performance of the
asset in its present service context (functions)?
II. In what ways can it fail to accomplish its functions (functional failures)?
III. What causes each functional failure (failure modes)?
IV. What transpires when each failure occurs (failure effects)?
V. In what way does each failure matter (failure consequences)?
VI. What should be done to forecast or prevent each functional failure (proactive
tasks and task intervals)?
VII.What should be done if an appropriate proactive task cannot be found (default
actions)?
The RCM process as conveyed below provides the most information and details
about asset functions, failure modes, criticality analysis and maintenance actions that
address the functional failures. The process analysis is explained as follows and further
reading in [18] and its developed case study application capturing highway assets [7].
System partitioning: This identifies all the technical information of functionally
significant items. The asset descriptive and operational information (e.g. asset defect,
defect categories, defect period, repair hours, repair cost) are gathered from traditional
expert judgment, maintenance literature and computerized maintenance management
systems as related to the individual asset.
Failure Mode Effect Analysis (FMEA): The principal causes of functional failure are
identified in this step. The functional failure is an unsatisfactory condition, which
results from an asset not adequately providing its intended function. It captures assets
functions, functional failure, failure modes and failure effects. The specific situation
causing the functional failure is known as a failure mode, while the arising
consequences are called failure effects. This step is the most critical phase in the RCM
analysis as it provides the basic information for decision logic analysis as well as the
quality of the proposed preventive maintenance programme.
RCM decision logic and criticality analysis: It uses decision rationality of Yes and
No questions to find an optimal balance between the best maintenance tasks since
making a final judgment from old-style expert judgments in highway asset maintenance
is difficult. Maintenance tasks are chosen based on 8 decision logic questions as shown
in Table I and Table II with consideration of the criticality classes.

Pollack Periodica 13, 2018, 1

 RELIABILITY BASED MAINTENANCE METHODOLOGY 103

In consideration of the most appropriate and effective task, the Question 1-3 in
Table I is used in developing Table II bestowing the criticality analysis using (Y~Yes,
N~No and N/A~Not applicable). The effectiveness in Question 4-7 is utilized in
producing the criticality class (A~Safety & environment, B~Mission, C~All others
failures and D~Hidden Failure).

Table I
RCM decision logic
RCM DECISION LOGIC TABLE
Q Logic Decision
1 Is the incidence of a failure Yes = Go to Question 2 Evident Failure
evident to the working crew while
it is performing its normal duties? No = Go to Question 7 Hidden Failure
2 Does failure cause loss of function Yes = Go to Question 4 Safety Capability
damage that has a direct and
adverse effect on operating No = Go to Question 3 Operational
safety? Capability
3 Does failure have a direct and Yes = Go to Question 5 Operational
effect on operational capability? Capability
No = Go to Question 6 All Others
4- Is there an effective and applicable preventive maintenance task or a combination of
7 tasks that can prevent functional failures?
Effectiveness Rules
Q4 Safety and Environment Probability of failure reduced to very low
Q5 Mission Risk of failure reduced to acceptable level
Q6 All Others Cost of maintenance is less than cost of repair
Q7 Hidden Failure Consequences of hidden failure
Yes=Describe/Classify; No= Go to Q8
8 Is a scheduled failure finding task Yes = Specify task
available and justified? No = Consider safety design

Table II
Criticality analysis and class
CRITICALITY ANALYSIS CRITICALITY CLASS
Q1 Q2 Q3
Y Y N/A A=Safety/Environments
Y N Y B = Mission
Y N N C = All Other Function
N N/A N/A D = Hidden Failure

Maintenance task improvement: The maintenance task improvement comprises

developing Preventive Maintenance (PM) tasks and combining effective PM policy.
Evaluating existing maintenance classifications and intervals are useful in developing
PM programme for assets.

Pollack Periodica 13, 2018, 1

104 E. O. EKPIWHRE, K. F. TEE

2.2. Reliability analysis and growth curves modeling

The instantaneous and cumulative effect of failure rate for repairable fielded systems
on the road network depletes the reliability of road network systems. A repairable
system is a system, which after failing to perform one or more of its functions
satisfactory can be restored to entirely satisfactory performance [19]. The restoration
can be by ways and means other than replacement of the whole system. Repairable
systems are classified into three categories namely minimal repair, normal repair and
perfect repair based on the outcome on its repair level. The most common used models
for the failure process of repairable systems are:
I. Renewal Process (RP) or Homogeneous Poisson Process (HPP) model ~
perfect/maximal repair;
II. Generalized renewal process (correction and rejuvenation) ~ normal/partial
repair; and
III. Non-Homogeneous Poisson Process (NHPP) model ~ imperfect/minimal repair.
The organization chart in Fig. 3 portrays the fundamental relationship between
established classes of models and their dependence to failure intensities.

Fig. 3. Failure intensity relationship between models of repairable systems

In this critique, reliability growth for the improvement of repairable road pavement
systems is anticipated. Reliability behavior of highway asset is assessable using
reliability estimates of it Mean Time To Failure (MTTF) for instantaneous failure time
of the event and Mean Time Between Failure (MTBF) for cumulative times of events.
Although RCM achieved the maintenance interventions predictions for highway assets
via FMECA, quantitative judgments on the MTTF and MTBF were not captured as
RCM is more of a qualitative approach. In the representations of reliability growth
analysis by Crow and Duane are the most frequent amongst the foremost in proposing
the idea that improvements times can be represented mathematically. The prominent

Pollack Periodica 13, 2018, 1

 RELIABILITY BASED MAINTENANCE METHODOLOGY 105

relationship between both Duane and Crow-AMSAA model is their observed

cumulative MTBF and cumulative test logarithm time having a linear association [20],
[21]. This extension ability of the CROW-AMSAA methodology has allowed for
estimations in failure intensity and cumulative failures.
The growth trend and parametric growth curves of HPP and NHPP power law are
presented using Maximum Likelihood Estimation (MLE) and Least Square Estimation
(LSE) as well as their computed Mean Cumulative Function (MCF) of failure times of
events as expounded below.
Homogeneous Poisson Process (HPP): The Poisson process of the HPP has a
constant intensity function. The Poisson process is perceived not suitable for systems
that are either improving or deteriorating. Thus the model is only appropriate when the
intervals between failures do not systematically increase or decrease.

Non-Homogeneous Poisson Process (NHPP): The process can either be a power law
process or exponential law process. The intensity function of the systems represents the
rate of failures or repairs. Thus, the system reflects an improvement over the operational
time. When simulating a validation to track improvement, deterioration or stability of
the system, the NHPP power law process is classified as an AMSAA model if MLE
method is used and a Duane model if LSE method is utilized.
Mean Cumulative Function (MCF): The MCF portrays the average cumulative
number of failures or cost of the overall system in the investigated time interval.
Conceding that the resultant output of the MCF produced by the NHPP indicates an
increased, constant or decreased rate of system failure, the HPP has a constant failure
rate resulting in a straight line.
Trend test: Testing of trends in inter-failure times is a key aspect of the analysis of
failure time of event data for the repairable system. Event plots and Total Test of Time
(TTT) are possible to define the trends of inter failure times. The right censored datasets
are computable with any of the three most prolific trend tests namely; MIL-Hdbk-189
(the military handbook test), Laplace and Anderson-Darling.

2.3. Risk-based inspection and assessment

The importance of inspection and maintenance for system and assets of civil
engineering infrastructures is of high priority. The consequences could be fatal and
severe if poorly maintained and the most importantly, depreciation, which can be very
costly. The connection of quantitative risk analysis to maintenance has not been
effusively studied. Also, there is an absence of systemic, risk-based maintenance
methodologies that can solve the problems facing highway agency maintenance
programme. These inspections are anticipated to identify defects with the potential to
cause harm, danger or serious inconvenience to road users of the network and the
community environs. An onsite inspection is conducted to identify and assess the risk of
the defect, and after that based on the extent, defects are categorized into Category 1
[CAT.1] and Category 2 [CAT.2] with appropriate response time. The risk associated
with the hazards or dangers on the site are identified, and risk-based analysis is often

Pollack Periodica 13, 2018, 1

106 E. O. EKPIWHRE, K. F. TEE

developed to address the findings of the safety inspections [22]. CAT.1 defects are those
defects requiring urgent attention since they signify an immediate or imminent hazard or
because there is a risk of structural deterioration within a short period. CAT.2 defects
consist of all other defects deemed not to represent an immediate hazard as CAT.1 and
are further categorized based on priority namely High (H), Medium (M) and Low (L).
Techniques used in risk assessment and analysis in literature are numerous, unique
and suitable for different applications. The most common risk analysis are in two
categories namely deterministic (quantitative & qualitative) and stochastic (statistics &
forecasting), [23], [24]. The stochastic approach used is presented in Table III.

Table III
Stochastic techniques
Categories Analysis Techniques
Classic Statistics Probability Distribution - Exponential, Normal, Lognormal,
Approach (CSA) Weibull
Event data models -MTTF/Mean Time to Repair (MTTR) Model,
Time at Risk Failure Model, Poisson Model
Accident Time-Series, Markov Chain Analysis, Grey Model, Scenario
Forecasting Analysis, Regression Method, Neural Networks
Modelling (AFM)

The risk assessment framework based on Risk Based Inspection (RBI) intertwined a
Dual Stochastic (MTTR-PDF) approach. The algorithm in Fig. 4 evaluates the RBI
assessment to enable a stochastic application to pilot its outcomes. The approach
overcomes the precincts of inherent risk practices for defects on highway assets offering
a systematic and coherent way to manage the assets [22] as it is applied in case
study [10].
The consequence/impact analysis: The captured impact of the defected asset is
estimated as the scale of the defect to the asset and on the social-economic significance
of the asset status. The defect size and the impending level are used to factor a score.
The likelihood/probability analysis: This is used for scoring the defect after
inspection. The severity of the defect and the potential failure that could arise from the
defect as well as the assumed rate of deterioration is phantom.
Defect category analysis: The resultant scores from the consequence/impact and
likelihood/probability analysis in union with the risk register for consistency are used to
classify the defect as CAT.1 or CAT.2 defect to aid maintenance response prioritization.
Repair response analysis: The repair response is divided into response time scale
based on the risk level (e.g. 1 to 9, 1 to 25). A risk level at 1 indicates low priority and is
classified as the lowest impact while 9 or 25 as high requiring fastest response time.
Stochastic classical statistical approach: The combined stochastic process Mean
Time to Repair (MTTR)-Probability Distribution Function (PDF) accurately estimates
the maintenance interval outcome of the defected period. The mean (µ) and standard
deviation (ı) generated from MTTR interval from the sampled safety inspection data set

Pollack Periodica 13, 2018, 1

 RELIABILITY BASED MAINTENANCE METHODOLOGY 107

is sampled to derive the best fit PDF trend of each defect category. The best fit PDF is
generated using the most precise interval ȝ and ı, predicting the most current MTTR
interval for the various defect categories [10].

Fig. 4. Flowchart of RBI-Stochastic assessment framework

2.4. Strategic management of asset deterioration decision

The accurate predictions of the Current State (CS) and Forthcoming State (FS) of
highway infrastructures are crucial for developing appropriate inspection and
maintenance regimes for newly created or existing highway infrastructures. The
likelihood of optimal performance is the function of the CS (before treatment), the
treatment type and treatment cost in obtaining its FS (after treatment). The condition of
the CS of these repairable assets is the base unit for forecasting and implementing their
desired maintenance [12], [25], [26]. These attributes have gained importance because
asset owners and managers are interested in knowing the relationship between the CS of
their asset and the output of the maintenance process. A strategic ideal for asset
deterioration decision is modelled in Fig. 5. The definition of the attributes of the
strategic model is described in Table IV for ease of reference.
The application case procedure [14] as it is shown in the strategic model in Fig. 5,
displays the trail of the strategic cost model to enable assets owners to determine the
best time to carry out treatment for their assets. The performance condition requires
variables from the current state of the asset and the utilized performance transition in
computing its future state. In the case of the treatment renewal of the current condition
bands, the treatment cost of the repair level is required. The asset performance condition

Pollack Periodica 13, 2018, 1

108 E. O. EKPIWHRE, K. F. TEE

is classified using condition band (e.g. 1-Very Good, 2-Good, 3-Fair, 4-Poor and, 5-
Very Poor) as the starting point CS for subsequent planning for the network.

Fig. 5. Strategic model for asset deterioration transition and decision

Table IV
Attributes definition for strategic model
i Current state: The state of which the asset represents before modeled for a
forthcoming state
ii Untreated quantity: The portion of an asset having no treatment intervention
iii Performance transition: The transition probability matrix in a deterioration context
iv Forthcoming state: The projected state from the current state of treated or untreated
quantity
v Performance condition: The forthcoming condition of the current state having no
treatment
vi Budget renewals: The annual budgeted figure needed for each of the treatment
vii Treated quantity: The portion of an asset that treatment intervention was conducted
viii Treatment transition: The transition probability matrix in an improvement context
ix Treatment renewals: The forthcoming condition of the current state having
treatment action
x Forthcoming state: The projected state from the current state of treated or untreated
quantity
xi Forthcoming category: The projected state from the current state of treated or
untreated quantity

Pollack Periodica 13, 2018, 1

 RELIABILITY BASED MAINTENANCE METHODOLOGY 109

The estimation for CS is imperative as it represents the definition of the asset

condition used in describing the condition bands. In the revolvement towards the FS
from the starting point, transition probability from two consecutive inspection periods
having no maintenance (untreated) or maintenance (treated) intervention is essential to
obtain the respective performance transition or treatment transition [14]. These helps in
determining the performance condition and treatment renewal knowledge required for
an optimized decision making of when the best to carry out maintenance activity.

3. Application to carriageway

The results from the application of the proposed models are used to develop
appropriate maintenance program for carriageway in order to ensure continuous
functionality and availability. The proposed method is applied and evaluated using
condition improvement between the two successive inspections from the Surface
Condition Assessment of National Network of Roads (SCANNER) survey of the United
Kingdom Pavement Management System. SCANNER surveys are used for the
collection of data for carriageway surface condition. It is developed on behalf of the
Department for Transport (DfT) to provide an accurate, consistent approach for
assessing the condition of all principal roads across the United Kingdom.
The result depicts the change that occurs at different condition states of the
carriageway. The deterioration profile is illustrating the 20 years spread between the
various condition states. The deterioration levels after every five years are presented
in Fig. 6.

Fig. 6. Predicted future year (FY) deterioration for 5, 10, 15 and 20 years

Pollack Periodica 13, 2018, 1

110 E. O. EKPIWHRE, K. F. TEE

The predicted condition state could assist asset managers to understand the possible
deteriorations of the carriageway and help to identify the best time to carry out
maintenance actions. Asset owners look forward to ensuring that the carriageway is
above the fair level. This will enable them to avoid reactive maintenance and provide
more preventive and condition based maintenance. The Markov chain model still
happens to be one of the most appropriate tools for predicting the future condition of
physical assets.

4. Conclusion
The research aims to propose methodologies of a reliability based asset management
for highway infrastructures. The methods presented in the paper enable highway asset
owners, and managers to develop reliability based maintenance strategies and to
maintain newly created and already existing fielded highway infrastructures. In the
maintenance of fielded highway assets, multiple maintenance methodologies are to be
taking into applications. This quad process would help to ensure that the maintenance of
the various aspects of the asset lifecycle is well-thought-out. The quad approaches
developed for the asset maintenance in this paper are in themselves valuable means for
asset owners and maintenance managers. The methods are useful in determining
appropriate maintenance type and time, thereby creating an excellent platform for
decision making. The methodology and processes developed in the paper have the
capability to support organization in enhancing their well-established maintenance
programs. The methods are organized to follow the generalized principle of reliability
based maintenance and should allow asset managers to implement them impeccably in
conjunction with their existing processes.

Acknowledgements
This work has been undertaking on as a part of a project funded by the Niger Delta
Development Commission (NDDC), Port Harcourt, River State, Nigeria and the Delta
State Government Scholarship and Bursary (DTSG), Asaba, Delta State, Nigeria.

References
[1] County Surveyors Society, Framework for highway asset management, City Surv Soc,
London, The Stationary Office, 2004.
[2] UK Bridges Board, Management of highway structures, A code of practice, London, The
Stationary Office, 2005.
[3] Frangopol D. M., Mohamed S. Life-cycle of structural systems: Recent achievements and
future directions, Structure and Infrastructure Engineering, Vol. 12, No. 1, 2016, pp. 1−20.
[4] UK Roads Board, Well- maintained highways, Code of practice for highway maintenance
management, London, The Stationary Office (TSO), 2005.
[5] Economic Development Research Group, Failure to act, The impact of current
infrastructure investment on America’s economic future, Virginia, 2013.

Pollack Periodica 13, 2018, 1

 RELIABILITY BASED MAINTENANCE METHODOLOGY 111

[6] Department for Transport, Highways Agency, 2010 to 2015 government policy, road
network and traffic, 2015, https://www.gov.uk/government/publications/2010-to-2015-
government-policy-road-network-and-traffic (last visited 25 July 2015).
[7] Ekpiwhre E. O., Tee K. F. Reliability-centered maintenance of road junction transport
assets, In: Kruis J, Tsompanakis Y, Topping B. H. V. (Eds) Proc. of Fifteenth Int. Conf.
Civil, Struct. Environ. Eng. Comput. Prague, Czech Republic, 1-4 September 2015, Civil-
Comp Press, 2015, Paper 279.
[8] Mariut L., Helerea E., Felea I. Underground power cables-life analysis and reliability
prognosis, Pollack Periodica, Vol. 7, No. 3, 2012, pp. 3−14.
[9] Ekpiwhre E. O., Tee K. F. Reliability based maintenance for sustainable transport asset
management, In: Iványi P., Bachmann B. Kvasznicza Z. (Eds.) 12th Miklós Iványi Int. PhD
DLA Symp, Pollack Press, Pécs, Hungary, 3-4 November 2016,
p. 38.
[10] Ekpiwhre E. O., Tee K. F., Aghagba S., Bishop K. Risk-based inspection on highway assets
with category 2 defects, Int J Safety Security Eng, Vol. 6, No. 2, 2016, pp. 372–382.
[11] Washer G., Connor R., Nasrollahi M., Reising R. Verification of the framework for risk-
based bridge inspection, Journal of Bridge Engineering, Vol. 21, No. 4, 2016,
paper 04015078.
[12] Ekpiwhre E. O., Tee K. F., Mordi O., Bull T. Carriageway deterioration prognosis
modelling using Markovian chain, In: Scarf P., Wu S., Do P. (Eds) Proc.of 9th IMA Int.
Conf. Model. Ind. Maint. Reliab, London, UK, 12-14 July 2016,
pp. 58–63.
[13] Wellalage N. K. W., Zhang T., Dwight R. Calibrating Markov chain–based deterioration
models for predicting future conditions of railway bridge elements, Journal of Bridge
Engineering, Vol. 20, No. 2, 2015, paper 04014060.
[14] Ekpiwhre E. O., Tee K. F. Cost modelling of carriageway treatment transition for strategic
maintenance optimisation, In: Institute of Asset Management, Institute of Engineering &
Technology, (Ed.) Profs. of Asset Manag. Conf, London, UK, 23-24 November 2016,
paper 7.a.2.
[15] Kota L., Jarmai K. Efficient algorithms for optimization of objects and systems, Pollack
Periodica, Vol. 9, No. 1, 2014, pp. 121−132.
[16] Khan L. R., Tee K. F. Risk-cost optimization of buried pipelines using subset simulation,
Journal of Infrastructure Systems, ASCE, Vol. 22, No. 2, 2016, 04016001.
[17] Orugbo E. E, Alkali B., DeSilva A., Harrison D. K. RCM and AHP hybrid model for road
network maintenance prioritization, Baltic Journal of Road & Bridge Engineering, Vol. 10,
No. 2, 2015, pp. 182–190.
[18] US Navy Sea Systems, Reliability-centered maintenance (RCM) handbook, vol. S9081–
AB–G. U.S Navy Sea Systems Command, 2007.
[19] Ascher H, Feingold H. Repairable systems reliability, modeling, inference, misconceptions
and their causes, NY, M Dekker 1984.
[20] Crow L. H. Reliability analysis for complex, repairable systems, technical report No. 10,
US Army Materials Alaysis Activity, 1975.
[21] Duane J. T. Learning curve approach to reliability monitoring, IEEE Trans Aerosp, Vol. 2,
No. 2, 1964, pp. 563–566.
[22] Transport for London good practice guide, Risk based inspection for highway structures,
Attkins, London, 2011.
[23] Fang Y., Xiong J., Tee K. F. Time-variant structural fuzzy reliability analysis under
stochastic loads applied several times, Structural Engineering and Mechanics, Vol. 55,
No. 3, 2015, pp. 525–534.

Pollack Periodica 13, 2018, 1

112 E. O. EKPIWHRE, K. F. TEE

[24] Khemis A., Hacene-Chaouche A., Athmani A., Tee K. F. Uncertainty effects of soil and
structural properties on the buckling of flexible pipes shallowly buried in Winkler
foundation, Structural Engineering and Mechanics, Vol. 59, No. 4, 2016, pp. 739–759.
[25] Zhang Y., Kim C. W., Tee K. F. Maintenance management of offshore structures using
Markov process model with random transition probabilities, Structure and Infrastructure
Engineering, Vol. 13, No. 8, 2017, pp. 1068−1080.
[26] Fang Y., Tao W., Tee K. F. Time-domain multi-state markov model for engine system
reliability analysis, Mechanical Engineering Journal, Vol. 3, No. 3, 2016, paper 16-00084.

Pollack Periodica 13, 2018, 1