a critical review of prioritization models for pavement ... · this will help in evaluating the...
TRANSCRIPT
2nd International Conference on Emerging Trends in Engineering & Technology, April12, 13, 2013
College of Engineering, Teerthanker Mahaveer University.
1
A Critical Review of Prioritization Models for Pavement
Maintenance Management Decisions
Yogesh U.Shah1, Dr. S.S. Jain
2, Dr. M.K. Jain
3, Dr. Devesh Tiwari
4
A primary purpose of a pavement management system (PMS) is to provide information so that roadway
improvements can be priority ranked. Ideally, prioritization is a consistent and justifiable process. It should involve
minimizing life cycle costs subject to minimum levels of serviceability and budget constraints. Prioritization is a
complicated process that requires sound engineering judgment and a good understanding of local conditions.
Current fiscal crises and rising roadway improvement costs have made prioritization decisions more important than
ever. Priority analysis is a systematic process that determines the best ranking list of candidate sections for
maintenance based on specific criteria such as pavement condition, traffic level, pavement functions, etc. Various
methods are used for priority analysis ranging from simple listing based on engineering judgment to true
optimization based on mathematical formulations.
This paper examines theoretical and pragmatic problems surrounding the prioritization process. This study
report a detailed review of various prioritization techniques and models developed for flexible pavements at global
level. This will help in evaluating the usefulness of the various models in some particular condition having the
similar prioritization parameters. A discussion on the limitations of the different models is also given in this study.
Keywords: Pavement management system, prioritization models.
1. Introduction
Pavements are the important component of the inland
transport system. It‟s a challenging job for engineers
and scientists to augment the performance of roads to
meet the needs of growing urban communities. It is very
important to maintain the existing pavement network in
its serviceable condition within the available resources.
And this can be achieved by adopting appropriate
maintenance strategy for the road network under
consideration. Managing the pavements is a systematic
approach, which includes different activities like
pavement performance evaluation, determination of
maintenance and rehabilitation (M&R) requirements,
optimization of resources and prioritization of pavement
sections for timely and resource effectively maintenance.
These activities collectively create the pavement
management system (PMS).
Priority ranking, as used in PMS, is a process used
to rank the pavement sections in an order of urgency for
maintenance and repair. The prioritization process is the
main step of PMS, before the decision maker‟s takes
final decision on execution of maintenance program.
1. Research Scholar, Centre for Transportation Systems
(CTRANS), Indian Institute of Technology Roorkee. Email:
2. Professor & Coordinator of Transportation Engg. Group,
Civil Engg. Dept., IIT Roorkee.
3. Assistant Professor, Dept. of Hydrology, IIT Roorkee.
4. Principal Scientist, Pavement Evaluation Division, Central
Road Research Institute of Technology (CRRI), New Delhi.
The quality of priority-setting is directly influencing the
effectiveness of available resources which are, in most
cases, the primary judgment of the decision maker
(Sharaf & Mandeel 1998). The priority ranking process
depends on various factors like pavement condition,
traffic volume, environmental effects, desired
performance standards, and budgetary constraints. Since
maintenance actions affect the scheduling of work and
allocation of resources, proper selection of such actions
(priorities) is crucial to the most efficient use of limited
resources. In the present study an effort has been made
to broadly classify various approaches/methods for
prioritization of pavement maintenance and the
applications of these models made by different
researchers at global level. A discussion on the
limitations of the different models is also given in this
study.
2. Prioritization Methods / Approaches
Priority setting techniques as used in the PMS cover a
wide spectrum of methods and approaches ranging from
simple priority lists based on engineering judgment to
complex network optimization models as shown in
Table 1 (Haas et al. 1994). These prioritization methods
can be further divided as: (i) Ranking Methods (ii)
Optimization Methods (iii) Artificial Intelligence
Techniques (iv) Analytical Hierarchy Process Method.
2nd International Conference on Emerging Trends in Engineering & Technology, April12, 13, 2013
College of Engineering, Teerthanker Mahaveer University.
2
Table 1. Different classes of priority programming
method
S.
N.
Class of Method Advantages and
Disadvantages
1 Simple subjective
ranking of projects
based on judgment,
overall condition
index, or decreasing
first cost
Quick, simple; subject to
bias and inconsistency;
may be far from optimal
2 Ranking based on
parameters, such as
serviceability or
distress; can be
weighted by traffic
Simple and easy to use;
may be far from optimal,
particularly if traffic
weighting not used
3 Ranking based on
condition analysis
and traffic, with
economic analysis
Reasonably simple; may
be closer to optimal
4 Annual optimization
by mathematical
programming model
for year-by-year basis
over analysis period
Less simple; may be
close to optimal; effect of
timing not considered
5 Near optimization
using heuristics
including benefit-cost
ratio and marginal
cost effectiveness
Reasonably simple;
suitable for
microcomputer
environment; close to
optimal results
6 Comprehensive
optimization by
mathematical
programming model
taking into account
the effects of M,R&R
timing
Most complex and
computationally
demanding; can give
optimal program
(maximization of
benefits)
(Source: Haas et al. 1994)
2.1 Ranking Methods
2.1.1 Composite Index Ranking Method
The ranking of pavement sections for maintenance is
done on the basis of the priority index calculated by
combining different pavement indices. These indices are
estimated by considering parameters like pavement
distresses, riding quality, traffic conditions, economic
analysis, functional class, accident details, geometric
deficiencies, structural capacity, skid resistance,
pavement age, engineering judgement, etc.
The prioritization program should not only be based
on current pavement condition, for making it efficient
future pavement condition should also be considered.
As generally the treatments are applied at least one year
after the condition surveys are carried out, giving the
time to require for relative prioritization and organizing
for funds.
There are different approaches to develop a priority
combined index for pavement maintenance. These
approaches include unique sum approach, utility theory,
Delphi method, factorial rating method, and fuzzy set
theory. The application of few methods is discussed in
following sections.
2.1.2 Economic –based Methods
The prioritization methods based on economic analysis
can be of two types: (i) using optimal benefit/cost ratio
(ii) using incremental benefit/cost ratio. In the first
method, prioritization process uses the optimal M&R
recommendations and corresponding benefit/cost ratios
(or effectiveness/cost (E/C) ratio) for each pavement
section of the network produced from the dynamic
programming. The higher the E/C ratio of a section, the
higher the priority of that section for repair. The
available budget is allocated to the pavement sections as
per the priority list till the budget is completely
exhausted.
The second methodology is a heruistic method for
budget optimization. In this mehtod all feasible M&R
alternatives of a section are identified and the
coressponding inflated initial cost, present-worth costs,
and weighted benefits are obtained. This information is
then used in the program to produce optimal M&R
recommendations for each pavement section, including
initial cost and type of treatment. The budget
optimization also gives the total network-weighted
benefits corresponding to optimal M&R
recommendations (Butt et al. 1994).
2.2 Optimization Methods
Priority programming by optimization is conceptually
different from others methods. It combines the function
of priority programming, program formulation, and
project scheduling into one operation which gives the
optimum schedule of projects through precise analytical
techniques such as linear and dynamic programming.
Generally, these method uses maintenance cost
minimisation or maintenance benefits maximization to
generate the optimal maintenance plans.
2.3 Artificial Intelligence Techniques
Artificial intelligence techniques include fuzzy
mathematical programming, ANNs, and evolutionary
computing (including genetic algorithms). These
techniques are particularly appropriate for pavement
management because the information may be uncertain
and incomplete. The data may involve combinations of
objective measurements, subjective rating, and expert
2nd International Conference on Emerging Trends in Engineering & Technology, April12, 13, 2013
College of Engineering, Teerthanker Mahaveer University.
3
inputs, such as those data used to create M, R&R
decision criteria and policy tables. Artificial neural
networks and fuzzy systems have been used for needs
analysis as alternatives to the traditional priority ranking
tools, such as decision trees.
2.4 Analytical Hierarchy Process Method
Analytical hierarchy process (AHP) is one of the multi
criteria decision making methods (MCDM) used to
scale and quantify measurements. AHP theory was
developed by Saaty in 1970. Fig. 1 shows a typical
schematic hierarchy of prioritization process. The first
step is the formation of the problem components
(hierarchy); the second phase of the AHP is the
evaluation which is based on the concept of paired
comparisons. The parameters considered at each level
of hierarchy are compared in relative terms as their
importance or contribution to a given criterion. This
process of comparison yields a relative scale of
measurements of priorities or weights of the elements.
These relative weights sum to unity. For project-level
evaluations where a few sections are to be considered
simultaneously, AHP is an effective method for
analysis.
Fig. 1 Typical hierarchy structure for prioritization
3. Studies on Pavement Maintenance
Prioritization
3.1 Based on Subjective Ranking
Fwa and Chan (1993) illustrated the feasibility of using
neural network models for priority assessment of
highway pavement maintenance needs. A simple back-
propagation neural network with three different priority
setting schemes viz. a linear function relating priority
ratings to pavement conditions, a nonlinear function,
and subjective priority assessments obtained from a
pavement engineer, was tested using general purpose
micro-computer based neural network software known
as Neural Works. The validation analysis was carried
out for all three schemes as described below:
Chen et al. (1993) developed a prioritization procedure
as a part of an Urban Roadway Management System
(URMS) at network level. The procedure combined two
matrices and an equation for computing priority index
(PIX). PIX was taken as a function of PCI, pavement
age, mixed traffic and street class. The process is shown
in Fig. 2.
(Source: Chen et al. 1993)
Fig. 2 Priority index procedure used in URMS
Sharaf (1993) compared three methods for priority
ranking. In the first model the priority index was
estimated as follows:
(1)
The traffic levels were classified as less than 2500
vehicle per day (vpd), between 2500-10,000 vpd, and
more than 10,000 vpd and the corresponding traffic
factors were selected as 1.0, 0.5 and 0.1 respectively.
The defect factor (values between 0.1 to 1.0) was
assigned on the basis of the distress type and required
treatment (lower values for major treatments). The
sections were then ranked in descending order after
calculating the section priority index, and converted to a
cost list by using appropriate maintenance treatment
unit cost. The second model was a modification of the
first, where the pavement sections were arranged
according to road type (i.e. desert or agricultural roads)
and traffic level into four classes. Also in this model the
budget shares were reserved for different maintenance
treatments. In the third model, the distribution of budget
among all the sections was done on the basis of
optimization method.
Jain et al. (1996) selected four sections of flexible
pavement on NH-21 & NH-22 in state of Himachal
2nd International Conference on Emerging Trends in Engineering & Technology, April12, 13, 2013
College of Engineering, Teerthanker Mahaveer University.
4
Pradesh and eight sections in state of U.P. on NH-2,
SH-45 & MDR-121 for maintenance prioritization.
Various field studies like measurement of surface
deflection, roughness, rut depth, cracks & cracking
patterns, pot holes and axle load studies were performed
on all pavement sections. Based on the collected data
M&R investment strategy for design life of 5 years was
suggested. Priority fixation for M&R was done based on
following factors:
i. Importance of road (i.e. road type and commercial
traffic);
ii. Characteristic deflection (Dcs) value;
iii. Road condition based on roughness, cracks, rutting
and potholes;
iv. Investment needs for maintenance and
rehabilitation.
Suma et al. (2000) developed a simplified ranking
methodology based on distresses to prioritise the
pavements for maintenance management. Data
pertaining to pavement distress viz., cracking and
patching along with their severity levels, unevenness
(roughness), and pavement condition ratings were
collected and weightages were assigned to each type of
distress based on the severity level and extent and
deduct value curves were developed. Pavement
Condition Indices were then determined and the
pavement sections were ranked for maintenance.
Agarwal et al. (2004) presented a rational approach
using AHP for prioritization of highway sections for
maintenance on the basis of present highway condition,
future highway condition and the highway importance.
The functional condition of highway sections was
evaluated considering the (i) safety (ii) efficiency of
traffic operation and (iii) riding comfort. The relative
weights of these factors have been conceived from the
expert opinion of field engineers which were analysed
using AHP. For evaluation of structural condition a
simple and cost effective statistical model based on
Long Term Pavement Performance (LTPP) database of
US Department of Transportation was developed. The
highway importance is assessed considering the (i)
highway class (ii) importance to community and (iii)
political importance. The relative weights of these
factors have also been conceived from the expert
opinion of field engineers and using the AHP. A total of
15 factors are considered and identified by thicker
borderline of the text boxes in Fig.3. The developed
approach was employed for prioritizing the sections for
maintenance in a small highway network.
(Source: Agarwal et al. 2004)
Fig. 3 Decomposition of the maintenance priority problem into a hierarchy
Chandran et al. (2007) formulated the prioritization
technique of low-volume pavement sections for
maintenance using fuzzy logic. Eight pavement sections
each of 500m length with different deterioration levels
Overall Highway Condition
Maintenance Priority
Overall Highway Importance
Percentage Highway
Condition
Future Highway
Condition
Present functional
Condition (PFC)
Present Structural
Condition (PSC) 6
Hwy Class
(HC) 13
Imp. To Community
(IC) 14
Political
Imp. (PI) 15
Future functional
Condition (FFC)
Future Structural
Condition (FSC) 12
Present Fatigue
Cracking
Present Fatigue
Cracking
Present
Traffic
Safety (PTS) 1
Present Efficiency
of traffic
Operation (PTO) 2
Present
Riding
Quality
Present
Fractured
Surface (PFS) 3
Present
Distorted
Surface
(PDTS) 4
Present
Disintegrated
Surface (PDIS) 3
Future
Traffic
Safety (PTS) 7
Future Efficiency
of traffic
Operation (PTO) 8
Future
Riding
Quality
Future
Fractured
Surface (PFS)
9
Future
Distorted
Surface
(PDTS) 3
Future
Disintegrated
Surface
(PDIS) 11
Future Fatigue
Cracking
Future Fatigue
Cracking
2nd International Conference on Emerging Trends in Engineering & Technology, April12, 13, 2013
College of Engineering, Teerthanker Mahaveer University.
5
were selected in the Trivandrum and Kollam Districts in
Kerala State. Fuzzy condition indices (FCI) were used
to prioritize the pavement sections by suitable fuzzy
ranking methods. Fuzzy membership functions were
formulated for severity, extent, and relative importance
of each distress with respect to maintenance. The
common index aggregating above parameters in form of
fuzzy condition index (FCI) was computed using
following expression:
(2)
Where wki = subjective weight of distress k at severity
level i, Aki = subjectively assessed extent of distress k at
severity level i, and Si = subjective assessment of
severity level i.
Fuzzy condition indices were developed for each
pavement section by using MATLAB software. The
FCI so obtained for various sections are shown in Fig. 4.
The ranking based on PCI values were also compared
with FCI ranking.
(Source: Chandran et al. 2007)
Fig. 4 Fuzzy condition indices for various sections
Sandra et al. (2007) proposed a Fuzzy Multi Criteria
Decision Making (FMCDM) approach for prioritizing a
pavement stretches. An expert opinion survey was
conducted to quantify the influence of various distresses
with respect to their extent and severity on the
functional condition of the pavement. For collecting
pavement condition data, four different stretches
dominated by different types of functional failures have
been chosen in the state of Rajasthan, for a total length
of 25km, representing different functional classes of
highways viz. National Highways (NH), State Highways
(SH) and Major District Roads (MDR).
Priority Index (PI) for all the pavement stretches was
computed from the fuzzy preference relation matrix
using the following mathematical form:
– (3)
Where, eij = real number indicates the degree of
preference between the respective ith and j
th pavement
stretches. It has been calculated using positive (Sij+) and
negative areas (Sij−) of difference between two fuzzy
values ( ). The eij was calculated using following
equation and Fig. 5.
(4)
Where, ( = Total area of ( )
(Source: Sandra et al. 2007)
Fig. 5 Computation of
The road link which had highest Priority Index (PI) was
given top priority and vice versa. A code was developed
in MATLAB in order to calculate the PI for the present
study.
Zhang (2004) presented an Analytic Hierarchy Process
(AHP) based quantitative approach to the prioritization
of data requirements for pavement management. Using
the Weighted Sum technique, the importance and
frequency of usage of data items were combined to
yield their comprehensive ranking score. The overall
methodology is illustrated in Fig. 6. The process of
establishing the hierarchy to prioritize pavement data
collection using AHP is illustrated in Fig. 7.
Farhan & Fwa (2009) made an attempt to explore the
use of an analytic hierarchy process (AHP) for the
prioritization of pavement maintenance activities. Three
forms of AHP were examined, namely, the distributive-
mode relative AHP, the ideal-mode relative AHP, and
the absolute AHP. For analysis purpose, three road
functional classes, three distress types and three levels
of distress severity were considered which gave 27
possible combinations of maintenance treatments. The
hierarchy structure used for the AHP analysis is as
shown in Fig. 8. The study concluded that the absolute
AHP was suitable for the pavement maintenance
prioritization process, on the basis of its ability to
provide priority assessments for pavement maintenance
activities in good agreement with the priority
assessments obtained by the direct assessment method
and its operational advantage in evaluating a large
number of maintenance activities.
2nd International Conference on Emerging Trends in Engineering & Technology, April12, 13, 2013
College of Engineering, Teerthanker Mahaveer University.
6
(Source: Zhang 2004)
Fig. 6 Overall process of the methodology
Moazami et al. (2010) generated AHP and Fuzzy logic
modelling for prioritization of pavement rehabilitation.
The developed models were executed for 131 sections
of urban streets of district No. 6 of Tehran municipality.
The modelling was done in two steps, in the first step
analytical hierarchy process was employed to compute
the relative weights of parameters. The pairwise
comparison between the following criteria and sub-
criteria as given in Table 2 was done by considering
ideas of about 200 PMS experts. Finally, with rating
approach in Expert Choice software prioritization was
done.
In the next step, fuzzy logic modelling was used to
obtain satisfactory precision. The three fuzzy sets were
defined and following Gaussian membership function
were used for demonstration of being high, medium and
low fuzzy sets.
(5)
where, μ(X) = Value of Membership of Variable X to
Fuzzy Set, G = Mean of Gaussian Function , σ =
Gaussian Function Bell Radius. The Singleton Fuzzifier,
a product inference engine was used to calculate the
priorities of the sections using fuzzy logic. The analysis
was done using the fuzzy toolbox of MATLAB 7.0
software and coded M-files.
It was concluded that, in AHP pairwise
comparison between criteria and sub-criteria is arbitrary,
while gauss mean and sigma matrixes constitutions are
based on reality. Also it was shown that AHP method
cannot differentiate between two almost identical
sections. Consequently, fuzzy logic programming was
considered as the best choice in the study.
(Source: Zhang 2004)
Fig. 7 Application of AHP to prioritizing pavement
data collection
3.2 Based on Composite Condition Index
Karan (1981) described the pavement management
implementation project for Prince Edward Island (PEI),
Canada covering a road network of 500km. It included
the field inventory, identification of needs for pavement
improvements, evaluating the rehabilitation alternatives,
priority analysis and budget analysis. The Condition
Index (CI) was computed considering the severity of 10
different distresses on road sections. The overall
Serviceability Index (SI) was calculated for all road
sections using following expression:
SI = a (RCI) + b (SAR) (6)
Efficient and Cost
Effective Pavement
Management
Aiding in
Policy
Decisions
Structural
Adequacy
Enhancement
Serviceability
Enhancement
Safety
Improvement
Pavement Management User
Needs
Identification of Core Pavement
Management
Activities to be Conducted
Grouping of Similar Core
Pavement Management
Activities
Use f Analytic Hierarchy Process
Priorities of Group of Core
Pavement Management
Activities
Identification of Core
Pavement of
Management Activities
for Each
Pavement Life - Stage
Grouping of
Similar Activities
Use of AHP to
Arrive at Priorities
of Activity Groups
Use of Delphi
Process to Arrive at
Consensus of opinion
about Priorities of
Activity Groups
Final Priorities of
Activity Groups
Adjusted Priorities of
Activities
Use of Weighted Sum Technique
to combine Frequency of Usage
of Data Items and Their
Importance
Comprehensive
Ranking of Data Items
Scoring of
Activities in
Each Activity
Group Based
on Levels of
Importance
Identification
of Major
Categories of
Pavement
Data Items
Frequency
Analysis of
Major
Categories of
data Items
for Identified
Activities
The Average
Score of Each
Activity
Multiplied with
Respective
Activity group
Priority
2nd International Conference on Emerging Trends in Engineering & Technology, April12, 13, 2013
College of Engineering, Teerthanker Mahaveer University.
7
where, SI = serviceability index (Rating Scale: 0-10, 0 –
totally unacceptable pavement, 10 - perfectly smooth
and strong pavement), RCI = Riding Comfort Index
(Rating Scale: 0 – 10, 0 - worst riding comfort & 10 –
perfectly smooth pavement), SAR = Structural
adequacy rating (Rating Score: 0 – 10, 0-4.9 indicates
structural adequate pavement, 5-10 indicates structural
inadequate pavement), a, b = weighting factors (a + b =
1.0)
The calculated values of CI, RCI, SAR, SI was given as
an input for development of PMS at network level. One
of the components of PMS was establishing the
pavement improvement priorities. A mathematical-
optimization (linear-programming) model was adopted
to set priorities based on benefit maximization and
budget constraints.
(Source: Farhan & Fwa 2009)
Fig. 8 Hierarchy structure for AHP analysis
Table 2. Criteria and sub-criteria for AHP process Level - 1 Criteria PCI Types of Urban Roads
(Road Width m – both
direction)
Traffic Volume (pcu/hr/
direction)
M & R Cost (USD per red)
Level - 2 Sub-
Criteria
0 – 10 : Failed
10 – 25 : Very Poor
25 – 40 : Poor
40 – 55 : Fair
55 – 70 : Good
70 – 85 : Very Good
85 – 100: Excellent
Local (5.5 - 6.5)
Minor Arterial (5.5 - 21)
Major Arterial (13 - 30)
Low Volume (<433)
Medium Volume (433-2660)
High Volume (2660 – 6200)
Others (> 6200)
Very Low (< 2000)
Low (2000-20000)
Medium (20000-50000)
Omit High (50000-100000)
Very High (>100000)
(Source: Moazami 2010)
Selection of a Pavement Section for Maintenance
Expressway Arterial Access
Pothole Rutting Cracking
High Moderate Low
Objective (Level 1)
Criteria (Level 2)
Sub-Criteria (Level 3)
Alternatives (Level 5)
Sub Sub-Criteria
(Level 4)
Section 1 Expressway
Pothole
High
Section 2 Expressway
Pothole
Moderate
Section 3 Expressway
Pothole
Low
Section 4 Expressway
Rutting
High
Section 5 Expressway
Rutting
Moderate
Section 6 Expressway
Rutting
Low
Section 7
Expressway
Cracking
High
Section 8
Expressway
Cracking
Moderate
Section 9
Expressway
Cracking
Low
Section 10
Arterial
Pothole
High
Section 11
Arterial
Pothole Moderate
Section 12
Arterial
Pothole Low
Section 13
Arterial
Rutting
High
Section 14
Arterial
Rutting Moderate
Section 15
Arterial
Rutting Low
Section 16
Arterial Cracking
High
Section 17
Arterial Cracking
Moderate
Section 18
Arterial Cracking
Low
Section 19
Access
Pothole
High
Section 20
Access
Pothole
Moderate
Section 21
Access
Pothole
Low
Section 22
Access
Rutting
High
Section 23
Access
Rutting
Moderate
Section 26
Access
Rutting
Low
Section 27
Access Cracking
High
Section 28
Access Cracking
Moderate
Section 29
Access Cracking
Low
2nd International Conference on Emerging Trends in Engineering & Technology, April12, 13, 2013
College of Engineering, Teerthanker Mahaveer University.
8
Zhang et al. (1993) developed a comprehensive ranking
index for flexible pavements using Fuzzy sets theory.
The model developed was named as overall
acceptability index (OAI) is given as follows:
(7)
Where W1, W2, W3 & W4 are the weighing factors for
roughness, distress, structural capacity and skid
resistance respectively. The sum of all weighing factors,
Wi = 1.0. The membership functions for each
parameter as developed using nonlinear regression
analysis for a subjective opinion survey about the level
of acceptance for selected pavement attributes and their
relative importance, are given in Table 3.
Table 3. Membership functions and weights for
secondary roads
Attribute Membership
Function R
2 Weight
Roughness
(PSI)
A1 = 1 - exp (-
0.01274*PSI4)
0.970 0.306
Distress
(D)
A2 = exp (-
0.00000185*D2)
0.971 0.244
Structural
Capacity
(SC)
A3 = 1-exp (-
0.207*(CS/50)5)
0.960 0.225
Skid
Resistance
(CF)
A4 = -0.22 +
1.6308*CF 0.979 0.231
(Source: Zhang et al. 1993)
Andres (1994) details the prioritization methodology of
local roadway improvements for the town of Eastharm,
Massachusetts. The system used a composite
performance based index to assess priorities. The index
was formed by first determining an effectiveness ratio.
This is accomplished by dividing “benefits” (areas
under deterioration curves after treatment is assigned)
by the cost per square yard of treatments. This
effectiveness ratio was then weighted by functional
class. This measure of the influence of traffic volume on
priority assessment can be modified by user. The
weighted effectiveness ratio was then used to prioritize
road segments in each category. The prioritization was
done considering different budget scenarios like zero
funding, full funding, moderate budget cut and severe
budget cuts.
Flintsch et al. (1998) presented the formula to prioritize
pavement rehabilitation projects based on expert‟s
opinion. It was used for the 5 yr pavement preservation
program of Arizona Department of Transportation
(ADOT). The preliminary list of candidate projects
were prioritized using a “rate” number computed
according to the following equation:
Rate = Cr + 0.2*Rg + 2*Rut + 0.0015*MC (8)
where; Cr = cracking (%), Rg = roughness (Maysmeter
units), Rut = rutting (in.), and MC = average
maintenance cost for last 3 years (dollars per year).
Pavement experts were instructed to examine carefully
all pavement sections presented in the questionnaire and
to assign a priority rating to each of the sections using a
scale from 1 to 10. A priority of 1 indicated the
pavements most in need of a preservation treatment.
The average response for each section was used for the
analysis. The average proposed treatment was computed
by averaging the numeric values corresponding to all
treatments recommended for each section. The average
was rounded and converted back to a treatment
recommendation. The average priority values computed
for each pavement section were used to obtain an
alternative prioritization formula. The following model
was selected to calculate the priority for pavement
sections:
Priority = 4.82 – 0.83A – 0.64B – 0.50C – 0.82D
– 0.81E – 1.11F (9)
Where, Priority = priority value, A = functional
classification code (Inter-state or State route), B =
traffic code, C = maintenance cost code, D = roughness
code, E = cracking code, and F = rutting code.
This model includes all the variables in the rate formula
plus two more: roadway classification and rutting.
A model for selection of maintenance treatment to
pavement sections was also developed as shown below:
Treat = 2.28 + 0.33A + 0.14B + 0.32D + 0.34E + 0.56F
-0.06G – 0.26EF (10)
Where, Treat = type of treatment, A = functional
classification code, B = traffic code, D = roughness
code, E = cracking code, F = rutting code, and G =
structural number code.
The result of the analysis were then implemented in a
computer program that automatically computes the
recommended treatment for each homogeneous
pavement section in the preliminary list of candidate
sections using a default assignation matrix that can be
modified by the user. The program also assigned a
priority value based on the alternative pavement
prioritization formula (priority) and sorts the projects by
priority within each roadway group (roadway
classification, traffic, and region). The sections with the
highest priority within each group are funded until they
2nd International Conference on Emerging Trends in Engineering & Technology, April12, 13, 2013
College of Engineering, Teerthanker Mahaveer University.
9
reach the budget recommendation provided by the
network optimization system, increased by a coefficient.
Ramadhan et al. (1999) used Analytical Hierarchy
Process for priority ranking of pavement maintenance.
The priority index (PI) used in this study for pavement
maintenance priority has the following form:
PI = Wj*Fj (11)
Where, PI = Priority index for any section (out of 100);
Wj = Factor “j” weight of importance to priority ranking;
Fj = Factor “j” value (out of 100); and SWj = 1.0.
A questionnaire was prepared in four parts for
collecting the respondent‟s opinion about the weight of
importance of the seven priority factors. The seven
factors considered for priority ranking were road class,
pavement condition, operating traffic, riding quality,
safety condition, maintenance cost and importance to
community. The procedure followed in the preparation
of this questionnaire is detailed in Ramadhan (1997).
The factor weights were computed using the AHP
method by analyzing the data collected. The factor
weights so obtained were further adjusted for the
individual experience factor, in the following manner:
(12)
Where, CWi = corrected overall weight of factor i; wj =
factor weight as estimated by individual j; Ej =
individual experience factor for individual j, i = factor
number (one to seven for priority factors, and one to six
for sub-factors); and j = individual number out of the
total number of individuals n.
The final form of the developed maintenance priority
ranking procedure was given as:
PI = CWi * Fi (13)
where, PI = priority index (out of 100), CW = Corrected
weights, Fi = priority factors.
It was concluded that the developed procedure can
adequately and efficiently rank a huge number of
pavement sections for maintenance.
Bandara and Gunarante (2001) tested a methodology
developed for pavement maintenance prioritization
based on rapid visual condition evaluation using fuzzy
logic, on the major pavement network of Sri Lanka.
Four commonly observed distress types and their
severity levels chosen in this study are given below.
1. Alligator cracking [low severity (AL), medium
severity (AM), high severity (AH)]
2. Potholes [low severity (PL), medium severity (PM),
high severity (PH)]
3. Edge failures [low severity (EL), medium severity
(EM), high severity (EH)]
4. Raveling [low severity (RL), medium severity (RM),
high severity (RH)]
The concept of triangular fuzzy numbers (TFNs) with a
linear membership function as described in the Fig. 9
was used for analysis.
The mathematical equations for the membership
function for TFN that has been adopted are given as
follows:
(14)
(15)
(16)
; (17)
(18)
(Source: Bandara and Gunarante 2001)
Fig. 9 TFN – A
The efficient fuzzy weighted average (EFWA)
algorithm was employed to compute the weighted fuzzy
pavement condition index with slight adjustments to
suit the study. The employed aggregation method can be
expressed as follows:
(19)
Further a ranking index was defined for each pavement
section i using the FCI values and the rating difference
matrix, given as follows:
(20)
Where, = element indicating the degree of
difference between the respective fuzzy condition
ratings of the ith and j
th pavement sections. The ranking
for maintenance as per the current rehabilitation
urgency was done based on these ranking indexes (Ri).
Also a fuzzy pavement condition forecasting model was
developed by incorporating subjective probability
assessments regarding pavement condition deterioration
rates, in the Markov transition process. And the re-
ranking of same sections was done based on ranking
indexes (Ri) for the forecasted pavement condition in
the second year.
2nd International Conference on Emerging Trends in Engineering & Technology, April12, 13, 2013
College of Engineering, Teerthanker Mahaveer University.
10
Reddy and Veeraragavan (2001) proposed a priority
ranking methodology based on overall distress index
model and traffic adjustment factors. A schematic flow
chart of the proposed priority ranking model is as shown
in Fig. 10.
The degree of acceptability on a 0 to 1 scale was
determined, for pavement distresses of different extent
based on the expert opinion survey. The degree of
acceptability and the extent of distress were correlated
to obtain best fit function as the membership functions.
Table 4 shows the mathematical relationships developed
as a member ship functions and the relative weightage
assigned to each of distress based on expert‟s opinion.
Table 4. Mathematical relationships as membership
functions for distress
Performance
Attribute
Membership
Function R2 N S.E.
Relative
Weight-
age
Cracked
Area
AL = Exp
(0.0137 –
0.024 CRA)
0.975 21 0.121 0.30
Area of Pot
holes
AL = Exp
(0.073 –
0.077 PT)
0.982 19 0.102 0.25
Unevenness
AL = 1.157 –
1.088*10-3
UI
0.987 21 0.040 0.25
Patched
Area
AL = Exp
(0.155 –
0.0398 PA)
0.984 20 0.079 0.10
Rut Depth AL = 1.03952
– 0.0351 RD 0.996 21 0.022 0.10
(Source: Reddy and Veeraragavan 2001)
Where, AL = acceptability level (0 to 1), CRA =
cracked area of pavement (%), PT = area of pot holes
(%), UI = unevenness index (cm/km), PA = patched
area of pavement (%), RD = rut depth (mm).
The surface condition for pavement sections was
assessed in form of acceptability and relative weightage,
defined in terms of Pavement Distress Index (PDI) as
PDI = [1 - (ALi * wi) / wi] 100 (21)
Where, ALi = Acceptability level of any distress, i, wi =
weightage of distress i. Based on above equation, PDI
values have been assigned as 0 to very good pavement
and 100 to completely deteriorated pavement. Further
the prioritization factors (F) based on functional class
and average daily traffic were also considered as given
in Table 5. These factors were used to adjust the PDI
values in order to assign greater priority to the higher
functional class roadways and also to roadways with
higher traffic levels within a given functional class. The
Priority Index (PI) for a pavement section was
computed as:
PI = F x PDI = F x [1 - (ALi * wi) / wi] 100 (22)
Where, PDI = Pavement Distress Index, F =
Prioritization factor based on functional class and
average daily traffic.
Using this priority index, different maintenance
strategies were assigned to pavement sections. Also, the
developed model identified the availability of budget
and compared with the total cost required for
strengthening. It was concluded that the Priority
Ranking Model (PRM) so developed provides a
consistent, reliable and facile method of evaluating
pavements at network level.
Table 5. Prioritization factors based on functional
class and average daily
Functional Class
Average Daily Traffic
Prioritization
Factor (F) Level
No. of
Commercial
Vehicles
Expressway/
National Highway
(NH)
High >5000 1.00
Medium 3000 – 5000 0.90
Low < 3000 0.80
State Highway
(SH)
High > 3000 0.80
Medium 1500 – 3000 0.75
Low <1500 0.70
Other Roads (OR) High >1500 0.75
Medium 500 – 1500 0.70
Low < 500 0.60
(Source: Reddy and Veeraragavan 2001)
Cafiso et al. (2002) provided a Multicriteria analysis
(MCA) framework for pavement maintenance
management. An AHP method of MCA was
implemented to compute the ranking values (RV) for
the maintenance alternatives. The overall computational
procedure for implementation MCA within HDM-4 to
compute the ranking vector is as shown in Fig. 11. The
HDM-4 was used to produce outputs that can be used as
attributes for each road section and investment
alternative (i.e., the average IRI values in Figure: Part b).
The above procedure produced a matrix of
“Multicriteria ranking numbers” for every year (or the
whole analysis period), and for all road sections
included in the study. The alternative with highest RV
was selected in case of unconstrained budget. When
there is budget constraint, as is always the case in
practice, this need to be considered by applying further
analysis of prioritization or optimization.
Narasimha (2003) defined a procedure prioritizing roads
in a network for periodic maintenance based on the
subjective rating technique as per the guidelines of
Asphalt Institute (AI), USA. The field surveys on
various categories of roads (SH, MDR & ODR) were
conducted and data regarding the present serviceability
conditions were collected for about 22 km of road
length from a total of 160.334 km under study area, in
2nd International Conference on Emerging Trends in Engineering & Technology, April12, 13, 2013
College of Engineering, Teerthanker Mahaveer University.
11
Tamil Nadu. The pavement condition survey was
carried out, by identifying and evaluating thirteen types
of distresses as defined by AI and then the pavement
condition rating (PCR) of each stretch of road was
computed. The distresses considered with their rating
score were: Transverse crack (Rating : 0-5),
Longitudinal crack (Rating : 0-5), Alligator crack
(Rating : 0-10), Shrinkage crack (Rating : 0-5), Rutting
(Rating: 0-5), Corrugations (Rating :0-5), Ravelling
(Rating: 0-5), Shoving (Rating: 0-5), Potholes and
Patching (Rating: 0-10), Excess bitumen (Rating: 0-10),
Polished aggregate (rating: 0-10), Deficient drainage
(Rating: 0-10), Riding quality (Rating : 0-10). For all
rating scores lowest value indicated good pavement
condition and highest value indicate worst pavement
condition.
(Source: Reddy and Veeraragavan 2001)
Fig.10 Schematic flow of priority ranking model
Finally, the PCI value was calculated using the charts
given in AI and the prioritization of maintenance for
road sections was done.
Kumar (2004) developed a methodology for priority
ranking of highway pavements for maintenance of
certain roads in Thiruvananthapuram city based on
composite criteria. The approach included the opinions
from the users in first step and opinions from the
experts in the second step. The questionnaire for user
opinion survey consisted two parts one with trip
information and second with pavement condition.
Average response of each user was found out using
Excel Macro Programming. Analysis of Variance
technique was used to decide that minimum sample size
should be 30 so that there is no variation between users.
From the user opinion different pavement section
was ranked for priority maintenance.
In next step functional evaluation of pavement sections
was done and following parameters were analyzed:
(i) Visual rating was done and sections were rated on
a scale of 0 – 5. The riding quality assessment was done
by driving the vehicle at uniform speed of 30kmph and
a subjective rating scale of 0-5 was adopted.
(ii) Crack measurement was done and crack index was
computed based on width of crack.
(23)
where, C1 = area of cracks with width 1mm or less, C2
= area of cracks with width 1 to 3 mm, C3 = area of
crack with width greater than 3 mm, w = pavement
width in m.
The Crack Index (CI) for each km was computed
by assigning weightages to the different severity levels.
Crack Index = (Crack Percentage x WF) x 100 (24)
Where, WF = weightage factor = 0.5 for low severity
level = 0.75 for medium severity level = 1.0 for high
severity level.
(iii) The number of small, medium and large potholes
per km of roads was recorded and the percentage of
potholes was compute as:
(25)
where, S = number of small potholes, M = number of
medium potholes, L = number of large potholes, W =
pavement width in m.
Then percentage of potholes were converted to Pothole
Index (PI) is calculated by assigning weightages to the
different severity levels. The PI is calculated as:
Start
Total No. Of Pavements
Pavement Condition Information
Cracked Area, Raveled Area
Patched Area, Pothole Area
Put Depth, Unevenness Index
Determination of acceptability
levels
Determination of PDI
Membership
Functions
Weightage of
Distresses
Budget
Priority Index
PI = PDI F
Functioned class
Traffic Volume
Prioritization Factor „F‟
PI > 2.5
10 < Pi < 25
Assign Strategy – I
PI Priority
>50 1
45 – 50 2
20 – 45 3
35 – 40 4
30 – 35 5
25 – 30 6
Assign Strategy –
II
PI Priority
20 – 25 1
15 – 20 2
10 – 15 3
Assign Strategy –
III
PI Priority
5 – 10 1
< 5 2
Total Cost of Improvements
Cost Budget
Allocated Budget to All
Pavements for
Improvements
Allocate Budget According to
Priority Index & Priority List
Outputs of Priority model & listing of overlay strategies as per
Budgets
2nd International Conference on Emerging Trends in Engineering & Technology, April12, 13, 2013
College of Engineering, Teerthanker Mahaveer University.
12
Pothole Index = (% of pothole x WF) x 100 (26)
Where, WF = weightage factor = 0.5 for low severity
level = 0.75 for medium severity level = 1.0 for high
severity level.
(iv) Roughness value in terms of Unevenness Index
(UI) was calculated indirectly by using the mean of ride
rating values termed as Present Serviceability Rating
(PSR). The following model developed at Bangalore
University was adopted:
PSR = -1.9326 loge (UI) + 14.3765 (27)
Based on above functional evaluation indices second
questionnaire was prepared and expert opinions were
taken for fixing priority for maintenance. Finally the
priority obtained from user and experts were compared.
(Source: Cafiso et al. 2002)
Fig. 11 AHP procedure for computing ranking vector
Pantha et al. (2010) developed a Geographic
Information Systems (GIS)-based highway maintenance
prioritization model for major link of 53.2 km to the
capital city of Kathmandu, Nepal. The highway sections
were prioritized considering the pavement and roadside
slope stability condition. Using the bivariate statistical
method (Information Value Method) also know as
statistical index method, the weight value for a
parameter class affecting the landslide and pavement
condition were calculated. Finally an integrated
maintenance priority index (MPI) was computed by
adding pavement maintenance prioritization index and
roadside slope maintenance prioritization index after
multiplying by weighted value of each component, as
given in following expression.
(28)
where, PMPI is the pavement maintenance prioritization
index, RMPI is the roadside slope maintenance
prioritization index, Wp is the weight given to pavement
maintenance and Wr is the weight given to roadside
slope maintenance. The selection of maintenance
sections were done based on these calculated MPI.
Khademi et al. (2010) presented a combined
Conference-Delphi-analytic hierarchy process (AHP)
model employed to prioritize the low-class roads in
Gilan, Iran. The main three objectives defined for this
study was determining the high-priority low-class roads
for maintenance (MA), improvement (IM), and
upgrading (UP).
2nd International Conference on Emerging Trends in Engineering & Technology, April12, 13, 2013
College of Engineering, Teerthanker Mahaveer University.
13
Each part of the designed Conference-Delphi-AHP
decision-making model function interdependently
through a series of systematic steps was as shown in the
Fig. 12.
(Source: Khademi et al. 2010)
Fig. 12 Frame work of the study
The first step included the creation of data bank for road
network in Arc View, GIS software. Then, three
categories for the prioritization were formed by the
classifying algorithm as shown in Figure 33 for three
objectives viz. MA, IM & UP. And all the sections of
low-class road network are tested by the algorithm and
placed in these three categories. Next step included to
finalize the decision criteria and list of decision makers
to participate in AHP. This was done by conducting
conference and Delphi Survey. The AHP hierarchies
decided to make the decisions for MA, IM & UP were
as shown in Fig. 14.
Lastly, the final weight of alternative number n was
calculated by the following equation with regard to all
of the criteria in the hierarchy:
(29)
where io=number of the criteria in Level 1; ji=number of
the criteria in Level 2 whose parent criterion is the ith
criterion in Level 1; and kji=number of criteria in Level
4 whose parent criteria are the jth criterion in Level 3
and the ith criterion in Level 2.
(A-1)
Network Survey
(A-2)
GIS Data Bank
(A-3)
Classifying algorithm
(Alternative Specification)
Decision Alternative Sets (that the
alternative in each set should be
Prioritized) with respect to its
related objective (MA, IP ,or UP))
(C)
Computer Program for:
1-Aggregation of the Views
2-Conversiion of the Pair –wise
Comparisons to the Vector of Wights
3-Calculation of the inconsistencies
GIS Based Routes & Sub-Routs
Data bank
(B-1)
Apply the Conference Group Decision Making Techniques to:
i) Fix the Problem Goal and Objectives.
ii) Construct the Pre- Final hierarchies (the Decision Criteria
Specification)
iii) Select the Group of experts who should be participated in
the AHP session
iv) receive the approval of the algorithm which specify the
alternative s (Step a -3)
(B-2)
Finalizing the hierarchies through the Delphi Survey
(D)AHP
(D-1)
Determine the Criteria Weights
(D-2)
For each decision alternatives set, perform the alternative
Prioritization with respect to its objective (MA, or IP, or UP)
Determine the routes & sub-routes weights
with respect to the objectives, MA,IM and
UP (route & sub-route) prioritization) Based
on the weights derived from step (C-2)
(F)
Negotiation
2nd International Conference on Emerging Trends in Engineering & Technology, April12, 13, 2013
College of Engineering, Teerthanker Mahaveer University.
14
(Source: Khademi et al. 2010)
Fig. 14 Three AHP hierarchies for three objectives
Kaysi et al. (2010) prioritized the National Road
sections based on a multi-criteria analysis (MCA)
approach, the linear additive model, which evaluates
each scheme on a set of criteria derived from the
various national goals and objectives of the Saudi
authorities.
The framework designed and implemented to set the
priorities was explained by flow-chart as show in Fig.
15. To set the priority an exhaustive set of criteria‟s
been decided. The selected criteria were grouped into 5
main criteria (Group Criteria) reflecting the broad
objectives of the key players: A- Road Network
Development; B- Socio-Economic Development; C-
Economic Efficiency; D- Serving Hajj and Umrah; and
D- Other Criteria including Security, Safety, and
Environment. Each main criterion was then divided into
primary criteria, some of which were divided further
into secondary criteria, thus resulting in a three level
hierarchy as given in Fig. 16.
In the next step the relative weights of each criterion
were derived using AHP method by developing the
questionnaire. The following Linear Additive model
was used to calculate the overall value score of a
proposed road scheme by adding its weighted score on a
set of criteria:
(30)
where Si is the total value score of scheme i; wj is the
relative weight of criterion j; Sij is the value score of
scheme i on criterion j; and n is the total number of
criteria. The schemes were then divided into two groups
based on the road category: primary and secondary
roads. The schemes were then ranked based on the
calculated total scores. The sensitivity analysis was also
carried out to evaluate the robustness of the Saudi
Arabian Road Prioritization (SARP) model.
2nd International Conference on Emerging Trends in Engineering & Technology, April12, 13, 2013
College of Engineering, Teerthanker Mahaveer University.
15
(Source: Kaysi et al. 2010)
Fig.15 The prioritization process
3.3 Based on Economic Analysis
Livneh and Craus (1990) suggested an economic-based
model for prioritizing the maintenance of pavement
sections. The model had a following mathematical form:
(31)
Where P% = the first year rate of return, SN = structural
number = 5 - 0.04DR, where DR is a distress grade (0-
100), PSI = serviceability grade (0-5), AADT = annual
average daily traffic, and k = numeric constant.
The first year rate of return was estimated for a
given rehabilitation investment, that is the agency and
road user‟s costs (before and after upgrading), and the
benefits gained from investment in the maintenance and
rehabilitation work. The governing factor of this model
was AADT.
Data
Collection Objective of
Key Player
GIS Database
Existing Roads
Railway Lines
Airports/Seaports
Regions
Cities
Industrial Area
Agriculture Areas
Topography
Proposed Roads
Mot
Proposed
Road
Schemes
Study
Team
Cost Estimation
Model
Road network Traffic
Model
Study
Team
Budget for Primary
Roads Schemes per
Phase
Criteria
Measurement
s
Set of
Criteria
Scoring
Guidelines
Scores Per
Schemes and
Criteria Co
nsu
ltation
Application
of AHP
Relative Weights of
Criteria
Linear
Additive Model
Total Score Per
Scheme
Ranking of
Primary Road
Schemes at a
national Level
Ranking of
Secondary Road
Schemes by
region
Choice of Primary&
Secondary Data Schemes
to be implemented during
phase
Any More
Phase?
Regional
Budget
for Phase
End of
Prioritization
Process
Repeat process for next
phase
Yes No
2nd International Conference on Emerging Trends in Engineering & Technology, April12, 13, 2013
College of Engineering, Teerthanker Mahaveer University.
16
(Source: Kaysi et al. 2010)
Fig. 16 The hierarchy of the set of criteria
Sharaf & Mandeel (1998) analyzed three techniques for
priority setting. The first technique was a simple ranking
based on four ranking measures viz.: (i) lowest life cycle
cost; (ii) worst condition first; (iii) highest traffic and (iv)
highest benefit/cost ratio. The second technique was a
combined ranking technique based on relative weights
assigned to the above mentioned four ranking measures.
Finally, the third technique was a linear programming
optimization model which considers both time (current
and future) and space (entire network). A comparison
between the three techniques in terms of network
condition over time and in terms of budget deficit over
time was presented. The results indicated a considerable
difference in future network performance under the three
techniques with the optimization technique produced the
best results.
Veeraragavan (2000) presented a methodology of
prioritizing the highway sections for maintenance based
on the current structural and functional condition of the
pavement and the traffic intensity. An economic analysis
was done to evaluate the benefit/cost ratio and net
present value. Based on the highest value of the
economic criteria, the pavement sections were selected
for treatment and the highway sections were ranked on
priority for maintenance.
Roy et. al. (2003) evaluated the uses of HDM-4 as a
versatile tool to study the economic viability of
alternative road projects and to prepare road investment
programmes for the selected sections of State highways
of Kerala. The M&R strategies were selected for
sections based on IRR and B/C values.
Aggarwal et. al. (2004) prioritized the yearly M&R
works for the National Highways based on the
decreasing NPV/Cost ratio. The „Project Analysis‟
component of HDM 4 was used for doing the analysis
PRIORITIZATION OF PRIMARY AND SECONDARY ROADS
Transport
Network
(A) 0.26/0.26
Socio Economic
Development
(A) 0.26/0.26
Economic
Feasibility
(C) 0.09/0.09
Hajj &
Umrah
(D) 0.12/0.12
Other
(E)
0.06/0.06
Traffic Demand
Supply
(A) 0.77/0.200
Connectivity
(A)
0.77/0.200
Connectivity (A)
0.77/0.200
Connectivity
(A) 0.77/0.200
Connectivity (A)
0.77/0.200
Connectivity
(A)
0.77/0.200
Time service
(C1) 0.09/0.09
Traffic Volume during Hajj
(D1) 0.26/0.031
National;
Security (E1)
0.16/0.011
Saving in VOC
(C2) 0.27/0.024
LOS Deterioration during Hajj
(No Subject)
Safety
(E2)
0.62/0.037
Traffic Volume during Umrah
(D3) 0.17/0.020
Environment
(E3)
0.12/0.007
Relative Cost
(C3) 0.30/0.027
LOS Deterioration
During (no project)
(D) 0.12/0.12
Political
Commitment
(E4)
0.08/0.005
Traffic demand
(A1-1) 0.30/0.060
Level of service Deterioration
(no Project)
(A1-2) 0.32/0.064
Extra Capacity
(A1-3) 0.18/0.036
Service Freight
Movement
(A1-4) 0.20/0.040
Multi Model Connectivity
(A2-1) 0.22/0.013
Connecting within National road
Network (A2-2)
0.62/0.037
International Connectivity
(A2-3) 0.18/0.036
Manufacturing
(B2-1) 0.41/0.057
Agriculture
(B2-2) 0.32/0.064
Tourism
(B2-3) 0.18/0.036
Miring
(A2-4) 0.20/0.040
Other Services
(B2-4) 0.15/0.021
Manufacturing
(B3-1) 0.29/0.027
Agriculture
(B3-2) 0.10/0.009
Tourism
(B3-3) 0.13/0.012
Miring
(B3-4) 0.22/0.032
Economic Cities
(B3-4) 0.26/0.024
Weight within Segment /
Relative Weight
Crit
eri
a G
rou
ps
Prim
ary C
rit
eri
a
Seco
nd
ary
Crit
eria
2nd International Conference on Emerging Trends in Engineering & Technology, April12, 13, 2013
College of Engineering, Teerthanker Mahaveer University.
17
and computing the NPV/cost ratio for different
pavement sections of NH under study. A section with
higher NPV/Cost ratio was considered first for
maintenance.
Singh and Sreenivasulu (2005) used programme analysis
of HDM-4 for prioritizing the maintenance of road
sections, under different budget scenarios. The
NPV/Cost ratios were determined for all candidate
sections and were prioritized accordingly.
Parida et al. (2011) & Shah et al. (2012) compared the
two methods for priority ranking of urban road
maintenance, viz. (a) ranking based on subjective rating
and (b) ranking based on economic indicator. The
subjective ranking was done using maintenance priority
index (MPI) which was a function of road condition
index, traffic volume factor, special factor and drainage
factor. The second ranking method was based on
economic indicator in which NPV/Cost ratio was
calculated for each pavement section using the HDM-4
software.
4. Discussion and Conclusions
The main aim of this paper was to build knowledge
about the pavement maintenance prioritization methods
and to present the research studies developed in this area
by different researchers at global level.
It can be concluded that the selection of prioritization
method is a case specific and not applicable everywhere.
The selection of pavement sections on the bases of fund
constraints, may lead to road condition deterioration for
other sections not in priority list. At the same time,
prioritization based on optimization is able to maintain
or even improve road conditions of the overall network.
Also it has been observed that:
- The complexity is added to the calculations as we
move on from simple engineering judgment based to
complex network optimization models, as a
prioritization method.
- Ranking based methods requires less data collection,
than other methods, where data like future traffic
projection, future funding, forecast of effects and
durability of maintenance and pavement deterioration
models need to be studied.
References
[1] Agarwal, P.K., Das, A. and Chakroborty, P., “A Rational
Approach of Prioritization of Highway sections for
Maintenance”, Proceedings 6th
International Conference
on Managing Pavements, Brisbane, Australia, 2004.
[2] Aggarwal, S., Jain, S.S. and Parida, M., “Development
of Pavement Management System for Indian National
Highway Network”, Journal of Indian Roads Congress,
Vol. 65-2, pp. 271-326, 2004.
[3] Andres, C.W. and Collura, J., “Network-Level
Prioritization of Local Pavement Improvements in Small
and Medium-Sized Communities”, Transportation
Research Record 1455, Transportation Research Board,
pp. 13-21, 1994.
[4] Bandara, N. and Gunaratne, M., “Current and Future
Pavement Maintenance Prioritization Based On Rapid
Visual Condition Evaluation”, Journal of Transportation
Engineering, ASCE, vol. 127 (2), pp. 116-123, 2001.
[5] Butt, A. A., Shahin, M. Y., Carpenter, S. H. and
Carnahan, J.V., “Application of Markov Process to
Pavement Management Systems at Network Level”,
Proceedings of Third International Conference on
Managing Pavements, Transportation Research Board,
Texas, Vol.2 pp.159-172, 1994.
[6] Cafiso, S., Graziano, A.D., Kerali, H.R. and Odoki, J.B.,
“Multicriteria Analysis Method for Pavement
Maintenance Management”, Transportation Research
Record No.1816, Transportation Research Board, pp.
73-84, 2002.
[7] Chandran, S., Isaac, K.P. and Veeraragavan, A.,
“Prioritization of Low-Volume Pavement Sections for
Maintenance By Using Fuzzy Logic”, Transportation
Research Record No.1989, Transportation Research
Board, pp. 137-144, 2007.
[8] Farhan, J. and Fwa, T. F., “Pavement Maintenance
Prioritization Using Analytic Hierarchy Process”,
Transportation Research Record No.2093,
Transportation Research Board, pp. 12-24, 2009.
[9] Flintsch, G.W., Zaniewski, J.P., and Medina, A.,
“Development of a Knowledge-Based Formula to
Prioritize Pavement Rehabilitation Projects”,
Transportation Research Record, No. 1643, pp. 54-61,
1998.
[10] Fwa, T. F. and Chan, W. T., “Priority Rating Highway
Maintenance needs by Neural Networks, Journal of
Transportation Engineering, ASCE, vol. 119 (3), pp.
419-432, 1993.
[11] Haas, R.C.G., Hudson, W.R. and Zaniewski, J.,
“Modern Pavement Management”, Krieger Publications,
Malabar, Florida, 1994.
[12] Haas, R.C.G., Hudson, W.R. and Zaniewski, J.,
“Modern Pavement Management”, Krieger Publications,
Malabar, Florida, 1994.
[13] Jain, S.S. and Gupta, A.K., “Development of
Maintenance & Rehabilitation strategy for Flexible
pavements”, Journal of Indian Roads Congress, vol.57
(2), pp. 367-418, 1996.
[14] Karan, M.A., Haas, R. and Walker, T., “Illustration of
Pavement Management: From Data Inventory to Priority
Analysis”, Transportation Research Record No.814,
Transportation Research Board, pp. 22-28, 1981.
[15] Kaysi, I., Al-Hathlool, H., Kahtani, S., Whiteway, M.
and Darwish, F., “Prioritization of National Road
Projects in Saudi Arabia: Framework and
2nd International Conference on Emerging Trends in Engineering & Technology, April12, 13, 2013
College of Engineering, Teerthanker Mahaveer University.
18
Implementation”, 12th
World Conference on Transport
Research, Lisbon, Portugal, 2010.
[16] Khademi, N. and Sheikholeslami, A., “Multicriteria
Group Decision-Making Technique for a Low-Class
Road Maintenance Program”, Journal of Infrastructure
Systems, ASCE, Vol. 16, No. 3, pp. 188-198, 2010.
[17] Kumar, S.M. and Kumar, V.N., “Development of a
Methodology for Priority Ranking of Highway
Pavements for Maintenance based on Composite
Criteria”, Highway Research Bulletin, IRC, No. 70, pp.
31-54, 2004.
[18] Livneh, M. and Craus, J., “Economic Models For Setting
Priorities For Road Maintenance And Rehabilitation In
Israel”, Transportation Research Record No.1268,
Transportation Research Board, pp. 205-208, 1990.
[19] Moazami, D. and Muniandy, R., “Fuzzy Inference and
Multi-criteria Decision Making Applications in
Pavement Rehabilitation Prioritization”, Australian
Journal of Basic and Applied Sciences, vol. 4(10), pp.
4740-4748, 2010.
[20] Narasimha, V.L., Sundararajan, T. and Raja, M.,
“Subjective Rating Technique- A Tool for Prioritizing
Roads in a Network for Periodic Maintenance”,
Highway Research Bulletin, IRC, No. 69, pp.1-14, 2003.
[21] Pantha, B.R., Yatabe, R. and Bhandary, N.P., “GIS-
based Highway Maintenance Prioritization Model: An
Integrated Approach for Highway Maintenance in Nepal
Mountains”, Journal of Transport Geography, Elsevier,
Volume 18 (3), pp. 426-433, 2010.
[22] Parida, M., Shah, Y.U., Jain, S.S., Tiwari, D.,
“Prioritization of Maintenance Management Decisions
for Urban Roads Using HDM-4”, Proceedings of
Seventh International Conference on Road and Airfield
Pavement Technology (ICPT), Bangkok, Thailand, 3-5
August 2011, pp. 1246-1255.
[23] Ramadhan, R.H., Modeling of Pavement Condition and
Maintenance Priority Ranking for Road Networks, PhD
Thesis, Civil Engineering Department, King Fahd
University of Petroleum and Minerals, Dhahran, 1997.
[24] Ramadhan, R.H., Wahhab, H.I. and Duffuaa, S.O., “The
Use of An Analytical Hierarchy Process in Pavement
Maintenance Priority Ranking”, Journal of Quality in
Maintenance Engineering, vol. 5 (1), pp. 25-39, 1999.
[25] Reddy, B.B., and Veeraragavan, A., “Priority Ranking
Model for Managing Flexible Pavements at Network
Level”, Journal of Indian Roads Congress, vol.62-3, pp.
379-394, 2001.
[26] Roy, N., Issac, K. P. and Veeraragavan, A., “Highway
development and Management tool (HDM-4):
Calibration to Indian conditions and its Application-
Case Study”, Highway Research Bulletin, IRC, No. 69,
pp.73-96, 2003.
[27] Sandra, A.K., Rao,V. Raju, K.S. and Sarkar, A.K.,
“Prioritization of Pavement Stretches Using Fuzzy
MCDM Approach – A Case Study”, Soft Computing in
Industrial Applications, Advances in Soft Computing,
vol.39, pp. 265–278, 2007.
[28] Shah, Y.U., Jain, S.S. and Parida, M., “Evaluation of
Prioritization Methods for Effective Pavement
Maintenance of Urban Roads”, International Journal of
Pavement Engineering, (Online on 9 February 2012).
[29] Sharaf E. A. and Mandeel F. M., “An Analysis of the
Impact of Different Priority Setting Techniques on
Network Pavement Condition, 4th
International
Conference on Managing Pavements (1998).
[30] Sharaf, E., “Ranking Versus Simple Optimization in
Setting Pavement Maintenance Priorities: A Case Study
from Egypt”, Transportation Research Record 1397,
Transportation Research Board, Washington, D.C.,
pp.34-38, 1993.
[31] Suma, N.H., Chakraborty, N., and Veeraragavan, A.,
“Simplified Methodology for Prioritization of Pavement
Maintenance Based on Pavement Condition Index”,
Proceedings, 1st European Pavement Management
Systems Conference, Budapest, Hungary, 2000.
[32] Veeraragavan A., and Chakraborty, N., “Prioritisation of
Pavement Maintenance based on Economic Criteria”,
Proceedings, 1st European Pavement Management
Systems Conference, Budapest, Hungary, 2000.
[33] Zhang, Z., “An AHP Based Approach for Prioritization
of Pavement Management Data Requirements”,
Proceedings 6th
International Conference on Managing
Pavements, Brisbane, Australia, 2004.
[34] Zhang, Z., Singh, N. and Hudson, W. R.,
“Comprehensive Ranking Index for Flexible Pavement
Using Fuzzy Sets Model”, Transportation Research
Record 1397, Transportation Research Board,
Washington, D.C., 1993, pp.96-102.