You are on page 1of 33

Value World

Volume 32 Number 2 Summer 2009


Published by SAVE International

Contents:
2

Editors Comments

Measuring Perceived Capability


and Motivation vs. Experience in
Value Engineering Workshops Using
Expectancy Theory

M. A. Berawi, Ph.D.

A. K. Medallah, AVS

11

Best-in-Class: An Executive
Level Application of the Value
Methodology

Ronald J. Tanenbaum, Ph.D., PE, CVS, GE,


F.ASCE & Gordon Johnson, PE

19

Can Value Methodology Enhance the


Competitiveness of the Supply Chain
Ferenc N
Ndasdi,
dasdi, Ph.D., CVs, FSAVE

26

Genetic Algorithm in Designing


Optimal Layout for Stock Plate
Storage with a Unique Dimension of
Material
T. Yuri Zagloel & Heryadi Pamungkas

Inside this issue:

Process
and Product
Optimization
Using Value
Engineering/Value
Management

Process and Product Optimization Using


Value Engineering/Value Management
EDITORIAL
After a series of innovation discussions, the theme of
engineering optimization is chosen for this summer issue
as another beneciary of value engineering/value management application. Optimization is the action of designing and developing systems to gain maximum advantage of the available resources or, in other words, is the
procedure(s) used to make a system or service as eective
or functional as possible. In the art of systems design and
operation, it is of crucial importance to reduce costs and
to improve performance, quality, system reliability and
on-time delivery of a project/product.
During the designing of a product for manufacturing or early phases of construction projects, some key issues and constraints arise in a large number of conicting
pressures regarding the engineering/construction design,
operation, cost, safety, environmental issues, etc. Identication of design optimization in a system turns into a
necessary attribute to create value in the form of productivity, performance, and competitiveness. In this sense,
value engineering/value management is applied as a tool
and as a process to optimize the value of design, engineering, services and conguration aspects. The initial
functional designs and value engineering/value management workshops utilize a guided process to lead the team
through the workshops with the objective to maximize
the expected output.
Process and Product Optimization
This issue of Value World presents four selected papers
from SAVE Internationals 2009 Annual Conference, the
11th International Conference on Quality in Research,
and the Value World journal submission, to stimulate a
discussion and to explore various optimization systems
employed in the context of construction and manufacturing industries.
The rst two papers examine the methods and activities for providing the value engineering/value management workshops a maximum output. The rst paper,
2

Volume 32, Number 2, Summer 2009

written by A. K. Medallah, emphasizes the perception of


the capabilities of members during VE workshops and
the impact of motivational outcomes by evaluating three
main factors using Expectancy Theory, namely: (1) member expected capability (Eort); (2) perceived outcomes
expected (Outcome), (3) desirability or undesirability
weight of each outcome (Valance). An empirical study
was conducted on 21 VE practitioners in a Saudi Arabian
governmental agency. As the result, this paper concluded
that by optimizing and updating tools and skills required
by facilitators in VE workshops can lead to a better outcome of VE study.
Furthermore the second paper, written by Ronald J.
Tanenbaum, Ph.D., CVS, PE, GE, F.ASCE and Gordon
Johnson, PE, describes how U.S. public agencies can improve their operating procedures in the quest to become
the Best-in-Class in their area of expertise by addressing three critical points: modifying the delivery methods,
providing more accurate cost estimation, and constructing a change for order management. As a conclusion, the
paper provides recommendations for performing successful Best-in-Class workshops for any agency, i.e. to shares
a goal to perform self-inspection on the operations, to
accept changes, and to possess a sincere desire to be better, if not the best.
The third paper, written by Ferenc Ndasdi, Ph.D.,
CVS, FSAVE, discusses how the value methodology can
be used eectively in the analysis of supply chain processes. Based on Hungarian and international experience,
he argues that it is important to ensure ecient cooperation among vertically related units regarding delivery
contracts or even regarding the R&D activities and the
planning of investments, where value analysis/value engineering can facilitate signicantly a more ecient operation of the entire vertical structure.
The last paper in this edition, written by T. Yuri Zagloel and Heryadi Pamungkas, provides a genetic algoV A L U EW O R L D

rithm mechanism to be used in creating an optimal layout


design, with regard to material handling in a manufacturing storage area, that aects the delivery time and cost
eciency. This model optimizes the stock plate storage
area having various types of plates, and every single plate
has a unique dimension and lifting intensity. In order to
optimize the process, crossovers and mutations are applied to dene a better solution with better tness values
as shown in a case study at a heavy equipment manufacturer in Jakarta. The correlation between optimization
and value engineering is shown in the values obtained
from the genetic algorithm, which has been applied in
the real storage condition.
I hope this edition of Value World conveys some new
insights in the way we conduct our value methodology
studies. Your valuable contribution and feedback are important to the success of our journal as it will guide the
journals future development.
With warmest regards from editorial desk,

Board the Queen


Mary for a
Value Adventure
on the High Seas!
Attend SAVE Internationals
50th Annual Conference
June 6 - 10, 2010
on the Queen Mary in
Long Beach, California, USA
Network with leading value practitioners from around the world.
Learn cutting edge innovations and
techniques.

Dr. M.A. Berawi


Faculty of Engineering
University of Indonesia
Jakarta 16424
Indonesia

Build skills at pre-conference workshops.


Earn credit toward certication.
Visit with partner organizations.
Interact with expert panelists and
speakers.

Check the SAVE International Web site for conference updates: www.value-eng.org.

V A L U EW O R L D

Volume 32, Number 2, Summer 2009

Measuring Perceived Capability & Motivation vs. Experience


in Value Engineering Workshops Using Expectancy Theory
A. K. Medallah
Abstract
The relationship between behavioral science and engineering practices is becoming substantially strong. Value engineering (VE) practitioners; deal with humans on daily basis inside
VE workshops, during information gathering or implementation phases. These people vary in emotions, experience and
level of motivation. This paper is intended to survey the perception of members capabilities during VE workshops and
the impact of motivational outcomes expected. This was studied by measuring three main factors (1) member expected capability (Eort); (2) perceived outcomes expected (Outcome),
(3) desirability or undesirability weight of each outcome (Valance). The methodology of the paper was done by using the
well-known Expectancy Theory as a basis of measurement focusing on non-monetary factors that inuences members signicantly. An empirical study was conducted on 21 VE practitioners in a Saudi Arabian governmental agency.

man behaviors in a sense to improve their productivity, like


Maslow Hierarchy of Needs, Herzberg Hygiene, Alderfer
Existence Relatedness and Growth ERG, Equity Theory, and
Vroom Expectancy Theory (Jones & George, 2008). As rst
developed by (Vroom, 1964), Expectancy Theory is a process
theory of motivation. It has held a major position in the study
of work motivation (cf. van Eerde and Thierry, 1996) and had
served as a rich source for theoretical innovations in various
domains, such as organizational behavior and compensation
(Lawler, 1971). The advent of Expectancy Theory is how to
motivate people using intrinsic measurements, such as, employee recognition, supervisor praising, training and improvements in the work eld. These factors are all considered nonmonetary because management doesnt compensate employees
directly. Non-monetary factors are considered a Need in most
of those theories mentioned above, named as Self-actualization, Esteem, Belongingness (Maslows), or Growth, Relatedness
in (Alderfers).

Keywords

Problem Statement

motivation, capability, value engineering, Expectancy


Theory

Introduction
VE is a problem solving methodology which magnies
problems by determining basic functions of project/product
and emphasizing symptoms with potential qualitative enhancements. VE is a systematic team analysis approach focusing on life cycle costs improving functions, qualities and beyond doubt costs (Al-Youse, 1998). Recently, in large rms
in Saudi Arabia the application of VE studies becomes a mandatory procedure for projects exceeding a certain cost.
The essence of VE workshops is to have a participative
team eort which enhances creative solutions to problems
with all concerned parties involved. VE workshops are peopleincentive and have to be controlled with full regard to peoples
backgrounds, cultures and intents. To motivate people in VE
workshops the management has to understand their behavior. In fact, management has to set their own procedures in
how to motivate and encourage people to be creative in VE
workshops. Many contemporary theories are focusing on hu4

Volume 32, Number 2, Summer 2009

The intention of this paper is to identify intrinsic measurements in how to motivate people and make them more committed to VE workshops by non-monetary outcomes. This was
done by searching for cognitive problems rather than explicit
ones aecting the performance of members coming from positive and negative expected outcomes. The survey tested 21 VE
practitioners to measure how they perceive their own capability and the expected outcomes, measuring that against their
level of experience. The results are to enlighten managements
to understand a better approach for managing VE workshops
with concentration on people not the process.

Value Engineering Workshops


The most important phase in any VE study is the workshop, which is a team eort meeting gathered to do functional
analyses and generate ideas. The essence of a VE study is letting members stride with their thinking abilities to create innovative ideas and generate applicable alternatives, this done
through functional analyses or generating ideas. This can be
facilitated by allowing a comfortable environment and enhance productivity by the outcomes of the work shop. The
productivity of the people involved in a workshop is aected

V A L U EW O R L D

by explicit factors a facilitator can monitor at a glance and


cognitive problems which are implicit and are very dicult to
control. Examples of explicit problems are external and internal physical environments of a workshop, tools and skills used
and pays and bonuses for team members. Cognitive problems
are coming from the daily job the employees, reactions from
a manager or work team and family problem all of these are
examples of factors cognitively or intrinsically aecting members. Another intrinsically motivating factor is the workshop
itself, the extent of dialectical negotiations to throw ideas and
not criticizing them once generated, or knowing that the implemented results are only biased ideas or management-inuenced ones.

Participating Members

The project understudied determines who is usually involved in the workshop. A facilitator is in charge of leading
the workshop, members perform the analyses and idea generations, and stakeholders may assign a representative to validate
results. It is not always recommended that only the bothered
[sic] project teams are the ones working in the workshops
although their contribution is essential. Involving people
outside the project team may allow a room for thinking outside-the-box which will enlighten the project team with more
alternatives and possible gaps to study. A facilitator should allow a room for controlled discussions and avoid members that
will modify the description of the problem of the project to
comfortably t his discipline (Kaufman, 1998). This involvement of multi-disciplinary teams with dierent backgrounds
from inside or outside the organization can inuence the implementation of ideas and the way the ideas being criticized
and generated.

Management Role

The management team must decrease their level of control


over the workshops; they should not be involved to resolve
problems of cost reduction impacts, although, their participation in evaluation and presentation phases could be preferable.
A contribution during the creativity phase may be useful, but
the main responsibility of management is to assure the process
is going to be in the best interest of the organization and members have a good environment to generate their work with setting procedures for produce more ideas and alternatives. These
procedures should be associated with outcomes that are considered valuable to members recognizing non-monetary outcomes and controlling what is outside the facilitator scope of
coverage.
Equally important is selecting the level of the team members that have a common factor among their cultures, beliefs
and management level. The number of participants attempted to have every discipline represented on the team could be

V A L U EW O R L D

more harmful than skipping others (Woodhead, 1998). The


selection of participants should be in regard that all members
should have the authority over their decisions without being
aected by other members in or outside the workshop to have
better unbiased ideas.

Facilitator Role
A facilitator administers a well planned strategy in dierent problem situations before and during the workshop (Norton & McElligot, 1995). Certainly, the skill of facilitator with
the methodology of the job plan provide a solid foundation
for success (Miles, 1972). The selection of a facilitator for a
VE study should be inuenced by the candidates ability to
empathize with the social, emotional, intellectual, and technical content of the workshop (Woodhead, 1998). Woodhead
presented a model that can be used to design a job specication for the selection of the appropriate facilitator to suit a
particular workshop. Some of them where:

The facilitator must have an adequate capability to maintain team creativity-borne behavior,

He should be aware of the sources of information,

He must create a culture of trust among all members,

He must review the procedures of the organization to


know what can aect members,

He must chat with team members and measure the extent


of vulnerability on them when leaving their actual work
duties.

Create a synergic environment and allow for quality contributions from members,

He is also responsible in the early beginning of the workshop to explain to members how they will not be aected
by others especially their actual work supervisors.

Motivation Methods

In general, productivity signies the measurement of how


well an individual uses its resources to produce outputs from
inputs. Motivation is the key skill to enhance this productivity, by understanding what encourages a member to be engaged in any kind of behavior. Therefore, motivation has been
dened as a physical or psychological drive to obtain the needs
to satisfy ones needs (Maloney, 1986). A vast amount of studies have been made in motivating people, whether in Needs
Theor
eories
ies which implies that a person has a high need for seeking achievement, aliation and power (McClelland), or Content Theor
eories
ies which implies that a human needs change with
time or with his experience level. However, Content Theories
are eective for people who want to focus their carriers in a
Volume 32, Number 2, Summer 2009

specialized business-line; i.e., value engineering. Expectancy


Theory has the tendency to allow for changes with time, it can
be adapted to any outcome perceived by people as long as the
outcome is expected and the valance was measured properly.
Alderfer (1972) classied the persons needs which require
intrinsic motivation explicitly by three factors his Existence,
Relatedness and personal Growth. In a creativity phase, managers who employ managerial practices that crush their employees intrinsic motivation or their internal motivated desires
will conclude with an implementation plan that is far away
from reality and from the aim of the workshop. A study by
Amabile & Teresa (1983) directly tested the hypothesis that
intrinsic motivation is conducive to creativity and extrinsic
motivation is detrimental. Seventy-two young adults participated in a creative writing test in individual laboratory sessions where people had been asked to write two brief poems.
Before writing the second poem, subjects in an intrinsic orientation condition completed a questionnaire that focused on
intrinsic reasons for being involved in writing. Subjects in an
extrinsic orientation condition completed a questionnaire that
focused on extrinsic reasons. Poems written under an extrinsic
orientation were signicantly less creative than those written
in the other two conditions. VE workshops are more important than writing poems; when members are intrinsically motivated their creative ability will be highly dependent on the
persons way of cognitive expansions to generate ideas and his
cognitive analyses of problems. Understanding what would
keep a members mind busy can inuence what is required
from him to generate creative ideas or analyzing of functions
or producing alternatives during the workshop. Although, extrinsic motivators are essential, intrinsic motivators have the
greater eect on people. VE practitioners have to realize that
members will generate better ideas when they are internally
motivated to do so.

Expectancy Theory
Expectancy Theory expresses that the more attractive the
performance of a task, the more motivated the worker will be
to perform it (Maloney, 1986). The greater the challenge in
the hardness to identify basic functions of problems or nding other alternatives induced in a VE workshop, the greater
the motivation that the member will perceive to do the job
and come up with new solutions. Expectancy Theory suggests
that a highly motivational work climate requires strong expectancies, instrumentalities and valance. Lawler divided the
concept of expectancy into two specic types in the mind of
the individual. First, each behavior has associated with it an
expectancy or belief about the probability of success. Second,
certain outcomes associated with every behavior. The nal element in the expectancy model is that the attractiveness or va6

Volume 32, Number 2, Summer 2009

lence of outcomes (Laufer, 1983). The question is how attractive or unattractive it is to an individual is determined by the
particular set of values an individual holds and reects many
factors in the individuals life. Expectancy Theory argues that
it is important to measure the attitudes individuals have in order to diagnose motivational problems. In Laufers hypothetical model to measure motivation:
M = (EP) [(P O)(V)] . (1)
Where,
M = Motivation level of workers
E = Eort exerted by workers
P = Performance level measured from outputs
O = Outcome received form eort exerted
V = Valance of the outcome perceived by workers
Laufer (1983) included in his study on labors (bricklayers) that the work motivation scores are determined by multiplying the strength of the belief that the eort leads to performance by the sum of the beliefs that the performance leads
to outcomes weighted by the attractiveness of the corresponding outcomes. Every VE member believes that an eort spent
leads to performance expected, and each performance has an
associated outcome that will vary in its valance depending on
each members own perception; i.e., a member who is generating 10 eective and creative ideas believes that it took him to
do some hard eort to produce these ideas; thus, he performs
very well during the workshop, so he, in return, expects a nonmonetary outcome in the form of praise from the facilitator,
which he perceives as a valuable compliment he intrinsically
needed to feel he has accomplished a goal or that he has contributed positively to the success of the project.

Motivation and Experience


Experience is directly related to how much work people
perceive themselves capable of doing; it has also another relation to the motivational impact of the work outcome. A study
made by (Bartkus, Hartman, and Parent, 1994) to investigate
the trade-o between motivation and experience focuses on
travel services. It concluded that highly experienced but less
motivated travel agents performed no better than highly motivated but less experienced ones. They [concluded] that recruiting decisions should be based on long-term considerations
and the internal competencies of the agency through managerial implications. Long-term considerations and competencies
can be well managed by the certication programs; e.g., SAVE
Internationals certication program. This paper considers
members level of experience and their perceived capability as
the major inuence to be highly motivated.

V A L U EW O R L D

1) General: this section is measuring general information of


respondents i.e. level of experience.

Table 1. Posive and Negave Movaonal


Outcomes of the VE
Feeling of accomplishment
Posive

This paper classies the sample based on years of experience and studies their capabilities, relating them with motivational impacts, existence and valance from the VE outcome.
To identify how intrinsic measurements of motivation will be
measured, the Laufer model will be used to determine peoples
productivity in VE workshops by measuring the perceived capability of the members and measuring the motivational outcomes that are more likely to exist at the organization with the
valance of these outcomes. In 2007, 50 questionnaires were
distributed to employees of the organization under consideration hand-by-hand or via e-mail. The questionnaire had three
sections:

rst section was scaling the rate of occurrence of all the variables in Table 1.

Supervisor parsing
Higher points in SAVE
Skill development and learning

Negave

Paper Methodology

Blamed for leading duty


Work team will be angry
Personal ideas not implemented
Final report not implemented

This was measured by a scale, where, always occurs = 5,


2) Capability: this section is measuring the level of producusually occur = 4, often occur = 3, sometimes occur = 2, never
tion members perceiving they are actually capable of peroccurs = 1.
forming:
The other section has two parts: one is rating valance of
Idea generation: this measured by a scale of quantity
positive outcomes as, always desirable = 5, usually desirable =
The second part, was
rating
the valance
of negative
= 5,
of ideas
generated
starting
from 20 outcomes
and aboveas,
= 5,never
20 desirable
4, often desirable = 3, sometimes desirable = 2, never desirable

11
=
4,
10

6
=
3,
less
than
5
=
2,
1
idea
=
1.
ometimes desirable = 4, often desirable = 3, usually desirable = 2, always desirable
= 1.
= 1.
Functional analysis: this is also measured by a scale of
The second part, rated the valance of negative outcomes
quantity, where, more than 15 = 5, 15 11 = 4, 10
as,
never
desirable = 5, sometimes desirable = 4, often desirable
ults
6 = 3, less than 5 = 2, 1function analyzed = 1.
= 3, usually desirable = 2, always desirable = 1.
Producing alternatives: this is measured by a scale
ranging
fromwith
always
do = 5, values
usuallywhich
do = 4,
often
= 3,net response
dback came from 21
engineers
no missing
mean
42%
rate.
Results
sometimes = 2, never = 1.
jority of the sample as shown in Figure 1 is from 4 12 years of experience. All of the
The feedback came from 21 engineers with no missing
3) Motivation: this section contains some possible negative
ents were working for the same governmental organization 7 Civil, 4 Architectural,
4 meant a 42% net response rate. The majority of
values,
which
and positive outcomes that might exist and measures the
ical, 3 Electrical, 1 Bio Med, 1 Computer, and 1 Systems Engineers showing in Figure
2. One as shown in Figure 1 (below) has 4 12 years of
the sample
valance of each outcome is perceived by all members Taexperience.
was a Consultant
Value Specialist CVS, another one was not certified at all and the rest of the All of the respondents were working for the same
ble-1.
governmental organization: 7 civil, 4 architectural, 4 mechaniwas all Associate Value Specialists AVS. This indicates that the sample size should be well
cal, 3 electrical, 1 bio med, 1 computer, and 1 systems engiThis section was split into two: one to measure the outwith the VE
methodology
and what
amount
of work expected
fromofthem.
neers showing in Figure 2 (previous page, bottom). One of
come
expected and
the other
to measure
the valance
it. The

Figure 1 - Respondents level of experience

Figure 2 - Respondents specializations

Figure 3 below shows clear variation on the extent of how members perceive themselves cap

V A L U EW O R L D

Volume
Numberespecially
2, Summer
2009
7
generating ideas with almost
little32,
variations
in mid
ages to analyze functions
or p

alternatives. The group of less than four years of experience perceives a mid capability range

The group of 13 20 years of experience perceives a mid range of idea generation and functional
Table 2 Capability analysis
analysis
with a high range of producing alternatives. The last group from 20 30 years of experience
Idea

Functional

Alternatives

generation

Analysis

generation

Comments

perceives a high range in idea and alternatives generation but low functional analysis capability.

This group perceive themselves capable but not to the level o


Mid
Mid
Mid
them was a Certied Value Specialist1 3
experienced people.
(CVS), another one was not certiThis group is lowering the perception of their own capability
High
Low
Mid
ed at all, and the rest of the sample4 7
were all Associate Value Specialists
analyzing functions,
but still can think outside the box.
High
(AVS). This indicates that the sample8 12
This
group
is
developing
a sense of knowledge in the process
Mid
Mid
High
size should be well familiar with the
Mid
and well familiar with similar products or projects.
VE methodology and what amount of
As previous.
13 20
Mid
Mid
High
work expected from them.
Low
This group is perceiving high capabilities but low in knowing
High
Low
High
Figure 3 (right) shows clear varia-20 30
processes and other alternatives.
tion on the extent of how members
perceive themselves capable of gen-0.5 and below = low, 0.5 0.69 = Mid, 0.7 and above = High
erating ideas with almost little varia-Idea generation = capability of throwing large number of ideas (quantity not quality), this depends on thinking out of the box.
tions especially in mid ages to analyzeFunctional analysis = capability of analyzing functions of the product or project, this depends on level of experience of products or projects under
replacing a product or project parts or as a whole by other alternatives, this capability depends on kn
functions or produce alternatives. TheAlternatives generation = the capability ofFigure
3. Perceived capability of the members
group of less than four years of experi-other similar products or projects.
ence perceives a mid capability range
for all three factors. The group of 4 7 years
- 10 of experience perceives a high capability to generate ideas, but low in functional analyses and
alternatives generation. The group of 8 12
High
years of experience perceives a mid range in idea
Mid
generation and producing alternatives and a low
functional analysis rate. The group of 13 20
Low*
years of experience perceives a mid range of idea
Low
generation and functional analysis with a high
Mid*
range of producing alternatives. The last group
from 20 30 years of experience perceives a
High*
high range in idea and alternatives generation
but low functional analysis capability. (See Table
2, previous page, below.)
As an average of sum of all capability ratings
*represents the scale of negative outcomes expected.
Figure 4 (right) shows the total average capabilFigure 4. The relationship between Capability VS. Motivational Outcomes.
ity rating, which indicates that the most critical

Table 2. Capability Analysis


As
an
average
of sum of all capability ratings Figure 4 shows the total average capability ra
Funconal
Alternaves
Analysis
Generathat
on theComments
which indicates
most critical age to be motivated is mid age group. Figure 4 also shows

Years
Experience

Idea
Generaon

1-3

Mid

4-7

High

members. It is
classified
as group
positive
and negative
factorson
onofmembers:
Low
Mid
This
is lowering
the percep
their own capability in
analyzing funcons, but can sll think outside the box

8 - 12

Mid

Mid

High

This group is developing a sense of knowledge in the processes and


- 11 is well familiar with similar products or projects

Mid

Mid

This group perceive themselves capable, but not to the level of


experience people

perception of both expected outcomes rating in Table-1 times the desirability rating perceived

13 - 20

Mid

Mid

High

As previous.

20 - 30

High

Low

High

This group perceives high capabilies, but low in knowing processes


and other alternaves

0.5 and below = Low, 0.5 - 0.69 = Mid, 0.7 and above = High
Idea Generaon = Capability of throwing large number of ideas (quanty, not quality); this depends on thinking outside the box.
Funconal Anlaysis = Capability of analyzing funcons of the product or project; this depends on level of experience or products or projects under study.
Alternaves Generaon = The Capability of replacing a product or project parts or as a whole by other alternaves; this capability depends on knowing other similar
products or projects.

Volume 32, Number 2, Summer 2009

V A L U EW O R L D

Table 3. Posive and Negave Factors


Years
Experience

Capability

Posive
Impact

Negave
Impact

1-3

Mid

Low

Mid

They work even if they are not posively inuenced.

4-7

Mid

Mid

Mid

Average on all factors.

8 - 12

Mid

Mid

Low

Less aected by negave outcomes.

13 - 20

Mid

Mid

Low

Less aected by negave outcomes.

20 - 30

High

Low

Low

They produce hard work and they dont care what happens.

Comments

Capability: 0.5 and below = Low, 0.5 - 0.69 = Mid, 0.7 and above = High
Posive Impact: 0.5 and below = Low, 0.5 - 0.69 = Mid, 0.7 and above = High
Negave Impact: 0 - 0.19 = High, 0.2 - 0.39 = Mid, 0.4 and above = Low

age to be motivated is mid age group. Figure 4 also shows


the perception of both expected outcomes rating in Table-1
times the desirability rating perceived by members. It is classied as positive and negative factors on members. (See Table
3, above.)
Which may be inferred that if an employee perceives himself less capable of doing work and he is less aected by positive outcomes and he is more aected by negative outcomes,
then he is the one who should be given the most attention;
i.e., he is the one who should get the focus of motivation. In
the study made to VE practitioners, the [experience] between
8 12 years is the most critical [period]. Maybe because 8
years of employment is between the beginner who thinks he
is not adding value to workshop, in a sense that he is feeling that his idea is old enough that everybody knew it or he
sees people with more experience in a better production coming from their intuition and knowledge throughout the years.
Other added points coming from the respondents that motivate them were, they have to feel the synergy between team
members, delegating responsibilities within team members
as positive motivator, or the feeling that the team is lacking
knowledge and experience as a negative demotivator.

skills attracts more people to participate and work harder in


VE workshops. A lot of research [is performed] on motivation,
but with no structured methodology to measure it; Laufer and
Jenkins (1983) used Vroom ET to describe a method to make
the measurement easier. This paper used this method to determine the relationships between the perceived capability of
members and the eect of negative and positive outcome and
its relationship with progressive experience.
The results are intended to enlighten management in how
to deal with their employees as customers who have needs and
requirements that should be fullled to produce hard work.
Those needs should not always be tangible, such as money
or bonuses; other outcomes can have greater impacts when
considered and properly awarded. It seems that more improvement opportunities can be made at least to align the probability with desirability levels of outcomes and more improvement
is more likely to exist if the probability is increased at all experience levels. There is a great relationship between experience
levels and the perception of the capability of doing hard work.
At early [experience levels], people consider they are more capable of doing hard work, when, with time, the perception
decreases and after 12 years of experience it may increase.

Conclusion

References:

Finally, this paper concludes that from 8 12 years of


experience, VE members in the organization studied perceive
themselves as less capable of doing work, more aected by
negative outcomes, and less aected by positive outcomes.
Management should give more attention to those people who
can be categorized as practiced the profession but yet not expert enough [to] contribute remarkably. These measures are
not absolutely xed; they vary from one organization to another: The weight of the outcome (Probability * Valance) is
nally determined then the perceived capability of members is
compared to their motivation level. Imperatively, facilitation
and management of VE workshops always has a chance of improvement. The proper facilitation with up-to-date tools and

Al-Youse, Abdulaziz S. (1998). Value Management: concept


& Techniques. KSA

V A L U EW O R L D

Jones, Gareth R. & George, Jennifer M. (2008). Contemporary Management, 5th edition, McGraw-Hill, New York,
USA.
Vroom, V. (1964).Work and Motivation, Wiley, New York,
USA.
Van Eerde, W. and Thierry, H. (1996). Vrooms expectancy
models and work-related criteria: a meta-analysis, J. Appl.
Psychol. 81, pp. 575586.
Lawler, E. (1971). Pay and Organizational Eectiveness, McGraw-Hill, New York.
Volume 32, Number 2, Summer 2009

Kaufman, J. Jerry (1998). Value Management: Creating Competitive Advantage. USA. Thomason Crisp Learning.
Woodhead, R., M. (1998). Can Any Facilitator Run a value
engineering Workshop? School of Construction & Earth
Sciences, Oxford Brooks University, UK.
Norton, B.R. and Mcelligott, W.C. (1995). Value Management in Construction, MacMillian Press, London, UK.
Miles, L.D. (1972). Techniques of Value Analysis and Engineering, McGraw-Hill, p295.
Maloney, William F. (1986). Understanding Motivation,
American Society of Civil Engineers, Journal of Management in Engineering vol.2 No. 4.
Amabile, & Teresa, M. (1983). Motivation and Creativity: Effects of Motivational Orientation on Creative Writers, Paper presented at the Annual Convention of the American
Psychological Association, USA

Join SAVE International now


and avail yourself of the opportunities to

Laufer, Alexander & Jenkins, G. Douglas (1983). Motivating


Construction Productivity: Learning From other Disciplines, Project Management Quarterly.
Bartkus, Kenneth R. Hartman, Cindy L. & Michael Parent,
C.R. (1994). The Performance Trade-o between Experience and Motivation: Its Impact on Travel Service Selling.
Journal of Travel Research, Vol. 33, No. 2, 15-20.

About the Author


Abdullah Khalid AlMedallah, AVS, Yellow Belt-Six Sigma is a
civil engineer employed by the Royal Commission of Jubail and
Yanbu in Saudi Arabia.

Call for Papers, Workshops

ADD VALUE TO YOUR CAREER.

Name
Job Title
Company
Mailing Address

City

State/Province

Postal Code

Country

Telephone

Fax

Email

SAVE International is accepting abstracts and


pre-conference workshop proposals for the
50th Annual Conference to be held June 6 - 10,
2010, aboard the Queen Mary in Long Beach,
California.
SAVE will only accept submissions made online
through the SAVE website: www.value-eng.org.
The call for abstracts and pre-conference
workshop proposals ends November 30, 2009.
10

Volume 32, Number 2, Summer 2009

Membership Type (check one)


Individual - $150
Corporate - $415

Sustaining Corporate - $1065

Payment Method (check one)


Check (enclosed)
MasterCard

Visa

American Express

Card Number Expiration


Cardholder Signature

Return your application and fee payment to

SAVE International
136 South Keowee, Dayton, OH 45402 USA
T (937) 224-7283 / F (937) 222-5794
info@value-eng.org www.value-eng.org

V A L U EW O R L D

Best-in-Class:
An Executive Level Application of the Value Methodology
Ronald J. Tanenbaum, Ph.D., PE, CVS, GE, F.ASCE and Gordon Johnson, PE

Abstract
Sometimes we, as value methodology practitioners, do not
see all of the opportunities to which the methodology, or even
selected components of the methodology, can be successfully
applied. Then, along comes a forward-seeing agency that has
a goal to self inspect its operating procedures and to change
these procedures in its quest to become the Best-in-Class in
their arena of practice and service to the public. This was the
case with the Metropolitan Water District of Southern California who approached Value Management Strategies, Inc. to
lead a series of unique workshops addressing three critical areas
where the District saw opportunities to improve: Alternative
Delivery Methods, Engineers Cost Estimates, and Construction Change Order Management. This paper will address the
following:

How a Best-in-Class workshop correlates with and utilizes


the VM methodology;

Composition of the VM Team(s) necessary to complete a


successful workshop;

How to conduct three workshops in three days with three


independent teams;

General ndings from each of the workshops including


selected Quick Hits;

Creating a living document that allows the District to


implement selected ideas and modify the document to
meet its ever-changing future needs.

tan to enhance the Engineering Services Section performance


in areas of alternative project delivery methods, engineers cost
estimating, and construction change order management. The
primary objective was to identify the best business practices of
leading-edge companies and agencies, and also identify pitfalls
to be avoided to increase the probability of project success.
With an objective of soliciting ideas and proposals from
outside experts and not from Metropolitan sta, a total of 10
independent consultants specializing in Alternative Project
Delivery (ADP) Methods, Engineers Cost Estimating (ECE),
and Construction Change Order (CCO) Management (three
to four each, relatively) were selected to participate in the Best
in Class Benchmarking Initiative PA study. The PA study was
organized in one pre-study meeting and three one-day workshops to address the three topics.

Goals
In identifying the Best-in-Class processes for each of the
three topic areas, Metropolitan encouraged each team member to focus on general, holistic, best in class methods utilized
in industry, specically the water industry, and not on the
specic practices of Metropolitan. Metropolitan assessed the
proposals generated at a later date, and determined how best
to adopt those proposals oering the most opportunity for
improvement of services. Below is a list of goals or suggested
objectives that were utilized to stimulate the creative thoughts
of the teams.

The paper will conclude with recommendations for performing successful Best-in-Class workshops for any agency
that shares a goal to self inspect their operations, accept change,
and possess a sincere desire to be better, if not the best.

Dene what other leading agencies/companies (the industry) are doing to provide the best practice in each area.

Provide expert assessment on where the Water Industry


is today with regard to each topic, and where it will be
tomorrow.

Introduction

Through the methodology of facilitated workshops conducted with respected industry consultants known to have a
wide degree of exposure and experience working with leadingedge companies, the Best-in-Class Benchmarking Initiative
Process Analysis (PA) study sought to identify best business
practices and/or procedures that can be adopted by Metropoli-

Identify how Metropolitan might then implement an improvement program.

Suggest how the measurement of results might occur.

Utilize the teams considerable experience in identifying


what they see as working and not working in the industry.

Rene what Metropolitan does to assist a consultant in

V A L U EW O R L D

Volume 32, Number 2, Summer 2009

11

providing a quality product.

Provide an inventory of what alternative project delivery


methods are used and what would be appropriate for this
industry.

Assess whether the ve class method of the Association


for the Advancement of Cost Engineering International
(AACE) is the preferred estimating approach, or are there
other methods to consider.

Identify external estimating programs that would be recommended to Metropolitan.

Identify other sources of labor units, etc. (other than


Means).

Identify external estimating standards that Metropolitan


might use.

Assess the track record of Partnering and whether it has


proven to be successful in the industry.

Address what other owners are doing to negotiate, handle,


monitor, and decide to proceed with their construction
change order process.

Based on these goals, the PA participants prepared a list of


proposals under each topic with rational and action items.

State of Practice
Alternative Project Delivery Methods

The traditional model for developing and operating public capital works in the United States has been the design-bidbuild-operate model. Three basic parties constitute this model: designer, constructor, and owner/operator of the facility.
Today, however, cities, towns and special districts across the
country are faced with the unenviable task of delivering capital
projects with budgets that are not always in line with project
requirements, time lines that are signicantly shortened, or
risk proles that do not align with owner capacities. While
the traditional model is well understood, it has some inherent
advantages, and is the basis for much of the public works law
across the country. There are shortcomings that owners have
addressed by turning to a variety of alternative project delivery
methods.
These organizations are looking for new, more innovative
means of project delivery, that range from construction management at risk to design-build and design-build-operate to
turnkey solutions that incorporate nancing and ownership.
Accordingly, there are many variations of alternative delivery
that may be applied by organizations to meet their capital project needs. Owners must choose what best meets their need; a
need that is not necessarily static. Requirements and context
vary from project to project and many owners nd that a tool
box of solutions is the best approach to maintain exibility
12

Volume 32, Number 2, Summer 2009

and provide the greatest return on capital works investment


funds.
The Best-in-Class workshop for alternative project delivery methods provided input on the large range of issues and
potential approaches for delivering capital works projects in
todays environment, presenting some the many innovative
approaches that have been successfully applied across the industry.

Engineers Cost Estimating

In order to allow owners to properly budget for future


needs, cost estimates must be available so that anticipated
needs can be adequately evaluated. These cost estimates can
be wide ranging, starting at a blue sky type estimate where
there is no data except a quantitywe will need a 50 million
gallon a day water treatment plant in the xyz area around
year 2015to the case where there is an exact location, 100
% design, and a denite anticipated start date. Handling this
range of estimating within the parameters of a variable and
rapidly changing pricing environment and global market conditions is a very dicult task.
The 5 class method of the Association for the Advancement of Cost Engineering International (AAEI) is the currently preferred estimating approach in the costing of water projects. Especially important to their concept of estimating is the
application of the appropriate contingencies and escalations.
To be successful and accurate within the parameters available
at the time the estimate is developed, the estimate should be
performed by people experienced in estimating, who have access to the proper tools: the Owners historical cost records for
similar projects, published manuals for labor units, material
pricing, equipment rental rates, and ination cost indices are
examples of required data. If large projects are routinely being
estimated, Owners are encouraged to use a proprietary estimating system.
Many projects in the water industry can be classied as
typical work, work that is done routinely and repetitively,
work such as installing runs of small diameter pipelines. Historical data with cost indexing can be used to price this type of
project. For larger more unique projects, a signicant amount
of time and eort is required to create a responsible estimate.
It is important for Owners to note that engineering estimates are a best guess at a specic point in time. For example, those estimates prepared in August of 2005 were no
longer valid in September of 2005 (Katrina hit August 29,
2005). The Best-in-Class workshop addressed this point and
examined the many options available to Owners to improve
the cost estimates generated internally or received from consultants while eectively communicating the accuracy and
meaning of these estimates to decision makers.

V A L U EW O R L D

Construction Change Order Management

Despite the ne eorts of Owners to dene project requirements, and design professionals to produce complete,
accurate and reliable construction projects, virtually every
project experiences the need to modify, add to, or otherwise
alter the project after opening the bids. Changes during construction are a routine occurrence within the water industry,
where projects often involve modications to existing complex
facilities (such as water treatment plants) or involve subsurface
construction projects that, by their nature, carry some inherent risk and uncertainty (subsurface construction such as pipelines and tunnels). Change orders (i.e. alterations considered
within the scope of the original contract) may be necessary
to incorporate contractor suggestions, respond to unforeseen
site conditions, or correct errors and omissions in the plans or
specications.
Even the most successful projects have change orders.
What often separates successful projects from unsuccessful
projects is how construction change orders are managed. Over
the last several decades the water industry marketplace participants including owners, designers, construction managers and
contractors have developed industry standard practices, procedures and tools to help manage construction change orders.
Leading agencies and companies in the water industry that
design and construct projects have developed best practices
which improve the eectiveness and eciency of managing
construction change orders. Some of these best practices include improvements in design procedures, construction contract writing, construction records management and tracking,
construction management stang, and improvements to construction management roles and responsibilities. During the
construction change order management Best-in-Class benchmarking workshop, many of these best practices were identied, rened and recommended for potential implementation
on future Metropolitan projects.

Findings from Workshops

After an intense three days, the three independent workshop value engineering teams generated numerous recommendations for consideration by Metropolitan. A total of 211 ideas (62 ADP; 69 ECE; and 80

CCO) were identied during the Creativity Process from


which 86 proposals (some consisting of combined ideas) were
prepared by the teams, with the balance rejected during the
workshops. These were categorized into Quick Hit Proposals
and Recommended Proposals. Each proposal contained a brief
description justifying its recommendation along with a series
of bullets indicating the steps or actions required for implementation. The proposal presented below provides an example
of the format utilized:
Use Alliance Contracting (Australian Model)/Form Joint
Venture between Metropolitan and Private Industry
This suggestion proposes to evaluate and adopt a version of
the alliance form of contracting. This approach has been very successfully applied in Australia and the U.K. for public infrastructure projects and within the energy exploration eld and heavy
industry for complex and high risk projects. For more in depth
information, see attached white paper Introduction to Project
Alliancing located in Appendix F. Benets of alliancing include
assurance of better performance in non-process areas combined
with likelihood of lower direct costs, greater certainty of on-time
or early delivery, better informed decisions on technical solutions,
better balance between capital investment and life-cycle costs, and
collaborative management and alignment of commercial interests.
Concerns include potential perception of uncertainty in cost outcomes, signicant involvement and commitment of Owner sta
and senior management to be eective, and overall cultural shift.
Implementation:
Review alliancing as a potential form of project delivery to be
included in the Metropolitan portfolio of delivery options.

Identify the potential legal hurdles that must be overcome to


make alliancing an option for Metropolitan.

Identify a team of Metropolitan personnel that would be specically charged with dening and chartering potential alliancing projects.

Prioritize a list of potential Metropolitan projects to be delivered under an alliance contract. The approach requires eort
and probably cannot be justied for all projects.
The Quick Hit Proposals were evaluated to be immediate-

Table 1. Type and Number of Proposals


Quick
Hits

Selected
Proposals

Not Selected
Proposals

Current
Metropolitan
Pracce

Total
Proposals

Alternave Project Delivery Mehthods

15

28

Engineers Cost Esmang

22

31

Construcon Change Order Management

27

10

23

46

86

Performance Area

TOTAL

V A L U EW O R L D

Volume 32, Number 2, Summer 2009

13

ly implementable to produce direct results


related to the goals of the PA study. After
review by Metropolitan management, the
balance of the Recommended Proposals
were either selected or not selected. Some
were found to be current Metropolitan
practice and were so indicated. This resulted in following quick hit/selected/not
selected/current Metropolitan practice
number of proposals. (See Table 1, previous page, bottom.)

Quick Hit Proposals


Ten quick hit proposals were selected
as a result of the ndings from the workshop. It was felt by Metropolitan that these
ten proposals could be implemented without any cross-department impacts or approvals. Reproduced below, as an example,
is one of these quick hit proposals. It is presented to demonstrate how Metropolitan
is recording and tracking the action items
and progress of each proposal that has been
selected. (See Table 2 right.)

Findings From Workshops


Appended at the end of this section of
the paper is a table listing all of the proposals considered by the Metropolitan Water
District of Southern California. The color
coding indicates how these proposals were
to be resolved (quick hits, selected proposals and not selected proposals). These decisions were based on the following considerations:

Table 2. Construcon Change Order Proposal


Construcon Change Order
Proposal # 8 (#71)
Perform Constructability Reviews/Lessons Learned and Proper Project Closeout
Documentaon
Summary:
It is recommended that lessons learned documentaon be developed at the
closeout of each project to address constructability issues that can be beer
implemented on future projects and to compare change order frequency and
cost to other similar projects. This is a process that should start at the beginning
of construcon using design feedback forms, to keep the design team informed
of problem areas unconvered in the documents or specic process issues noted
during installaon and startup. Use this informaon as a transfer of knowledge
to the design or construcon teams, as necessary, to limit future dicules or to
promote technology or construcon pracces that went very well and should be
implemented on future projects.
Acon Item

Develop and use


standardized forms
to relay informaon
from the eld back to
the design team during construcon.

Perform a closeout
review and recommendaons document to idenfy the
pialls and posive
lessons learned on the
project.

Maintain budget to
perform this task.

If a proposal contained benets which appeared to outweigh the cost of resources anticipated

for research and implementation, it was selected for further development.

If a proposal was considered to be cost prohibitive, it was


not selected.

If there was a legal impediment, that is the current regulatory framework of the State of

California prohibited certain actions/methods by the District, it was not selected.

If a proposal was already partly or completely implemented, it was so identied and not selected for further development.

14

Status

Volume 32, Number 2, Summer 2009

Current documentaon
provides the Project
Manager and Designer
with infomaon about
the changes and issues
that are occuring on the
contract. As-built drawings
also indicated what had to
be changed in the eld.

Date of Compleon
Ancipated

Actual

2007

Completed

Current pracce (for larger


projects) is to perform a
lessons learned analysis,
although not always
performed.
Need Task Descripon for
budgeng purposes.

Of the Quick Hit Proposals, three have been implemented, one is pending, four are anticipated to be completed by
winter 2009 and two remain undecided. Of the 23 Selected
Proposals, three have been completed, ve are anticipated to
be completed by winter 2009, no decision has been made regarding 12 proposals, two are classied as being on-going, and
one has now been considered to no longer be applicable due to
cost concerns to the District if they were implemented.
One accomplishment that came out of the best practices
process was that Metropolitan created a prequalication list
for urgent pipeline repair projects. That was covered under
one of the proposals (16 & 17-QH below) that suggested the
creation of such a prequalication list. Successfully making
this change has created more exibility in contracting for these
repairs. It is believed that this single accomplishment is a big

V A L U EW O R L D

General Managers Best-in-Class List of Proposals


Legend
Study Categories
QH

Quick Hit Proposals

SP

Selected Proposals

NS

Proposals Not Selected

CP

Current Metropolitan Pracce

PROPOSAL
NO.

PROPOSAL

STATUS

A. Alternave Delivery Methods


1-NS

Acvely Seek Statutory Authority for Design/Build; Issue Aggressive Legal Interpretaon of Exisng Status; Study
Statute of Other States (Arizona); Use Caltrans Model to Establish Statute Authority

2-NS

Structural Design/Build Delivery Model that Fits Metropolitan and Industyr Constraints

3-SP

Develop Project Delivery Selecon Process

Sept. 2009

4-SP

Use Alliance Contracng (Australian Model)/Form Joint Venture Between Metropolitan and Private Industry

Undecided

5-NS

Consider Appropriate Involvement of Third-party Advisors, Consultants, and Program Managers to Facilitate
Successful Negoaons and Delivery of Projects

6-NS

Form Joint Powers Authority with Agencies that Have Design/Build Authority

7-QH

Survey Member Agencies and Private Water Companies to Understand How They Deliver Contracts

8-NS

Opmize Standard Forms and Contracts

9-NS

Develop Alternave Project Delivery Methods (APDM) Implementaon Team; Organize Separate Organizaonal Unit
within Metropolitan for APDM

Sept. 2009

10-QH

Review and Update Metropolitans Exisng Metrics; Dene Acceptable Metrics

Sept. 2009

11-SP

Rene or Revise Metropolitan Business to Incorporate Alternave Project Delivery Methods

Undecided

12-QH

Dene Dierent Criteria for Lifeline versus Non-Lifeline Facilies; Use Design/Build in Non-Process Facilies

Undecided

13-NS

consider Use of Design-Construcon Management at Risk

14-SP

Dene the Opmum Risk Allocaon that Meets its Needs and Conforms to Best Pracces in the Industry

Undecided

15-SP

Ulize Construcon Management at Risk as a Means of Projects Delivery

Undecided

16 & 17-QH

Consider Task Order-Based Contracts for Small Projects and Develop List of Pre-Qualied On-Call Contractors and
Design-Builders

Sept. 2009

18-SP

Idenfy Small Business Enterprise Opportunies and Challenges with Alternave Project Delivery; Ulize MentorProtogee Program with Small Business Enterprise Program

Undecided

19-SP

Ulize Incenve Contracng to Improve Project Delivery and Overall Performance of Delivery Teams

Undecided

20-NS

Use Design/Build for Greeneld Projects

21-NS

Use Design/Build for Plant Modicaons with Mulple Measurement and Payment Provisions; Ulize Flexible
Measurement and Payment Provisions for Rehabilitaon Work

22-NS

Do Not Use Design/Build for Projects that Will Employ Unproven Technology

23-SP

Tie the Delivery Approach to Programmac Use of Building Informaon Systems or Three-Dimensional Models or
Asset Management

24-NS

Ulize Design/Build/Operate

25-NS

Ulize Cost Reimbursable Contracng with Fixed Fee for Complex Facilies

26-NS

Consider Target Price Contracng

27-NS

Consider Open Book Negoated Design/Build Contracts

28-NS

Use Guaranteed Maximum Price Approaches with Shared Savings and Allowances

V A L U EW O R L D

Volume 32, Number 2, Summer 2009

Undecided

15

General Managers Best-in-Class List of Proposals


Legend
Study Categories
QH

Quick Hit Proposals

SP

Selected Proposals

NS

Proposals Not Selected

CP

Current Metropolitan Pracce

PROPOSAL
NO.

PROPOSAL

STATUS

B. Engineers Cost Esmang


29-NS
30-NS
31-NS
32-NS
33-NS
34-SP
35-NS
36-SP
37-NS
38-NS
39-NS
40-QH
41-SP
42-SP
43-SP
44-NS
45-NS
46-NS
47-NS
48-NS
49-NS
50-NS
51-NS
52-NS
53-SP
54-QH
55-NS
56-SP
57-NS
58-NS
59-NS

16

Fine Beer Index than Engineering New Record Construcon Index


Reduce/Modify/Idenfy Onerous Front End Condions (Schedule, Liquidated Damages, Project Labor Agreements,
Unforeseen Condions, M/W/SBE
Evaluate Current Esmang Tools to Determine if They are Best-in-Class (Timberline, Mc2, US Cost, Success, Excel,
Etc.)
Develop Database of Costs; Metropolitan Should Hire a Third Party to Develop Database; Find Industry Standard of
Database for Esmang Accuracy (AACE)
Consider American Naonal Standards Instute Method for Dening Cost Esmang Methods in lieu of Associaon
of Advancement Cost Engineering Internaonal
Invesgate Benchmarking Data from Construcon Industry Instute
Conduct Market Analysis Prior to Preparing Engineers Esmate
Establish Esmang Standard for Mulple-Firm Esmates; Metropolitan (Possibly with Third-Party Consultants) to
Develop Esmang Manual
Provide Addional Budget to Complete Engineers Esmates; Establish Reasonable Budget for Engineers Esamte
Develop Cost Factors
Perform Monte Carlo Analysis to Idenfy Risk and Conngency and to Evaluate Which Risks are Driving the Upfront
Costs and Whether it May Make Sense to Transfer Risk
Provide Esmates Throughouth Design Process
Educaon Organizaon, Both Board and Management, on Goals for Engineering Esmates; Reeducaon of Board
and Management
Retain On-Call Consultants to Perform Cost Esmang
Educate and Train Metropolitan Esmators
Evaluate Salary Structure for Cost Esmang Sta (Adjust to Industry Standard)
Have Construcon Type Rather than Engineering Type Conduct Cost Esmates; Ulize Consultants Constructors
Division to Prepare Esmates
Try Mulple Esmates on Several Jobs (Both Internal and External) to Compare and Review Dierences
Avoid Unfriendly Interface of Owner/Engineers/Designers with Contractors
Meet with General Contractors to Discuss Front Ends and Possibly to Discuss Projects in Early Final Design (Ongoing
Dialogue)
Provide Incenves for Accurate Esmates to Metropolitan Sta and Consultants
Be Aware of Bonding Problem for Projects Over $100 Million or Lasng Longer than Three Years; Have Discussion
with Bonding and Insurance Companies to Beer Dene Front-End Documents
Understand the Consultant Engineers Track Record in Producing 100% Quality Design
Hire Cered Esmators (in Dierence Disciplines) and Hire Esmang Manager or Chief Esmator
Apply Total Quality Management Tools to Cost Esmang
Incorporate the Review of the Cost Esmate in Value Engineering Studies
Control the Addendum Process Where You Have Many or Complicated Addenda (Especially Close to the Bid Date);
Honor Requests from General Contractors to Extend the Bid Date
Comprise List of Esmang Consultants with Specialized Fields of Experse
Consider Escalaon Clause Where the Owner and General Contractors Share Risk
Allow the Modeling Tools Used by the Design Engineer to be Available to the Bidders (Three-Dimensional Modeling)
- See also Proposal 23
Allow Contractors to Create Alternate Bid on Modied Schedule

Volume 32, Number 2, Summer 2009

July 2009
Completed

Completed
On-going
N/A
On-going

June 2009
Completed

June 2009

V A L U EW O R L D

General Managers Best-in-Class List of Proposals


Legend
Study Categories
QH

Quick Hit Proposals

SP

Selected Proposals

NS

Proposals Not Selected

CP

Current Metropolitan Pracce

PROPOSAL
NO.

PROPOSAL

STATUS

C. Construcon Change Order Management


60-CP

Perform Formal Risk Analysis of Design Documents; Idenfy Undened Scope Items

61-NS

Upgrade and Enhance Construcon Administraon Management System

62-CP

Track Discipline of Change Orders (Structural, Mechanical, Instrumentaon, Electrical)

63-NS

Communicate the Reality of Change Orders to the Board of Directors and Research Industry Standards

64-NS

Consider More Use of Allowances

65-NS

Include Unit Price Costs in Contract for Change Orders (for High-Risk Items)

Current Pracce
Current Pracce

66-SP

Reduce Modicaon of Standard Specicaons

67-NS

Aract Qualied and Experienced Trade Contractors

68-SP

Form Internal Instrumentaon and Control Review Plan Check Team or Consider Commissioning a Task Force
Aimed at Reducing Electrical and Instrumentaon and Control Change Orders

Undecided

69-SP

Metropolitan to Develop Construcon Standard Review Checklist for Use During Design Review (30%, 60%,
90%) and Maintain to Keep Current

Undecided

70-QH

Streamline and Improve the Change Order Process

71-QH

Perform Constructability Reviews/Lessons Learned and Proper Project Closeout Documentaon

72-CP

Construcon Schedules: Establish Realisc Owner Goals and Mandate the Baseline Schedule; Ulize the Full
Contract Duraon

73-NS

Conduct Presidents Walks on Job Site

74-CP

Consider Improving Operaons and Maintenance and Field Inspecon Inputs

75-NS

Improve As-Built Drawings

76-CP

Implement Pending Issue Log During Design

Current Pracce

77-CP

Update Field Inspecon and Construcon Administraon Standards Manual

Current Pracce

78-SP

Perform More Pre-Design Exploratory Potholing and Site Invesgaon

Undecided

79-QH

Provide Adequate Budget for Design Services During Construcon

Undecided

80-SP

Create an Educaon Program for Internal Sta and Resident Engineers

Completed

81-SP

Iniate Designer Evaluaon; Perform Contractor Evaluaon (i.e., Model Both Aer Corps of Engineers Format)

Undecided

82-SP

Implement Three-Dimensional Design - See Also Propoals 23 and 58

Undecided

83-CP

Idenfy Hazardous Materials Pre-Bid and Contract On-Call Services for Remediaon Services

84-SP

Rate and Ulize Partnering Facilitator

85-NS

Consider Management of Escalaon in the Contract Documents

86-NS

Obtain Field Inspecon Sta Input and Bid Packaging.

V A L U EW O R L D

Undecided

Pending
Completed
Current Pracce

Current Pracce

Current Pracce
Completed

Volume 32, Number 2, Summer 2009

17

win coming out of the best practices process.

Future Actions

Once the assessment of all of the proposals has been completed by Metropolitan, it is planned that the results will be
shared with all of Metropolitans member agencies. In this
way, each of the member agencies will be able to benet from
the lessons learned in this process analysis workshop so that
the implementable Best-in-Class concepts would benet all of
the agencies operations. The member agencies would not be
required to implement the same or all of the concepts accepted
by Metropolitan, but could pick and choose those most appropriate to their operations. In the long run, a more consistent
operating procedure among the various agencies may evolve
and would facilitate better communications and more uniform contracting, estimating and construction monitoring.
The Best-in-Class Process Analysis Study is considered by
Metropolitan to be a living document, that is, as new ideas
and approaches are implemented and tested to improve the
Districts practices, changes may be made to the document
to keep it up to date. These changes may come in the form of
altering a selected proposal to better meet the new needs of the
District, selecting a previously unselected proposal because of
changed conditions (for example new regulations that permit actions previously prohibited), and removal of previously
selected proposals because they are no longer deemed necessary or economical or they have been superseded by improved
methods.

Process Analysis Team Member


Acknowledgment
Metropolitan selected consultants from leading-edge consulting rms to identify best business practices and/or procedures that can be adopted by Metropolitan to enhance the
engineering services section performance in areas of alternative project delivery methods, engineers cost estimating, and
construction change order management. These individuals
made up the process analysis (PA) teams and were responsible
for generating the proposals presented in this report. As such,
they deserve signicant recognition and thanks for the eort
required to complete these studies. In addition, key individuals from Metropolitans sta in each of the three topic areas
were included in the team.
The studies were performed between January and February 2007. The PA studies were led by Ron Tanenbaum, CVS
from Value Management Strategies, Inc. The PA team members are listed below:

18

Ron Tanenbaum, CVS, Value Management Strategies,


Inc. - Team Leader
Volume 32, Number 2, Summer 2009

Ashley Carson, AVS, Value Management Strategies, Inc.


- Team Assistant
Solomon Moore, Metropolitan Water District - Engineering Services Section
Tony Perez, Metropolitan Water District - Engineering
Services Section
Brian Folsom, Metropolitan Water District - Engineers
Cost Estimating
Eric Anderson, Metropolitan Water District - Alternative
Project Delivery
Ron Brown, Metropolitan Water District - Construction
Change Orders
Peter Tunniclie, Camp Dresser & McKee - Alternative
Project Delivery
Mark E. Alpert, CH2MHill - Alternative Project Delivery
Dan Wetstein, CH2MHill - Alternative Project Delivery
Gary Meyerhofer, Carollo Engineers - Alternative Project
Delivery
Mike Doyle, Camp Dresser & McKee - Engineers Cost
Estimating
Daniel Reynolds, CH2MHill - Engineers Cost Estimating
Charles George, Parsons Water & Infrastructure - Engineers Cost Estimating
Thomas Haid, Parsons Water & Infrastructure - Construction Change Orders
Kevin McEvoy, Camp Dresser & McKee - Construction
Change Orders
Larry Johnson, MWH Constructors - Construction Change
Orders

Author Information
Ron Tanenbaum is a senior value engineer with Value Management Strategies, Inc., and president of GeoVal, Inc. With over 40
years of teaching and consulting experience in geotechnical and
civil engineering, Tanenbaum is actively involved in value engineering, having participated in and/or facilitated over 100 VE
workshops of which over 40 have been for the U.S. Army Corps
of Engineers, 30 for transportation-related projects, and the balance mostly for water/wastewater systems. As a member of SAVE
International, he serves as the San Diego Chapter vice president
of membership and is a Certied Value Specialist.
Gordon Johnson is the chief engineer for the Metropolitan Water
District of Southern California. Johnson has responsibility for all
engineering activities, including management of a $4 billion capital improvement program. Johnson is aliated with the American
Water Works Association, the International Ozone Association,
and the American Water Resources Association.

V A L U EW O R L D

Can Value Methodology Enhance the


Competitiveness of the Supply Chain?
Ferenc Ndasdi, Ph.D., CVS, FSAVE

Abstract

Introduction

Essentially, the supply chain may be seen as several processes linked together. Hungarian and international experience
shows that value methodology (VM) can be used eectively
in the analysis of particular processes (technologies, services,
administrative processes, etc.). We have seen that in addition
to enhancing the eciency of certain stages of processes VM
can produce excellent results in linking these stages together.
Both Hungarian and the foreign experts have conrmed that
managers are mostly interested in the eective operation of the
part of the process they oversee and are often little concerned
about the problems how successive phases of these processes
can be linked together. The problem is further worsened by
the fact that the linkage between each vertically related unit is
controlled by traders, and the main criteria include time of delivery, price, payment method, as well as a guarantee for quality set forth in the relevant documents (standard, specication by the buyer, administrative specications, international
agreements, etc.) Experimental projects (leather and footwear
industry, furniture industryconstruction industry, bridge
building - tinning industry, etc.) all demonstrate that it would
be important to ensure more ecient cooperation between
each vertically related unit in the case of delivery contracts or
even in the case of R&D activities and the planning of investments. Commercial agreements can not prescribe any technical conditions that would facilitate eective further processing.
In our view Vertical Value Analysis/Value Engineering can
signicantly facilitate the more ecient operation of the entire
vertical structure. Each unit in the vertical structure should
recognize that under globalized circumstances the chances of
long-term survival can be greatly enhanced by successfully
connecting to a supply chain. Economic-nancial analyses
also need to be changed considerably. Companies often prematurely stop producing marketable but high cost products.
However, VM can often make these products protable. Relying on the results of several previously successful projects we
will present several alternative solutions. We did not intend to
present entire projects but we wished to call attention mostly
to some methodological solutions.

The economic crisis makes the market position of companies increasingly more critical. With markets growing narrow
competition continues to intensify. The most recent economic
research demonstrate that corporate problems should be investigated within the framework of the supply chain (SC). It
appears that in the case of a large number of products it is not
the individual products that compete but the supply chains.
We have already seen cases in Hungary when a Hungarian
product with excellent quality was not accepted by the SC,
despite the fact that its price was much lower than the market price. Most countries support innovative processes. When
creating and developing support schemes, the emergence and
explosive development of SCs must denitely be taken into
account.
The Management Institute of the College of Dunajvros
considers the research into SC as a major eld. Within the
framework of this research our goal is not only to learn more
about SC but to work out solutions that enable us to further
develop it.
Part of our research is also to explore the application of
value methodology in the eld of SC.
In what follows we will summarize the most important
information about SC. According to Mentzer et al. (2001):

V A L U EW O R L D

A supply chain is dened as a set of three or more entities (organizations or individuals) directly involved in
the upstream and downstream ows of products, services,
and/or information from a source to a customer.
Encompassed within this denition, we can identify three
degrees of supply chain complexity: a direct supply chain,
an extended supply chain, and an ultimate supply chain.
A direct supply chain consists of a company, a supplier, and a
customer involved in the upstream and/or downstream ows
of products, services, nances, and/or information (Figure 1 a,
next page). An extended supply chain includes suppliers of the
immediate supplier and customers of the immediate customer,
all involved in the upstream and/or downstream ows of products, services, nances, and/or information (Figure 1b). An ultimate supply chain includes all the organizations involved in
all the upstream and downstream ows of products, services,
Volume 32, Number 2, Summer 2009

19

Figure 1. TYPES OF CHANNEL RELATIONSHIPS


SUPPLIER

CUSTOMER

ORGANIZATION
FIGURE 1 a. DIRECT SUPPLY CHAIN

SUPPLIERS
SUPPLIER

SUPPLIER

ORGANIZATION

CUSTOMER

USTOMERS
CUSTOMER

FIGURE 1 b. EXTENDED SUPPLY CHAIN


THIRD PARTY
LOGISTICS SUPPLIER
ULTIMATE
SUPPLIER

SUPPLIER

ORGANIZATION
FINANCIAL
PROVIDER

CUSTOMER

ULTIMATE
CUSTOMER

MARKET
RESEARCH FIRM

FIGURE 1 c. ULTIMATE SUPPLY CHAIN


(Source: Mentzer, John T., DeWitt William, Keebler, James S., Min, Soonhong, Nix, Nancy W., Smith, Carlo
D., Zacharia, Zach G.: Defining Supply Chain Management. JOURNAL OF BUSINESS LOGISTICS, Vol.22,
No. 2, 2001.) Figure 1. page 5.

nances, and information from the ultimate supplier to the


ultimate customer (Figure 1 c).
The emergence of SC naturally brings about the evolution
of Supply Chain Management (SCM). SCM is characterized by the following:

Using a systemic approach to SC, controlling the


entire material process from procurement to the nal
consumer.

Synchronization of internal and external processes for the


entire SC process.

Putting consumer demand into the focus.

The emergence of SCM presents a new challenge for managers, since it is no longer enough to solve the problems of
their own company; they need to negotiate processes with the
other participants of the SC when establishing their own strategy.
According to the Hungarian and international literature,
SCM is concerned in short with the creation and development of the following processes:

20

Management processes
Information technology processes
Volume 32, Number 2, Summer 2009

Logistic processes
Alignment of technological processes

After studying the most recent literature it can be established that there has been enormous development in the eld
of logistics, warehousing and information technology partly
with a view to developing SC. We have found relative little in
the literature on connecting technologies and on the alignment of R&D activities.

The Situation Of SC in Hungary


The Hungarian industry is characterized by modern multinational companies and less developed small and mediumsized companies (SMEs) in Hungarian ownership. Joining in
the supply chains of the multinational companies is an important market opportunity for the Hungarian-owned SMEs.
99% of the Hungarian industrial companies have less than 11
employees; for them it is extremely dicult to establish good
market relationships.
We will present the position of the automotive industry in
the SC as an example. According to the research done by Krisztina Demeter et al. (2004) the automotive SC is characterized
by Figure 2 (next page) (Demeter et. al., p. 4). Level 1 integrators are mostly companies in foreign ownership, while Level

V A L U EW O R L D

Mulnaonal OEM (Original Equipment


Manufacturer, End Product Assembler)

Subsidiary

Subsidiary

Subsidiary

Integrator
1.

Integrator
1.

Integrator
1.

Level 2
Supplier

Level 2
Supplier

Level 2
Supplier

Figure 2. The automove industry pyramid


2 suppliers include several Hungarian-owned companies too.
According to the analysis of the researchers Audi established a
subsidiary in Hungary on the basis of resourceswell-qualied and cheap labor force and investment allowanceswhile
Suzuki on market and resource basisvicinity of the European market and favorable costs. High quality, delivery on time
and competitive phase price are prerequisites for joining in the
SC. All the factors listed above can be eectively improved by
means of Value Methodology; this is why connecting Value
Methodology and SC should be treated as a high-priority issue.

The Possible Connection Between Value


Methodology (VM) and SC
VM can be eectively applied in the following areas:

Within the materials, parts and products process: procurement of materials, manufacturing of basic materials,
manufacturing of parts and products as well as technology
development,
Establishing and developing logistic systems,
Establishing and developing warehousing systems,
Establishing and developing IT processes,
Et cetera.

In the past few decades we have mostly used the Direct


Supply Chain solution (Figure 1 a.) in Hungary, though we
called it Vertical Value Analysis rather than SC. The military
industry in Hungary can not be seen as signicant, so most
of our projects are market-oriented. Another special feature is
that the resources within a company were often not enough to

V A L U EW O R L D

implement a new solution and we found the solution in the


supply chain.
The establishment of SC in Hungary requires a signicant
change in the attitude of the management of companies. Market economy is still in the process of evolution in Hungary
and according to our analysis there are signicant shortfalls in
management. Managers are mostly interested in the ecient
operation of workows they are responsible for and often do
not worry about problems regarding the connection between
the previous and the subsequent stage of the workow. The
problem is further worsened by the fact that the connection
between each vertical system is controlled by traders, where
delivery time, price, method of payment and the guarantee
for quality specied in certain documents (standards, customer specications, governmental requirements, international
agreements, etc.) are the main criteria. Experimental projects
(leather and footwear industry, furniture industry, construction industry, building and maintaining bridges etc.) all demonstrate that it would be important to ensure more ecient
cooperation between each vertically related unit in the case of
delivery contracts or even in the case of R&D activities and
the planning of investments. Eective technological connections between each vertical system have high priority. Commercial agreements can not prescribe any technical conditions
that would facilitate eective further processing.
The issue of to manufacture or to procure emerges in
the case of every vertical system when planning dierent parts
or units. value methodology can facilitate a sound and professional decision.

Experiences Gained through Value


Volume 32, Number 2, Summer 2009

21

Analysis Projects

Value Analysis of Providing Protective Clothing


for Workers

The material of the work wear did not keep size (after
washing a gown 20 to 25 times it shrank by some 10 to
15 cm).

Certain components of the chemical agents used in washing accumulated in the fabric and could irritate the skin.

The company under investigation is a signicant Hungarian meat industry company with a professional history of
90 years. As a private company, it was nationalized in 1949,
just like every other private company, and became a stateowned company. In 1992 it was transformed into private limited company. Today it employs nearly 1,300 people. It was
the rst company in Hungary to receive the MSZ EN ISO
9001certicate in 1992. Its major export markets include the
USA, South Korea, Japan, Spain, Germany, Sweden, Poland
and Slovakia.
Exploration of the problem: Providing workers with protective clothing is a sensitive issue at the company. Due to the
technology used at the company the workers continuously get
dirty, so the continuous supply of good quality work clothes
is crucial. Due to the high rate of the washing cycle, the work
clothes quickly got damaged and costs were constantly increasing. The procurement department could not solve the
problem; they said they purchased the cheapest work clothes
and the problem was most likely caused by the washing technology. To increase the eciency of the supply of work clothes
we launched a value analysis project.
When preparing the project, we explored the following
potential causal factors:

Work wear
Machines and equipment
Workforce
Washing technology

By including experts we came to the following conclusions:

The machines and equipment operated properly and did


not need to be replaced.

The workforce was properly qualied and its reduction


was not feasible.

The washing technology met demands and similar technologies could not improve the quality of the supply of
work wear.

We prepared a FAST diagram (Figure 3, next page) for the


analysis of work wear.

The analysis of the FAST diagram proved that the main


factor responsible for the increased costs was cheap work
clothes.
Detailed analyses showed that the high rate of waste in the
case of currently used work wear was due to several reasons:
22

Volume 32, Number 2, Summer 2009

We requested another opportunity from the department


responsible for the supply of work wear. On the basis of washing tests we suggested two new versions to be implemented in
the supply of work wear.
The costs are shown in Table 1 (below).
Table 1. Annual Work Wear Costs in Million HUF
(1 USD = 200 HUF)
Currently
Used Work Proposed
Wear
Work Wear

Deviaon

Washing Costs

26

26

Work Wear Costs

24

10

-14

Total

50

36

-14

Note: The unit price of the proposed work wear was 25%
higher than that of the currently used work wear. The higher unit
price was oset by the signcant decrease in waste.
Mention must also be made of certain false assumptions
concerning costs. The general attitude: Lets buy the cheapest materials, parts, etc. When examining the procurement of
materials at companies we established that there is no cheap
or expensive material; it is only function cost that can determine the framework of eective resource utilization.

The Application of Value Methodology in the


Leather and Footwear Supply Chain

Due to the important role of exports in this sector, the


footwear industry was a pioneer in the application of value
methodology. It soon became obvious that the leather industry played a key role in the competitiveness of the footwear industry. The Ministry of Light Industry, and later the Ministry
of Industry, in collaboration with other administrative bodies,
supported several tender projects to spread the application of
VM. Experts and the management of certain companies received the application of VM with some reservation in a eld
that mainly produced fashion products. The vertical structure
of the leather and footwear industry is presented in Figure 4
(next page). Figure 4 presents a complex structure, which is
further complicated by practice. It is up to the management
of a footwear factory to decide what parts it produces and
what parts it purchases. With a view to increasing exibility,
footwear companies manufacturing fashion products mainly produce soft leather parts, the upper leather part and the

V A L U EW O R L D

Protects against mechanical


impact
Ensures protecon
Protects against dirt

Preserves quality
Ensures supply
of work wear

Resists stress
Keeps size and shape

Resists cleaning agents


Ensures easy cleaning
Repels dirt

Ensures repair

Ensures an aesthec
outlook

Applies a classical form

Uses decoraon
Carries an adversement
Idenes a company

Figure 3. Customer Oriented FAST Diagram


ready shoes, while they purchase any other parts and auxiliary
materials. Smaller footwear factories purchase the design and
the manufacturing documentation from external companies
(SERVICE PROVIDERS) too.
In the past few years we have worked out a VM model
for procuring raw leather, for the leather industry technology,
the footwear industry technology and for footwear product
planning. Essentially, we can say that we have a VM model for
the main processes of the footwear vertical system that can be
instantly applied and taught. The procurement model worked
out for raw ox hide has been adapted to the procurement of
sheepskin too.

V A L U EW O R L D

The Role of Value on the SC System


Most recent research calls attention to the importance of
the Value of business relationship. According to Mandjk and
Simon: The Value of business relationship is an existing concept that expresses usefulness and motivation recognized in,
or assigned to, a business relationship. Usefulness means the
economic side and motivation indicates the social side of value. Usefulness and motivation are value types that determine
the value of business relationships. Business relationship value
types both exist and change over time. The economic and social value types exist at all three levels of business relationships, that is, at the exchange, relational, and network levels of
Volume 32, Number 2, Summer 2009

23

business relationships. Utility values are the economic components of business relationship value, while motivation values
are the social components of business relationship value. The
two value types are in close, dialectic interaction with each
other. In addition to this close interaction, there are determining economic values.
If we wish to work eectively in the SC system, it seems
appropriate to get acquainted with the Value of business relationship in detail and align it with our VM activities.

Summary
SC is a new type of economic structure that changes the
former competitive situation. It is not the individual products
but rather the individual SCs that compete with one another.
Our analyses show that the eciency of the SC can be signicantly enhanced by the application of VM. One important
precondition for eective R&D activity is to ensure the acceptance of results through negotiations with the decision-makers
of the SC even before work begins. Getting acquainted with
the SC seems to be quite feasible. According to our experi-

ence the closed nature of individual SCs dier signicantly


with one SC to the other. Certain SCs, e.g. in the automotive industry, are highly structured, and the manufacturers of
parts operate in accordance with detailed specications. On
the other hand, the participants in other SCs are just loosely
connected. When investigating SCs it appears to be feasible to
study the special features of the Value of business relationship,
which can signicantly inuence decision-makers.
The experience gained through several value analysis projects show that the resources required for the proposed changes could be procured from SC participants outside the given
company. This is why it is important for each company to
learn more about the potentials and the framework oered by
the SC.

References
Adam, Eric, Value Management, Longman Professional,
Melbourne, Australia, 1993.
Adams, M. S. and Lenzer, W. F., Facets of FAST, SAVE In-

Buyer of fashion shoes

Retail stores selling footwear

IMPORT

Wholesale companies collecng footwear


(producon of domesc footwear companies + import)

EXPORT,
LEASED WORK

FOOTWEAR FACTORIES
ASSEMBLY OF FOOTWEAR
ASSEMBLY OF UPPERS MANUFACTURING OF SOLES PARTS
CUTTINGS OF LEATHERS CUTTING SOLES (INNER SOLE)
PROCUREMENT (SOFT LEATHERS, PARTS, AUXILLARY MATERIALS)
SUPPLIER COMPANIES
LEATHER FACTORIES MANUFACTURERS OF PARTS MANUFACTURERS
OF AUXILLARY MATERIALS SERVICE PROVIDERS
RAW LEATHERS

Figure 4. The vercal structure of the leather and footwear industry


24

Volume 32, Number 2, Summer 2009

V A L U EW O R L D

ternational Proceedings, 1997. 10- 18. pp.


Bolton, James D. and Nayak, Bijay K., Implementation of
Value Management with the Manufacturing Supplier
Community. 44. SAVE International Conference, 2004.
Jlius, Montreal, CANADA.

Zach G., Dening Supply Chain Management, Journal


Of Business Logistics, Vol.22, No. 2, 2001.
Miles, L.D., Techniques of Value Analysis and Engineering
Engineering,
Mc.Graw-Hill Book Company, New York, 1961, 1972.

Bytheway, Charles W., FAST Creativity & Innovation, J. Ross


Publishing, USA, 2007.

Parker, D.E., Value Engineering Theory Lecture outline and reading supplement
supplement, The Lawrence D. Miles Value Foundation,
Washington, D.C., 1975.

Clancy, D. F. and L. M. Dennnis, The Innovation and Application of the Value Based Design Charette Start Your
Project Right to Ensure a Successful Completion, SAVE
International Conference, 2004.

Sato, Yoshihiko and J. Jerry Kaufman, Value Analysis Tear


Down: A New Process for Product Development and Innovation, Industrial Press Inc. and Society of Manufacturing
Engineers, New York, USA, 2005.

Demeter, Krisztina and Gelei Andrea and Jenei Istvn, A


vllalati stratgia hatsa az elltsi lnc menedzsment eszkzeire. 43. sz. Mhelytanulmny. Budapest University
of Economic Science Public Administration. Vllalatgazdasgtan Tanszk, 2004.(Eect of a rm strategic on tools
of Supply Chain Management, Case study of Workshop
No. 43., Department of Business Economy,)

Snodgrass, T.J. and Kasi, M., Function Analysis. The Stepping


Stones to Good Value, University of Wisconsin, 1986.

Hannan, Donald, Conceptual Design and Development:


Value Management at its Best, SAVE Proceedings Conference. May 5-8., 1991. Kansas City, Missouri, pp. 197203.

Ferenc Ndasdi is a certied mechanical engineer, certied industrial engineer, and professor of the College of Dunajvros where
he delivers lectures on value management in a semester. During
the past 25 years, he participated in about 500 value methodology
projects as head of theme, head of team, member of team, expert,
thesis consultant, etc. Ndasdi has written more than 100 studies, professional articles, and books. He is the vice president of the
Society of Hungarian Value Analysts.

Hoer, Ilona and Lux, Victoria, Possible means to consensual decision making, International Value Engineering
Conference 2008, Society of Hungarian Value Analysts
SAVE International, Budapest, Hungary, April 10-11,
2008.
Hoer, Ilona and Plinks, Jnos, Value Driven Management at BVM PELEM Ltd., International Value Engineering Conference 2008, Society of Hungarian Value
Analysts SAVE International, Budapest, Hungary, April
10-11, 2008.
Ivnyi, Attila Szilrd and Ilona Hoer, Development of Business and Innovation, Budapest, Hungary, 2002.
Kaufman, J. J. and Woodhead, R., Stimulating Innovation in
Products and Services with Function Analysis and Mapping
Mapping,
Wiley Interscience, USA, 2006.
Kaufman, J.J., Value Engineering for the Practitioner, North
Caroline State University, 1990.
Lavingia, N. J., Pacesetter Project Perfomance with Value
Improving Practices, SAVE International Proceedings,
1997. 148-150. pp.
Mandjk, Tibor and Simon, Judit, An Integrated Concept on
the Value of Business Relationships. How could it be useful? 20th Annual IMP Conference, Copenhagen.
Mentzer, John T., DeWitt William, Keebler, James S., Min,
Soonhong, Nix, Nancy W., Smith, Carlo D., Zacharia,

V A L U EW O R L D

Stewart, Robert B., Fundamentals of Value Methodology, USA,


2005.

Author Information

Value World
Is Accepting Submissions
Value World is seeking scholarly papers
about the value methodology (value
analysis, value engineering, value
management). Do you have a new
application, a new innovation, a case study?
Send your paper to SAVE International at
info@value-eng.org
Your paper will be forwarded to the Value
World editor for a peer review.

Volume 32, Number 2, Summer 2009

25

Genetic Algorithm in Designing Optimal Layout for Stock Plate


Storage with a Unique Dimension of Material
T. Yuri Zagloel and Heryadi Pamungkas

Abstract
Layout design, in connection with material handling in a
storage area, aects the delivery time and cost eciency. Genetic algorithm approach is investigated to determine materials position in a chromosome, and allocate them in the most
ecient location. Genetic algorithm in layout design gives the
most ecient way to relocate the materials and seeks for the
shortest path to unload materials.
This paper aims to nd an optimum total distance for a
crane to lift materials. The optimum total distance makes the
stacking managable and the materials easily lifted. A number
of previous studies support the correlation between optimization and Value Engineering. This paper disclosed the values
obtained from Genetic Algorithm to be applied in the real
storage condition.

Keywords
genetic algorithm, optimization, layout design of storage
area, value engineering

Introduction
Layout design for facility purposes generally relates to material handling and storage area, which signicantly aected
the delivery time to customers (Tomkins, et al, 2003). The
importance of faster delivery, for instance, has been studied
in outbound logistics which includes the warehouse management, transportation, and distribution, and consequently, has
a signicant inuence on the eciency of the time-to-market
of the supply chains (Lee and Leu, 2006). Therefore, most of
material handling studies focus on time-to-deliver the materials to customers. According to Ballou (1999), the eciency of
material handling may be increased in four ways; capacity of
usage, layout, storage equipments, and transportation.
In terms of layout, material handling is related to stacking
problems (Dekker et al, 2006). The study focused on simulation studies and used algorithm as a method of calculation.
Furthermore, genetic algorithm was also used in Layout Design of Deng, Lai, and Wu (2006), without considering the
various dimensions of materials. Based on this background,
this paper deals with storage design for dierent types of ma26

Volume 32, Number 2, Summer 2009

terials, in which each material has a unique dimension and


lifting intensity.

Genetic Algorithm in Layout Design


Genetic algorithm is a way of solving problems by mimicking the natural process based on optimization of function.
Principally, Genetic algorithm works with random solutions,
in which initial random solutions are rst generated, and then
combined to create new solutions. Subsequently, all solutions
are compared to eliminate the worst ones (Hicks, 2002).
Performance of solutions in genetic algorithm is measured
by tness value, bigger tness value means more ecient solution. Mechanism of genetic algorithm starts with encoding the
information into gene, then all genes are randomly combined
to generate population (early solutions) formed by chromosomes. Genetic operations use chromosomes of a population
to seek for a better solution through crossover (combining
two solutions to generate two new solutions), and mutation
(choosing randomly one of the new solution to be ipped in
order to obtain a better tness value).
For storage layout design, some adjustments are needed;
the genes in storage layout design are the number of components, the rectangular dimension of each component, and
location (Gunasekaran, et al, 1999).
That information is then converted into genes; the number of the location is called plot (including the number of plot
required for each component), the number of component,
lifting intensity, and the distance between plots (provided
in matrix). Then, those genes form chromosomes that will
be randomly selected to generate initial solutions to create a
population. Those chromosomes are again randomly selected
to undergo the crossover. The characteristic of those chromosomes are passed down to the descendants. Those descendants
will be compared based on their tness value. In this case, tness value shows the total distance of lifting and dropping operations. The bigger tness value shows the shorter path of the
transportation process (Pai, et. al., 2006).

Modeling a Genetic Algorithm


This paper denes the implementation of genetic algorithm in a stock plate stacking. The dimension of each plate is

V A L U EW O R L D

Table 1. Type of Plate, Material Specicaons, and Liing Intensity


ID

Type of
Plate

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59

SS41
SS41
SS41
SS41
SS41
SS41
SS41
SS41
SS41
SS41
SS41
SS41
SS41
SS41
SS41
SS41
SS41
SS41
SS41
SM490
SM490
SM490
SM490
SM490
SM490
SM490
SM490
SM490
SM490
SM490
EH360
HS780
HS780
HS780
HS780
HS780
HS780
SHT 780
SHT 780
SHT 780
SHT 780
SHT 780
SHT 780
SHT 780
SHT 780
SHT 780
SHT 780
SHT 780
SHT 780
SHT 780
SHT 780
Wear Plate
Wear Plate
Wear Plate
Wear Plate
Wear Plate
Wear Plate
Wear Plate
Wear Plate

Thickness Width
(mm)
(mm)

V A L U EW O R L D

4,5
5
6
6
8
8
9
9
10
12
12
16
19
22
25

40
6
8
10
12
16
20
25
32
50
60
65
8
10
12
16
20
25
32
8
9
10
12
16
20
25
32
40
50
60
80
90
120
6
8
10
12
16
20
25
32

C
1524
1524
1524
1829
1524
1829
1524
1829
1524
1524
1829
1524
1524
1524
1524
B
A
1524
1524
1524
1524
1524
1524
1524
1524
1524
1524
1524
2500
2500
1524
1524
1524
1524
1524
1524
2500
2500
2500
2500
2500
2500
2500
2500
2500
2500
2500
2500
2500
2500
2500
2500
2500
2500
2500
2500
2500
2500

Length
(mm)
6096
6096
6096
6096
6096
6096
6096
6096
6096
6096
6096
6096
6096
6096
6096

6096
6096
6096
6096
6096
6096
6096
6096
6096
6096
6096
6000
8000
6096
6096
6096
6096
6096
6096
8000
8000
8000
8000
8000
8000
8000
8000
8000
8000
8000
8000
8000
8000
6000
6000
6000
6000
6000
6000
6000
6000

Weight/Unit
(Kg)
74,65
328,18
364,64
437,57
525,15
583,43
700,19
656,36
787,72
729,29
875,15
1050,29
1166,86
1385,65
1604,44
1823,22
653,23
840,00
2917,16
437,57
583,43
729,29
875,15
1166,86
1458,58
1823,22
2333,72
3646,44
4375,73
7653,75
1256,00
729,29
875,15
1166,86
1458,58
1823,22
2333,72
1256,00
1413,00
1570,00
1884,00
2512,00
3140,00
3925,00
5024,00
6280,00
7850,00
9420,00
12560,00
14130,00
18840,00
706,50
942,00
1177,50
1413,00
1884,00
2355,00
2943,75
3768,00

Liing Intensity/
Month
(pcs)
135
150
20
150
100
150
100
60
0
20
40
0
20
8
8
8
2
1
2
170
180
44
39
33
36
33
3
1
1
1
20
2
4
3
2
4
1
33
4
24
60
24
19
9
1
4
3
1
1
4
1
219
32
111
122
159
75
12
5

Area
(m2)
2,97
9,29
9,29
9,29
11,15
9,29
11,15
9,29
11,15
9,29
9,29
11,15
9,29
9,29
9,29
9,29
2,97
2,97
9,29
9,29
9,29
9,29
9,29
9,29
9,29
9,29
9,29
9,29
9,29
15,00
20,00
9,29
9,29
9,29
9,29
9,29
9,29
20,00
20,00
20,00
20,00
20,00
20,00
20,00
20,00
20,00
20,00
20,00
20,00
20,00
20,00
15,00
15,00
15,00
15,00
15,00
15,00
15,00
15,00

Weight/Area
(kg/m2)
25,12
35,33
39,25
47,10
47,10
62,80
62,80
70,65
70,65
78,50
94,20
94,20
125,60
149,15
172,70
196,25
219,80
282,65
314,00
47,10
62,80
78,50
94,20
125,60
157,00
196,25
251,20
392,50
471,00
510,25
62,80
78,50
94,20
125,60
157,00
196,25
251,20
62,80
70,65
78,50
94,20
125,60
157,00
196,25
251,20
314,00
392,50
471,00
628,00
706,50
942,00
47,10
62,80
78,50
94,20
125,60
157,00
196,25
251,20

Maximum
Stack
Height
(s)
(Unit)
358,28
0,38
254,78
0,59
229,30
0,09
191,08
0,79
191,08
0,52
143,31
1,05
143,31
0,70
127,39
0,47
127,39
1,00
114,65
1,00
95,54
0,42
95,54
1,00
71,66
1,00
60,34
0,13
52,11
0,15
45,86
0,17
40,95
0,05
31,84
0,03
28,66
0,07
191,08
0,89
143,31
1,26
114,65
0,38
95,54
0,41
71,66
0,46
57,32
0,63
45,86
0,72
35,83
0,08
22,93
0,04
19,11
0,05
17,64
0,06
143,31
0,14
114,65
0,02
95,54
0,04
71,66
0,04
57,32
0,03
45,86
0,09
35,83
0,03
143,31
0,23
127,39
0,03
114,65
0,21
95,54
0,63
71,66
0,33
57,32
0,33
45,86
0,20
35,83
0,03
28,66
0,14
22,93
0,13
19,11
0,05
14,33
0,07
12,74
0,31
9,55
0,10
191,08
1,15
143,31
0,22
114,65
0,97
95,54
1,28
71,66
2,22
57,32
1,31
45,86
0,26
35,83
0,14
Total Stacks

Volume 32, Number 2, Summer 2009

Stack(s)
Needed
1
1
1
1
1
2
1
1
1
1
1
1
1
1
1
1
1
1
1
1
2
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
2
1
1
2
3
2
1
1
66

27

dierent and the ling intensity is also dierent (unique) .


Based on the concept of the genetic algorithm, every plate
must have a value to decide its location in the layout. The
value is set based on their types and lifting intensity. Plates
with higher lifting intensity are closer to the centre point in
order to make the unloading faster. Every plate has dierent
number of stacks; plate with higher lifting intensity will have
more stacks.
The function for maximum height of a stack is (Ballou,
1999):
=

Hn max

Txtxwx1
Wn

. . . . . . . . . . (1)

The function for number of stacks is:


Sn

roundup

Pn
Hn max

. . . . . . . . . . (2)

The function for tness value is:


Fi

N P xD
n
n=1 n

. . . . . . . . . . (3)

In which following denotations are used:


H
T
W
t
w
l
S
P
F
D

=
=
=
=
=
=
=
=
=
=

n =
N =

Height of stack (m)


Tonnage of the ground (kg/m2)
Weight of plate (kg)
Thickness of plate (m)
Width of plate (m)
Length of plate (m)
Stack(s) needed for plate
Lifting intensity of plate (pcs)
Fitness value (m)
Distance between plate and the centre of the area
(m)
Plate variant
Total variants

Every solution in this method has level of performance


that can be measured by calculating the total distance needed
to unload all of the plates. The best solution is the solution
with smallest value.

Numerical Example
A case study for this paper is performed at a heavy equipment manufacturer in Jakarta which integrates their two
plants into one. As the consequence, they have to provide a
new storage to accommodate all materials of the two plants
in one single area. This study aims to design a new stock plate
28

Volume 32, Number 2, Summer 2009

storage area layout using the facilities provided by the company, focusing in designing the most ecient layout based on
the total distance required by crane to unload all materials.
The experimental conditions are as follows.
1) Type of plate
2) Material specication of each variant of each type of
plate.
3) Lifting intensity of each variant of each type of plate.
4) Ground tonnage = 9000 kg/m2
5) Stock plate storage area = 84 m x 25 m
The data of (1), (2), and (3) are shown in Table 1 (previous page).
Every plate has a unique dimension. In this method every
plate is interchangeable in order to nd the most ecient position. All plates are assumed to have the same dimension as the
biggest one (8 m x 2.5 m). Some of SS41 variants, however,
are too small to be converted, i.e. plate A, B, and C,, so they
are grouped into new dimension as shown in Table 2 (below).
In the proposed layout, the layout must have a room between
each plate with minimum of 0.5 m. Also, in the centre of the
area there should be a 7 m space for allowing transportation.
In this case, the company uses lorry to move the plate outside
the storage. If all the plates size is converted, the area cannot
accommodate all 66 plates with this particular dimension. As
a result, the area is temporarily expanded (84m x 26.5m). After all plates have been positioned, the area will be returned to
the normal dimension (8 m x 2.5 m).
Table 2. Group of SS41
Group

Type of
Plate

Dimension

Liing
Intensity

Thickness
(mm)

Width
(mm)

Length
(mm)

SS41

36

1219

2438

SS42

50

1219

2438

SS43

60

1219

2438

SS44

75

1219

2438

SS45

28

1219

2438

SS46

32

1219

2438

SS47

45

1219

2438

SS48

3,2

1219

2438

150

SS49

1219

2438

120

The new storage can accommodate 72 positions. Based


on the mathematical approach, all of these positions must be
lled. Therefore, there are 6 dummy plates needed in order to
fulll the requirement. Figure 1 (next page) shows the plot of

V A L U EW O R L D

the expanded area . In order to nd the t7 mm allowance


ness value for each solution, it is necessary to
a
b
c
d
e
f
g
h
measure the distance between the centre of 1
1
18
19
36
37
54
55
72
each plate and the centre of the storage area.
2
2
17
20
35
38
53
56
71
Also, in order to comply with the mathemati3
16
21
34
39
52
57
70
cal programming, it is better to present the 3
4
15
22
33
40
51
58
69
data as a matrix which is called distance ma- 4
trix. The distance matrix is shown in Figure 2 5
5
14
23
32
41
50
59
68
(right). In genetic algorithm, all information 6
6
13
24
31
42
49
60
67
are converted into genes. These genes will be 7
7
12
25
30
43
48
61
66
used as source of data in the entire process of
8
8
11
26
29
44
47
62
65
genetic algorithm.
9
10
27
28
45
46
63
64
In general, there are two major parts in 9
Figure 1. Plot of the storage
this study. Each part uses the same method
and steps. The rst part is positioning each
a
b
c
d
e
f
g
h
type of plate in the plot, and the second part
is positioning all variants in the plot. The 1
45,0
36,5
28,0
19,5
19,5
28,0
36,5
45,0
process is divided into two parts, because the 2
42,0
33,5
25,0
16,5
16,5
25,0
33,5
42,0
plates need to be grouped based on their type.
3
39,0
30,5
22,0
13,5
13,5
22,0
30,5
39,0
For the rst part, the required data is dierent
36,0
27,5
19,0
10,5
10,5
19,0
27,5
36,0
from the data of Table 1 (above right). The 4
33,0
24,5
16,0
7,5
7,5
16,0
24,5
33,0
lifting intensity used is the average value of 5
36,0
27,5
19,0
10,5
10,5
19,0
27,5
36,0
lifting intensity of all variants of a particular 6
type and the stacks needed is the total stacks 7
39,0
30,5
22,0
13,5
13,5
22,0
30,5
39,0
needed for all variants of the particular type. 8
42,0
33,5
25,0
16,5
16,5
25,0
33,5
42,0
The data is shown in Table 3 (below).
9
45,0
36,5
28,0
19,5
19,5
28,0
36,5
45,0
The dummy plates are not set for 1 type
Figure 2. The distanct matrix (meter)
with 6 variants, but for 6 types with 1 variant, since each plate will be grouped based on
the end of the process, that it is considered better if they are
their type. However, the dummy plates will be eliminated at
positioned separately.

Data is calculated using the distance matrix. And as the


rst, step population is set with 10 initial solutions as follows:

Table 3. Data for Genec Algorithm Part One


No

Type

Average Liing
Intensity

Stack(s)

Inial soluon 1

: 1-2-3-4-5-6-7-8-9-10-11-12

Inial soluon 2

: 9-6-12-4-5-1-3-11-10-2-7-8

Inial soluon 3

: 12-5-8-1-2-3-11-4-10-9-6-7

SS41

46,64

20

SM490

49,18

12

EH360

20,00

Inial soluon 4

: 5-7-12-11-6-8-1-10-9-2-4-3

HS780

2,67

Inial soluon 5

: 1-7-2-3-12-10-4-11-6-5-9-8

SHT780

13,43

14

Inial soluon 6

: 8-2-4-6-7-1-5-10-9-3-11-12

Wear Plate

91,88

13

Inial soluon 7

: 4-9-12-2-11-10-7-5-8-1-3-6

Dummy

Dummy

Inial soluon 8

: 9-6-4-2-3-7-8-5-1-12-10-11

Dummy

Inial soluon 9

: 10-6-5-7-8-1-3-12-11-2-9-4

10

Dummy

Inial soluon 10

: 11-3-6-4-8-9-12-5-2-7-1-10

11

Dummy

12

Dummy

Total

72

V A L U EW O R L D

For each initial solution in a population, some values need


to be determined: the chromosome, location of chromosome,
route, stacks, and tness value (the traveled distance). The loVolume 32, Number 2, Summer 2009

29

cation of chromosome of each type is taken


from the centre plate. If there are 7 plates
Chromosome:
in a particular type, then the location of
chromosome for the type is at the 4th plate, Locaon of
chromosome:
but if there are 10 plates, then the location
of chromosome for the type is at the 5th Route:
plate, The unloading route always returns Stacks:
to the centre (0) after lifting a plate. . And Fitness value:
the tness value is calculated by multiplying the distance of plates in the location of
chromosome of each type and the average
lifting intensity. The results are shown below.
The rst initial solution is shown in Table 4 (above right).
This process continues until the tenth initial solution.
The next step is crossover. Two initial solutions are randomly chosen to be the parents in order to generate two new
children solutions. Solutions 3 and 7 are randomly chosen as
the parents.
P3

12-|5-8|-1-2-3-11-4-10-9-6-7

P7

4-|9-12|-2-11-10-7-5-8-1-3-6

C1

7-|5-8|-1-3-6-4-9-12-2-11-10

C2

7-|9-12|-5-8-1-2-3-11-4-10-6

There is only one child that will undergo the mutation


step. Therefore, the two children solutions are further compared for their tness value, and a child with better tness
value will be used.
The mutation process is a process to reassemble the chromosome, unfortunately the result sometimes gives smaller tness value. Thus, in every step of the mutation tness value
will be evaluated. By the end of the process, solution with the
best tness value will be used. Solution with the best tness
value will be added to the population and compared with other tness values. Solution with the worst tness value, either
the new one or the initial solution, is eliminated. This whole
process is repeated to nd the best solution. The iteration limit
is set at the beginning; for this study 20 iterations are used.
After 20 iterations, the rst part of designing new layout is
nished. The result is shown in Figure 3.
The second part will be run ve times, each type of plate
will run the genetic algorithm, except for EH360 which only
has 1 variant. When the second part is nished, location of every single plate in the plot can be established. The nal process
is to convert plates and the storage area into the real dimension. Finally, the nal layout design of the stock plate storage
is set.
The result for the second part is shown in Figure 4. However, it is still a plot needs to be converted. Figure 5 shows the
nal layout of the storage area.
Here in below, Figure 3, Figure 4, and Figure 5 are shown
30

Volume 32, Number 2, Summer 2009

Table 4. Inial Soluon


1-2-3-4-5-6-7-8-9-10-11-12
10-26-33-36-44-60-67-68-69-70-71-72
0-1-0-2-0-3-0-4-0-5-0-6-0-7-0-8-0-9-0-10-0-11-0-12-0
0-10-0-26-0-33-0-36-0-44-0-60-0-67-0-68-0-69-0-70-0-71-0-72-0
36.5+36.5+25+25+10.5+10.5+19.5+19.5+16.5+16.5+27.5+
27.5+36+36+33+33+36+36+39+39+42+42+45+45 = 733

respectively. (See Figures 3, 4, and 5 next page.)


All mathematical calculation in this research is completed
by mathematical programming software (MathLab), and after 20 iterations with 5 times running the program, which is
equivalent to 100 iterations, the nal layout (Figure 5) is set.
Every plate is positioned in groups based on their type. Wear
plate, SHT 780, and HS 780, which have bigger dimension,
are positioned on the right side and the remaining plates are
positioned on the left side. Distance between each plate is 0.5
m (minimum). In some spots the distance is unavoidably wider, because of the dimension dierence . The shortest distance
to unload all materials is 78655.9 m.

Conclusions
Genetic algorithm in this study is used to compare solutions for optimal layout, and the best solution will be used,
and the worst one will be eliminated. To optimize the process,
crossovers and mutations are applied to seek for a better solution with better tness value. The correlation between optimization and Value Engineering is shown at the values obtained
from Genetic Algorithm to be applied in the real storage condition.
The focus is designing new optimal layout with the shortest distance to unload all the plates inside the storage. The layout is successfully designed with the unloading distance equals
to 78655.9 m. This result is the output of Genetic Algorithm
optimization model, which has been adjusted to the real condition in the plant. This model optimizes the stock plate storage area having various types of plates, and every single plate
has a unique dimension and lifting intensity.

Acknowledgement
This research was conducted with the support of PT
United Tractors Pandu Engineering, Cikarang, Indonesia. The
authors thank to all the United Tractors sta for the support
during this study.

V A L U EW O R L D

Dummy

SS41

SS41

WearPlate

WearPlate

SHT780

SHT780

Dummy

SM490

SS41

SS41

WearPlate

WearPlate

SHT780

SHT780

Dummy

SM490

SS41

SS41

SS41

WearPlate

SHT780

SHT780

HS780

SM490

SS41

SS41

SS41

WearPlate

SHT780

SHT780

HS780

SM490

EH360

SS41

SS41

WearPlate

SHT780

SHT780

HS780

SM490

SM490

SS41

SS41

WearPlate

SHT780

SHT780

HS780

SM490

SM490

SS41

SS41

WearPlate

SHT780

SHT780

HS780

SM490

SM490

SS41

SS41

WearPlate

WearPlate

Dummy

HS780

SM490

SM490

SS41

SS41

WearPlate

WearPlate

Dummy

Dummy

ID - 47

ID - 51

Dummy

fID - 39

g ID - 49

Figure 3. Result of the rst part

Dummy

ID - 12

ID - 17

ID - 28

a ID - 9

ID17
ID - 1
26 Dummy
ID - 18 ID12 ID - 10

ID ID58
-6

ID54
ID - 56

ID47
ID - 43 ID51
ID - 45DummyID - 35

ID - 2
23 ID28
ID - 16 ID9 ID - 8ID14

ID ID58
-6

ID56
ID - 56

ID39
ID - 40 ID49
ID - 46DummyID - 34

ID - 3
20 ID26
ID - 31 ID18 ID - 5ID10

ID -ID6
2

ID56
ID - 52

ID43
ID - 41 ID45
ID - 44 ID35 ID - 33

ID - 4
25 ID23
ID - 21 ID16 ID - 11ID8

ID -ID6
4

ID56
ID - 52

ID40
ID - 38 ID46
ID - 50 ID34 ID - 36

ID - 5
27 ID20
ID - 21 ID31 ID - 13ID5

ID -ID2
1

ID52
ID - 55

ID41
ID - 42 ID44
ID - 48 ID33 ID - 32

ID11
ID - 6
29 ID25
ID - 22 ID21 ID - 15

ID -ID4
7

ID52
ID - 55

ID38
ID - 57 ID50
DummyID36 HS780

ID - 7
30 ID27
ID - 24 ID21 SS41ID13

ID1
SS41

8
9

d 3 : Result of the first parte


Figure
ID - 58
ID - 54

ID - 14 c

ID - 58d

IDe- 56

Dummy

ID55
ID42
WearPlate
WearPlate ID48
DummyID32Dummy

ID29

ID22

ID15
ID7
ID55
Figure 4.
Result of the second
part

ID57

Dummy

ID37

ID30

ID24

ID19

ID59

Dummy

Dummy

ID3

ID53

Figure 4 : Result of the second part

Figure 5. Final layout of the storage area


Figure 5 : Final layout of the storage area

Volume 32, Number 2, Summer 2009


V A L U All
EWmathematical
O R L D calculation in this research is completed by mathematical
programming software (MathLab), and after

31

References
Ballou, Ronald H., (1999), Business Logistics Management
Management,
th
Prentice-Hall International, 4 Edition, New Jersey.
Dekker Rommert, Voogd Patrick, and Eelco van Asperen
(2006), Advanced Method for Container Stacking,
Computers and Operations Research Journal
Journal.
Gunasekaran, A, H.B. Marri, dan F. Menci (1999), Improving the Eectiveness of Warehousing Operations: a Case
Study, Industrial and Management Data System, MCB
University Press.
Hicks, Christian (2002), A Genetic Algorithm Tool for Designing Manufacturing Facilities in The Capital Goods
Industry, Faculty of Engineering, Newcastle University,
United Kingdom.
Kim, C.W. and Lee, J.S., (2005), Optimal Design of Laminated Composite Plates for Maximum Buckling Load
Using Genetic Algorithm, Computers and Operations
Research Journal.
Pai P.-F., Deng S., Lai C.-C, and Wu P.-S (2006), Genetic
Algorithm In Simulating Optimal Stacking Sequence of
A Composite Laminate Plate With Constant Thickness,
Computers and Operations Research Journal
Journal.
Lee, L.J.H and Leu, J.D (2006), Outbound Logistics Optimization of Manufacturing Enterprise based on SCOR
Model, Value World
World, Journal of the Society of American
Value Engineers (SAVE) International, vol.29, number 2,
Summer Edition.

Ahoy, Mateys!
Dont miss these deadlines for the SAVE
International 2010 Annual Conference!
September 1: Calls for pre-conference
workshops and presentation abstracts
begin; conference logo contest begins.
October 31: Conference logo contest
ends.
November 30: Calls for pre-conference
workshops and presentation abstracts
end.
January 31: Workshop and presentation
selections announced.
April 1: Early registration begins.
April 30:
30 Early registration ends.

Tomkins, James, John White, Yavuz Bozer and J.Tanchoco,


(2003), Facility Planning
Planning, John Wiley and Sons, 3rd Ed,
USA.

May 1: Full registration begins. Early


registration discount ends.

About the Authors

May 7: Due date for sponsorships and


program advertisements.

T. Yuri Zagloel, Ph.D. currently serves as head of the Industrial


Engineering Department and Heryadi Pamungkas is a member
of the Industrial Engineering Department at the University of
Indonesia in Depok, Indonesia.

May 28:
2 Online registration closes. Walkin registrations will be accepted.

SAVE International
2010 Annual Conference
June 6 - 10, 2010
J
For more information
www.value-eng.org

32

Volume 32, Number 2, Summer 2009

V A L U EW O R L D

Value World Editorial Policy


Value World is published by SAVE International and is distributed
internationally. Value World welcomes articles on value engineering
and related disciplines. Reprints or abstracts from other journals
and periodicals are acceptable, provided that prior permission is
obtained from the copyright holder(s). Value Worlds policy is to
provide a medium for contributors to express themselves professionally on advanced in the state of the art. The views expressed in
Value World are neither approved nor disapproved by SAVE International.

SAVE International Board of


Directors
President: David C. Wilson, P.Eng., CVS
Executive Vice President: Craig L. Squires, CVS
Vice President-Finance & Administration: J. Jerey Plant, AVS
Vice President-Global Aairs: James D. Bolton, PE, CVS, PVM
Vice President-Construction: Randy Barber, PE, CVS
Vice President-Manufacturing: Mahesh Natarajan, AVS
Vice President-Government: Brandy J. Alexander, AVS
Vice-President Education: Don H. Staord, PE, CVS-Life
Vice President- Membership: Daniel P. Fanning, PE, AVS
Vice President-Conferences: Richard L. Johnson, PE, DEE, CVS
Vice President-Communications: John E. Sloggy, CVS
Immediate Past President: R. Terry Hays, CVS-Life, FSAVE

SUBSCRIPTION
Value World is published electronically and may be downloaded
from the SAVE International website at no charge. To subscribe
to Value World, simply send a subscription request to info@valueeng.org.
SAVE International
136 South Keowee Street
Dayton, OH 45402 USA
T (937) 224-7283

F (937) 222-5794

info@value-eng.org

www. value-eng.org

You might also like