Download as pdf or txt
Download as pdf or txt
You are on page 1of 23

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/266146310

A flow-chart-based methodology for process improvement

Conference Paper · October 1993


DOI: 10.13140/RG.2.1.2559.4327

CITATIONS READS
12 11,834

2 authors, including:

Brian Prasad
California Institute of Technology
281 PUBLICATIONS 3,335 CITATIONS

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Concurrent Engineering Fundamentals: Volume 1 Chapters View project

Book reviews View project

All content following this page was uploaded by Brian Prasad on 20 October 2014.

The user has requested enhancement of the downloaded file.


A FLOWCHART-BASED METHODOLOGY
FOR
PROCESS IMPROVEMENT

Biren Prasad.
'
EDS, Die Ma_nagenent Gioup (DMG),
Mid-SizE Car Divisioir,
and

Neal Strand.
GM,' Die Ma4ager_nent Group (DMG),
Mid-Siz-e Car Divisioh,

ABSTRACT

The paper describes a new methodology for capturing the process information (both "activities" and
"connectivity). lt is based on capturing the flow information into designated graphics icons and connec-
tors, and as such this methodology is named here as "lnformation Flow-based' (IFLOW) methodology.
IFLOW distinguishes each "activity" by functional types and each "connectivity" by data types that link the
process flow information. Thus, IFLOW has multiple representations of "activity" and "connectivity" types.

The IFLOW methodology has been implemented on a window-based PC environment. The modeling of
activity types and connectivity types is supported through a flow-charting program. ABC Flowcharter
program has been employed forthis purpose [1]. However, the methodologycan be implemented in any
suitable "flow-chart" programs. Eight shape-icons are used for "activity" modeling and three data types are
used for "connectivity" modeling. They together constitute the flow of information through the entire
process domain under improvement. Examples of process re-engineering, and snap-shots of "activity"
and "connectivity" are used to illustrate the significance of this methodology.

The methodology is further integrated with a "spread-sheet" to perform analyses on the activity and data
flows. The latter is useful for identifying bottlenecks and disconnects in the "as-is" process. Many criteria
such as timing, cost, manual, human factors, business and operation practices (issues), can be combined
to obtain an effective area of improvements. With appropriate weighting factors on the criteria, a global
activity ranking can be obtained.
AN INFORMATION FLOWCHART-BASED METHODOLOGY
FOR
PROCESS IMPROVEMENT

INTRODUCTION

For many companies, to date, product development is characterized by long lead times, a large number
of engineering changes, manufacturing complications, and, ultimately, excessive costs to satisfy the
customer requirements. "Process improvement" is often perceived as a functional service to be called
upon periodically for incremental improvements in manufacturing process, product quality, and for reduc-
ing product lead times and costs. However, the perception is clearly different in companies following lean
production principles where process improvement is seen as a pervasive set of process re-engineering
activities that form the life blood of the company's regenerating profit potential,

Many progressive companies are interested in maintaining a competitive edge in the world market and in
producing high quality products. They would like to do this at a lower net cost of production than their
competitors. To accomplish such lean production goals, today many organizations are following princi-
ples such as synchronous manufacturing [1-15]. The walls between various distinct groups and depart-
ments that existed not very long ago are crumbling. Today it is more important to get inputs from all facets
of organization since no single group is expected to know everything. An organization has to look at every
detail of how to run his business and determine if he can improve it some way. Everyone in the organiza-
tion not only knows what activity he or she is involved in, the rest of the team should know it as well, in
case if data is to be shared, they can work mutually. Knowing what information is required and who needs
it facilitates smooth communication and eliminates possible bottlenecks. By knowing what we do today
and how we do it, we will be in a better situation to improve our operations, if opportunities arise.

Process improvement is a concept often used to accomplish many lean production goals. Process
improvement is a concept of product and process re-engineering over its life cycle. Here, a team collec-
tively comes up with a process that takes into considerations needs of all individuals and groups and
above all the needs of the company as a whole. Re-engineering implies that (a) an organization is in
touch with new technological advances; (b) the latest advances in all related fields such as engineering,
computers and systems are regularly reviewed for improvements in the product life cycle; and (c) an
enterprise frequently employs them, whenever appropriate, to improve an existing product. ln actual prac-
tice however, these are difficult goals.

One of the major roadblocks in implementing "process re-engineering" is that there is no formal method-
ology of visually monitoring and applying this concept on a regular basis. During process design, teams
need to determine at a conceptual level how the product and the related process will function when a
particular technology or process is changed. There is a need to capture schematically the function and
behavior of the system without any physical model or prototype so that the relationship can be studied at
a conceptual level [4]. One way is to capture pictorially (graphically) representations of both work flow
and data flow on the same chart. Such conceptual model is often referred to as "information flow-chart."

54
ln addition, the methodology used to create the information flow-chan must be capable of

(a) providing "a simple mechanism" towards easy understanding of the current works, methods and
practices,
(b) encouraging an open forum of discussion among inside and outside groups to promote a free
flow of information and activities.
(c) On a stand-alone basis, the methodology must provide visual means to improve the "current"
process by

o seeking inputs from the people who were not involved in the interview process on the first
place when the information was initially captured (peer inputs).
o Seeking inputs from customers and plants' personnel, and getting their feedbacks on the
effectiveness of the process being used.

Following the principles of "Synchronous Manufacturing," and "Japanese Product Development Team
(PDT) approach" this "as-is" process is likely to become both the catalyst as well as the means for the
evolution of a more "efficient process" [6-10]. Thus, the methodology provides natural means of accom-
plishing continuous improvements without interfering with the day{o-day business.

I N FORMATION FLOW.CHARTING

Process improvement is evolutionary, requiring a systematic approach - capturing the "as-is" process in
enough detail with relevant activities and their connectivity. The capture and categorization of "activities"
within the process flow are important to determine their relative importance and values. Similarly, types of
connectivity are useful to serue as an aid in identifying any information flow's bottleneck. Not obvious to
persons actually performing the work, bottlenecks do exit in most processes. They can cause delay in
information movement, deterioration in the data accuracy, or problems with communications. A formal
mechanism is required to identify these specific types of bottlenecks from the current process and provide
solutions to overcome them.

FORMS OF REPRESENTATIONS

There are many ways the information about a process can be captured or represented. Some schema
provides better flexibility than others. lntegrated Definitions, IDEF is an example of such tools [2-3]. lnfor-
mation flow charts can be represented in four major ways (see Figure 1):

BIock Diagram: Block diagrams are most suited for representing the function of a product or a process,
i.e., a description of its purpose. lt often has an input and output forms -- very similar to how a human
being thinks in terms of cause and effects. This representation is suitable for analyzing information content
of a system, such as data flow (input, outputs), requirements (specifications, controls), constraints
(mechanisms, dependencies, logistics, environments, etc.)

Context Diagram: Conte>t diagram is an example of a block diagram, where the relationship between "a
child of" or "a paft of" is represented by encircling the child or part information within a parent activity. The
activity "is made out of" or "requires" information from another activity and is captured by a directed arrow.

55
Block Diagram

Context Diagram

Double-Q Diagram
Fish-Bone Diagram

--}----> --}-----> --)----> --}---->


--j> --j> {. ---i-{ --+!{ State B
StateA \
___>
__>/ __>/ )r. __>/
--7 --7 --7 --7

Bond Graph

Schematic Diagram
.+-Y l-1
7 I /m
45'n\
I I fGl t--.-----r
Telepho Dn

Figure 1: Different Forms of Representations

56
Fish-Bone Diagram: This is also commonly referred to as Double-a (aa) or cause-and-effect diagrams.
The Q's stand for qualitative and quantitative factors. The diagram is made out of a heavy line or a box,
slanted arrows and parallel lines. Each line represents an activity and arrow points to the issue being
addressed. There are three pafts to a fish-bone diagram. The first part is a solid line (or a combination of a
Box and a solid line) - a heavy horizontal arrow that runs straight from one state (say a tail) to other state
(say a head). There is a family of directed arrows that are placed on top of the horizontal line representing
major "hard" (quantitative) factors, or a line of functions. The "soft" (qualitative) factors run along the
bottom (say computer tools that incorporate these functions). Arrows leading off the major factors repre-
sent sub-factors. The QQ diagram provides a visual map of key factors at one place. lt often serves as a
useful tool for brainstorming since both sides of issues remain equally visible.

Bond Graph (BG): BG is most suited to capture the behavior of physical systems instead of its function.
Most behavior is often initiated by the energetic effects that take place across the multiple domains. Bond
graph captures this using a directed graph technique. The vertices denote the energetic effects of the
subsystems and the edges indicate the energy transfer between the subsystems. lt forms a good basis
for deriving a set of differential and algebraic equations for analytical or other forms of representations.
Use of such multiple-representations-forms gives both qualitative (such as simulation) and quantitative
(such as sensitivity) insight of its energetic and dynamic behavior.

Schematic Diagrams: This is also referred to as the most primitive form of physical model. lt primarily
shows a form of abstraction that tends to be near-complete with regard to domain-specific information
and very rudimentary with respect to the functional details of its components or parts. This is often useful
in determining the missing elements and improving quality of abstractions (see Figure 1.).

COMMON METHODOLOGIES

IDEF is one of the most widely used methodologies for conceptual modeling. Air Force's lntegrated Com-
puter Aided Manufacturing Program (ICAM) initially developed the methodology for use in its definition
program [2-3]. There are two versions of IDEF methodology, namely IDEF0 and lDEF1X. The configura-
tion parts of IDEFO and IDEF1X are shown in Figures 2 (a) and 2(b), respectively. IDEF has been found
most successful in capturing the business process. However, its usage as a tool to bring about the
needed continuous improvement in engineering and manufacturing process management has been very
limited.

IDEFO is mostly used for "functional representations" of activities during conceptual modeling. IDEFO
functional model is composed of ICOM's (lnputs, Controls, Outputs and Mechanisms), arrows and boxes.
Each activity or a function is conceptually represented by a rectangular box, with three inward arrows and
one outward arrow. Each activity can be decomposed into many levels, each following the same conven-
tions. An entire IDEFO model is a hierarchical representation of a process composed of activities or func-
tions through levels. Lower-levels are the activities making up an upper level function. A solid line arrow
indicates the flow of ICOM's at each level. The arrows do not cross between levels. The interaction of
activities is managed through the boxes representing the activities. Because of the uniformity and limited
interaction facilities, IDEFO has been widely used for business modeling. lts application for modeling of
engineering and manufacturing process encounters difficulties, since it does not allow iterations between
the levels.

IDEFlX is most suited for "relationship representations" between the activities. ln the lDEFlX, a piece of
information is represented as ECARs; an entity (or objects), cardinality, an attribute to an entity, and a
relationship. ln place of attribute, some type of assertions can also be used. Figure 2b shows a common
IDEF1X syntax and conventions where parent child relationship is shown using ECARs. Entities or objects
are represented by rectangles and arrows. IDEF1X follows a top-down structure, where top entities are
composed of bottom-entities. Some entities can be represented with dashed rectangular outline to show

57
CONTROLS
(factors that consEain the activity)

OUTPUTS
(results of the activity)
(information, materials,
or other that is
changed within
the actMty)

MECI{ANISMS
@eople, tools, equipment
that perform or support
the activity)

Figure 2a: IDEF0 Methodolog5r Diagram

Parent Entity

contains (this verb describes


A the relationship
(solid lines denotes an
of the parent to the child)
identi$ing relationship)

Dependent Entity Name

attribute

attribute Child Entitv

Figure 2b: IDEFIX Methodologr Diagram

58
incompleteness, an open relationship or to clarify a relationship.

MODELING CONVENTIONS

Many ways of conceptual modeling including IDEF methodology have been described in the earlier sec-
tions. IDEF has been found most successful in capturing the business process [2-3]. lts use for capturing
the engineering/manufacturing process is often hampered by the following difficulties:

(a) IDEF has no formal mechanisms to indicate iterations among a group of activities that are
present.
(b) IDEF has no inherent mechanism to represent various forms of data communication, such as
manual, oral orverbal (voice-based), paper-based, computer-based, orthrough a physical-aid
based. IDEF has a single representation form for an "activity," a rectangular block type and a
single representation form for "connectivity"- a solid!nelype.

IFLO\IV CONVENTIONS

ln this paper, we have developed a methodology, which is based on a convention similar to IDEF and
overcomes the above limitations (see Figure 3). The methodology captures the variations in product and
process information (both "activities" and "connectivity"), besides their functional characteristics. The
functional variations of activities are captured through designated graphics icons and connectors, (see
Figure 3) and hence this methodology is named here as "lnformation Flow-based" (IFLOW) methodology.

ICONS & BLOCKS

lcons and blocks (see Figure 3) are used to reflect the various types of activity and data connectivity.
During the modeling of "as-is" process, the corresponding functional types were characterized as:

lnformation/Activity Types:

Computer (Math-based) Activity


Look-up Sheets or Drawings (paper form)
Format or Transfer (Up-load or Download)
Logic Test (Decision making)
Data Storage or Retrieval (local or global)
Comments or Remarks
Vendor-supported (activity performed by an outside source)
lnterfaces (sources or origins of information)

Data/Connectivity Types:

Dotted Line --> Manual orverbal Data


Solid Line ---> Computer generated Data
Chain Line ---> Physicalaid-based Data

Eight shape-icons and three line-icons are used to represent the various types of activity and flow situa-
-
tion, respectively. Although each graphic icon is designed to represent a specific work function, activity or
data flow, they can be used in combination to get deeper insights into a particular product or process
situation.

59
LEG EN D S:

Activily: C onnectivilT,:

Shapslcons Description Line-Icons Descrintion

Computer (Math-based Computer-based


---------------_
Activity) Data

G
- Look-up Sheets or
Drawings (paper)
Format or Transfer
..................>
Paper or Verbal
Data

Prototype or a
VA (Up-load or
Download)
- Physical Part

L-J Data Storige or Retrieval


(local or global)

O
-
Logic Test (decision)

O
rx
Comments or
Remarks

Vendor-supported

Interfaces

Figure 3: ICONS & BLOCKS

60
ENTERPRISE MODELS:

Most firms review their operations at least once a year. They look at a logical view of how they conduct
business (the current practices), or how they would like to do the business in the future. They view the
enterprise conceptually through some paper representation of its operations. This is often called enter-
prise modeling. lt is a set of models that provide a precise and logical description of the types of activity
and flow of information in an enterprise. These models capture the flow of information through an organi-
zation unit (say through a Strategic Business Unit (SBU)) that is responsible for the product manufactur-
ing. Most models of such types reflect the information needs of both the manual and computerized
aspects of the manufacturing operation.

There are two types of enterprise models that are often invoked. They are "as-is" and "to-be" models. An
objective of "as-is" model is to identify the current process and flow of information including timings'
aspects associated with designing a product. The idea of a "to-be" process is to def ine an improved
process (improved in relation to "as-is") - detailing how the SBU should conduct their operations in near
future, say three years from now. Coming up with a "to-be" model from the "as-is" model is a challenging
task. lt is important to identify the parts and processes, and the human teams that are essential. They form
the "bricks and mortar" of a product manufacturing cycle. This may require pooling all relevant core and
support teams including suppliers into a multi-disciplinary coherent work force.

,AS.IS" MODEL

"As-is" model defines the steps in the current process. lt involves identifying types of activities performed,
and the flow of data across the activities that may, in any way, influence the parts and processes. The
following are the five steps that lead to defining an "as-is" process:

Step 1: A knowledge acquisition team must be empowered to interview and identify the current process.

o The members of the knowledge acquisition team interviewing the groups should be out-
siders (preferably not currently part of the group being interviewed). This is because it is
easy to overlook many of the simple steps by those, who are close to the job. People on
the job including managers and supervisors should be preferably a part of a Validation
Team not such an Acquisition Team.
o The sponsoring organization should inform the appropriate managers and supervisors of
the real purpose of conducting the "as-is" modeling. They must understand that the
purpose is not to identify who is doing what rather to find out the process in place - what
are the activities and how it is being performed. This determination is important to elimi-
nate the fear amongst employees. Also to emphasize that it is not a part of performance
appraisal or anyway related to it but a "documentation" of how the business is done and
the procedure in place of achieving that.

Step 2: ldentify the information or activities performed and categorize them by types of activities. Some
of the activities are bidirectional, such as Data translation or lnterfaces (see Figure 3).
The mutually dependent functions commonly performed in "as-is" modeling are:

o Manual or manual-aided: This designation is used when someone performs an activity


manually, or a step towards an activity is performed by a person, for the purpose of
producing intermediate outputs, drawings, soft-prototypes, physical prototypes, etc.,
o Computer-based or computer-aided is used when a computer program or a software is
used to do a required step or to support an activity.
o Look-up Sheets or Drawings: When modes of communications other than electronics,
such as paper, orverbal are used.

61
o Format or Data Transfer: This designation is used when converting an electronic data
from one format to other such as IGES, PDES, one CAD system to other, CAD to paper
drawing or vice versa.
o lnterfaces: lnterfaces are the groups other than primary group from which information is
sought to complete an activity or to whom the information is being sent.
o Decision: This designation is used to identify conditional logic, when a process spans into
multiple or parallel activities.
o Vendor-supported function: This icon is used for services or activities, which are not
performed in house. For example, services that are contracted. Suppliers and vendors in
many companies perform a number of supporting roles.
o Comments and Remarks: Often this icon is employed to identify checkpoints or mile-
stones in the process. lt has little relevance to the flow of information and is meant to
explain circumstances or group involvements.

Step 3: ldentify the types of data that flow through these activities. The types are identified by icons (see
Figure 3).

o Computer-based Data: This is represented through a solid line and usually connects
functions or activities which are computer-based.
o Paper or verbal data: The information transfer between the two adjoining activities is
carried via paper or through verbal communication.
o Prototype or Physical Data:This designation is used when data that is passed to an
adjoining function ls not computer based, or in a form of paper or verbal. lt is employed
when data takes the form of a prototype, which aids in the realization of the desired
product.

Step 4: Map these functions and data steps into flow charts. ABC flow-charter program is used in this
paper, though any suitable program can be employed in its place.

Step 5: Create back-up sheets: This contains additional information, such as statement of work, who is
doing it, how long it takes, and additional inputs and outputs requirements. The informa-
tion is captured in a standard form developed for this purpose. lnformation is entered in
designated fields so that later, if the needs arise, they can be processed using some type
of spreadsheet or database programs. An example of a sample back-up sheet form is
shown in figure 4.

TO-BE MODEL

The "as-is" model provides a good basis to study the current process and identify the redundant functions
that are either repetitive or present little or no value. The idea of a "to-be" model is to reorganize and
streamline the "as-is" activities or even combine or eliminate the tasks so as to control the waste. ln order
to lead a company to the 21 century, one must increase the throughput and improve productivity. This
means "doing it right the first time," and eliminating time-consuming and costly engineering change
requests. There are two ways of achieving a "to-be" model: Top-Down Approach and Bottom-up Ap-
proach.

62
I

DMG DATA
Source

Data Flow
Task No.

Statement
of Work

Task
Responsibilit

What
Triggers the

Task
No. of People

Deliverables

Source of
lnput6

(sAA/): {H/VV) ts (SA|/):


Computer
System
Requirement

Disconnects

Figure 4: A Sampre Back-up sheet for


capturing other rnformation
Top-Down Approach

ln this approach the management of the company puts together a team of experts from various disciplines
and empowers them for coming up with a vision of how the future operations should be run. Usually, with
such objectives little or no constraints on the existing systems, such as cultures, cost investments,
automation, etc., are imposed.

ln the top down approach, the team analyzes the "as-is" model to study the current process and deter-
mines a new process that meets the objectives towards a company long-term vision. This requires creat-
ing another model of an enterprise called "to-be." The following is a step by step process that can be used
for coming up with a "to-be" model.

Step 1: Strategic Vision: The team members understand their groups' vision in relation to the company
overall vision. They either know or learn the direction the company is heading, how the
units will be supporting it, the lines of products they will continue supporting, what's being
refined and what's the global needs are.
SIgp 2: Global Participation: Begin to conceive a concept where there exists no barriers of disciplines or
groups' participations. A global manufacturing philosophy, where geographical limitation
is no barrier, is followed, Global means conceiving a product in one place, designed by
another group or country and manufactured in another place or plant.
Step 3: Common Obiectives: Consider a set of common objectives, such as cost and quality, customer
satisfaction, etc., that is uniform across various departments and related units. Use this as
a basis for step 4. Each activity should be analyzed from its own right and how it impacts
customers, suppliers, engineering teams, product teams, manufacturing teams, plants,
etc.
Step 4: Simplify/Reorqanize: Reorganize the activities to do one or more of the following: eliminate dis-
connects between data change over, for example, manual to computer, computer to
paper, etc. Eliminate duplications, rework, repetitions, and waste. Most design teams
"see" activities but have difficulties identifying the types of data needed to execute them.

Bottom-up Approach:

Here, the "to-be" process is evolved from the "as-is" model as opposed to being defined a priori. A process
called "synchronous manufacturing" or "value analysis/engineering" is employed to transform the "as-is"
model into a "to-be." Multi-disciplinary teams that are formed at the early concept stage are given far
greater responsibility. They meet frequently and review the process in detail for all sorts of bottlenecks or
possible disconnects as discussed earlier. Thus, many such decisions are made lower in ranks. Manage-
ment retains rights to review and to reject, but the teams shape the vision, streamlines parts and process
work flow, recommends solutions, and plans improvements. The philosophy is based on the concept that
employees are close to the problems, they deal with them everyday, they know more about them and
therefore should also be capable of solving them. The approach encourages teamwork and enforces faith
and confidence in each other. The major drawback of this bottom-up approach is that it could be very
time consuming. The rate of change could be far too slow. Often, it takes too long to reach a consensus.
ln today's rapidly changing market conditions, one may not have so much time to wait. Often, far reaching
and bold decisions are quickly required to change the failing course.

64
IFLOW METHODOLOGY

ln the following an IFLOW methodology is proposed which is based on combining both top-down and
bottom-up approaches and gluing it with value engineering. ln this methodology, first, one start with a
bottom-up proposal, then a top-down scenario followed by a bottom-up revision. The trio combination
often leads to a more balanced result. The following are the major steps of the combined process:

o Planning for "as-is" Activity Charting: Activity chaning is an important step of value engineering
for building quality into manufacturing. The initial steps involve (a) identifying who and what has
been done so far; (b) reviewing and prioritizing the work to be done; (c) putting together a policy
document outlining purpose and scope and deliverables; and (d) reviewing with the management
and getting their Approval. lt focuses on the entire manufacturing process and their interfaces,
rather than its subset. The normal procedure is to start with one key area and use this as a team
exercise.

ldentify the Function ol the product or Service: The next major task is to review the "current"
process for this area and to get all the relevant facts. For example, the team gets (a) information
about speciality products, materials, processes, vendors; (b) talks to company and industry
experts to understand the functions they are performing.

a. Begin Drawing a process llow chart: This step includes the work flow laid in minute detail identi-
fying "what is planned to be done." A structured format is used to emphasize the impact of
sources of variations on the process. The emphasis is on "procedure" source not its agent, or the
person who is running the procedure. ln addition, the tasks involved in this step are:

- Define a candidate process model or value tree: The result of this charting is the devel-
opment of specifics called "candidate" process model.
- Draw value graph - process description sheets: the team should now be able to break
down the flow into smaller chunks in a form of a process description sheet so as to
accurately describe the associated manufacturing method.
- identify sequence: lt identifies sequence, the main assignments, part flow, timings, etc.
- ldentify indirect expense and control parameters (such SQC, SPC scheme, etc., associ-
ated with Just-in-Time manufacturing).
- ldentify process parameters: During this step, process assumptions are identified,
machines and people are sketched, Critical Process Characteristics are confirmed.
- ldentify New investment: New machinery or equipment cost is also identified and esti-
mated, and noted against the appropriate entry in the Bill-of-Material.

b. Establish a value for that lunction or service: The process begins with looking at the "candi-
date" process model of an enterprise and performs functional economic analysis in support of
process redesign and new investment justification, if any. The teams model the process using
icons and blocks in addition they do the following:

- List sources of variations: The team lists all sources of variation potentially affecting each
operation.
- ldentify value-added activities: Brainstorming and team dynamics are tools often used to
identify value-added activities from non-value added activities in the "candidate" process.
- Develop fishbone diagrams: The completion of this phase is typical of most brainstorming
sessions and is equivalent to developing a fishbone diagram for each operation.
- Brainstorm to identify the new "improvement" Process. Continue elimination of the most
obvious barrier in the "as-is" process.

65
- Create and refine: Based upon the consensus, the concepts are redefined, combined or
subdivided in order to provide a clear identification of activities that add values from those
that do not. A Graphic tool for placing the values can be used to aid this process.

c. Seek Creative ldeas: The creative ideas and the graphic pictures obtained during analysis phase
is evaluated here. Here team thinks creatively, analyze costs, such as putting "$" sign on toler-
ances, applies judgment and timings. The team also looks at the outgoing dimensions or charac-
teristics affected or changed at each step in the process. Generally speaking, activities can be
categorized to fall into four definite ranks:
o Analyze the Information and make changes: this includes identifying gaps or the activities
that are impacted by gaps.
o lnitiate a control plan for each operation: Each outgoing dimension or characteristic and
its associated description are recorded. lf "gages" are available for measuring process
results, they are recorded on the control plan. Only "perform" activities imply a transfor-
mation process from inputs to outputs that adds value. Others are means to get to the
end but are generally not an essential part of the process.
o Evaluate alternatives: Brainstorming and other tools are used to provide alternate meth-
ods of doing the same work. The team looks at each activity one at a time and studies the
measurement system variability. Participants are encouraged to suggest ideas no matter
how difficult or costly it may be. The objectives are to overcome gaps by use of stand-
ards, use of speciality products or using ones own judgment. lnappropriate ideas may
lead to good ideas.

d. Redesign an improved process: Redesign phase aims at emerging into a process that are error-
free designs and embed performance monitoring with provisions for self-correcting. The
common steps are (a) mark the process flow activities that are impacted; (b) lnvite Ex-
perts to critique the charts or seek inputs; (c) Revisit and look at every detail of how we
run our business and determine if we can improve it every step of the way. During this
phase, the team overcome roadblocks, construct a characteristic matrix which displays
relationships between product requirements and operations. The matrix helps to evaluate
the importance of characteristics (value-added) while showing important upstream rela-
tionship. Based on this, an improved process is redesigned. These processes use stand-
ards, effective employee time, eliminate wastes or redundancies. lt contains a significant
portion of value-added activities in its flow ("perform" type activities). Outputs of one activ-
ity normally are made inputs for a subsequent activity in a pipeline mode. The prepara-
tions and testing time are kept at minimum. However meeting this threshold consistently
at a lower cost is often difficult. Complexity, process variation, and uncontrolled situation,
such as dealing with consequences instead of outputs are some of the elements creating
such difficulty.

Do a global gap category Ranking as a group. First, identify the system level gaps, timings,
bottleneck and disconnect. Then, identify the system's level impact on the "to-be" Process
and the enterprise's functions. Then, invite experts to critique the numbers and the gaps.

Define ldeal Process: The goal is to provide the services at the lowest possible cost without
diminishing performance or affecting quality. The possible steps are:
- Think creatively: This can be a paradigm shift in terms of how we do the business today.
Though the goal is to find the best possible solution within the limits of organizational and
cost barriers, however, it is possible to cross them if the plan provides extraordinary
savings and adds to company values.
- ldentify top level gaps/waste and Create gap categories.

66
!#,il'xlws,#I$d x:iiriit lnnnaalirdl No. o'
nr EqJN Countsl iffilcons * cons # zageg
ii:;!!!!:!::ii;:i:!:;i:i.i;:::!1:!i::::::i:::: ;:;;;i !i:
i:!i!i::::i:: ii:::::::: !i: i i:i;::::::::::i:i
r

:Dft ;P
oj Ts,
i:iiiiiiiii6l iii: 67 !!i
,, !i : Ig.-99 i i : i :::
:: :::::i :r:

Die-Layout 96 103 4

::::::::!:::i::::::t:i::1:!i::a:r::!:l
!!!:^, , ::,,,i,:
ii!rufe ueslgn i:ii:..:l :::::i::iii:i48 :i!;i:i ;iiiiiii:::ii58::
i:iii:;ii:iiiiiii i ii:ii ii ii iiii;:iii|

Total 205 228 10

Table 1: Counts of Activity and Connectivity lcons in the 'As-is"


Process Charts
- Eliminate waste and gaps: Eliminate all the waste/gaps inherent in the process.
- Define ideal situation: The team looks at all the possible alternatives and recommends the
ideal situation that ranks well as in step (e). Document the "improved" Process timings
and ratings.

g. Repeat steps (a) through (f) for each functional areas: lnvite the experts and repeat this step as
often as necessary

PROCESS EXAMPLES

At GM/Die-Management Group (DMG), Mid-size Car Division Headquarters, we applied the IFLOW
methodology to most DMG operations, namely Die engineering seruices, platform support and Die plant
operations. A number of persons having different backgrounds (body-in-white, computers, research, were
assigned to define and capture the "as-is" process. Three groups were involved. One group did the "as-is"
for Die engineering services. The second group did the DMG three plant operations: Flint Tool and Die,
Grand Rapids, and Pontiac plants. Many of them are listed in the acknowledgment section. The third
group did the platform support flow-charting. Since Die Engineering Services (DES) was the first attempt,
we will concentrate the paper in relation to the DES "as-is" work. ln Die Engineering Services, we inter-
viewed experts from three distinct die-engineering areas: Die-Processing, Die Layout and Die-Design. We
capture the "as-is" flow for each area in three separate charts. We identified all activities in each area
through eight functional types and three data types as discussed in section 3 and 4. The information (both
activity and data) was flow-charted using the ABC Flowcharter program. Table 1 lists the counts of activity
and connectivity found in each of the three units. Die-processing chart took three pages, dieJayout four
pages and die-design three pages. Thethree charts altogetherwere ten pages (8-1/2"r'11") long. When
we were all done flow-charting the three charts were merged together. lt was then blown up 300o/o. When
we displayed them all together in larger prints, it spanned 4' by 30' wall area at the Die Engineering Serv-
ices in Pontiac. To illustrate the methodology, an excerpt from a die-layout area is shown in Figure 5.
Figure 6 shows a typical portion depicted in figure 5. ln each case the information contained within the
shape-icons and line-icons has been omitted. Figure 6 shows the "as-is" process (in skeleton form) at the
start of a die-layout phase. Two types of inputs (comprising of six manual activities, 1 through 6, and one
database activity # 7l are shown to fed into a decision box #7. Later, information is shown breaking up
into two paths in Figure 6.

After we completed the DES "as-is" process flow charts, soon we recognized the needs for improvements.
A number of areas where some local improvements can be possibly made were identified. However, it was
not clear looking at the charts, which of these activities ought to be changed and why. We didn't know
what basis to use in setting the change priorities. What was lacking was a rational basis for deciding what
was important and what was not. Though, value-analysis provides one such basis, many other bases do
exist. Some indexes often used in this context are [15]:

Critical lndex: This is a measure of criticality. This index is assigned a value of 'l if an activity falls on the
critical path and 0 if not. Thus critical index takes a logical value of o or 1.

Time-index: This is a measure of how long it takes to perform an activity.


Cost-index: Some activity cannot be measured by time but requires capital investment. Cost-index is a
measure of the cost associated with performing an activity.
lssues-index: Many times, some activities are subject to a lot of management and implementation issues
not related to either time or cost. lssues-index provides such measures.
Value-index: As defined earlier, it measures the value of performing the activity.

It was difficult to decide which one of these indexes should be used as a basis of determining the impor-

68
Wffiv.@7,192
f
Prd6 RaE.k

--l Yc''r<
\./
''D
@o
.-tulPhn*h

*-&
f, e) ' a(
---fa--
<-
>1

V A -o
TFdWdDdE:

J Of
Cqrrffi
(ffid)
ld-@
s6{Pprl
ooh)
Orts.(
Ekb
m \

DIE.LAYOUT DATAFLOW
CI q oL3
cl-
:-L
_J
),'"\
1v>* J-l.s.r ,,\
C--(

ru \

o x [.

-.-ffX
Figure 5: An Excerpt From the Dielayout Area -"As-is" Process Flow

69
Wedrresday. October 7, 1992

[f Stage lV or Higher
Product Relsas€ Data

efl I

I
. --- Shell Plastor lrom

[(
G@
{
G
(_f Stage lV or Higher
Produc'l R€leas€ Dala

Stage lV or Higher
--Producl Release Dala.-__=--_--
t,E(i ENI)S:

e
I

V
'[Jpes 0f Work ]'uncllons:
J

Compul6r Furction Rsmarks Look-uP


Download (one Upload (one Way
l-_J
f.-

Databaso (
ll

commonlv Sh6ets (Pap€r)


(Math based) way Datallow Datatlow) Eleclronic
)

Figure 6: A Typical portion of "As-is" at the start of Die-layout Phase


tance of an activity. lf one is considered more impoftant than others, it might be true for one set of activi-
ties but may not for the others. We needed a schema for arriving at a combined rating from the individual
ratings. This paper describes such a schema. Table 2 shows a partial result of combining the individual
ratings. A total of 23 activities is shown ranked out of a total of 44. The principles, on which the combined
rating is obtained, are derived on the basis that each index is equally important from process improve-
ment considerations. lf this is not true, weighting factors can be assigned to the indexes to balance the
results out. The following describes the approach:

Let us assume S;; is a measure of jth index for an ith activity. Then the combined ratings can be formed as
follows:

Normalized lndex N;1 = S;1/ (1)


!;
h
where, Lj = soRrf,Ttt,l (2)

lf W; is the weighting factors associated with the indexes.

The combined Ratings can be obtained as:


'i 5
l Combined Ratings, ni =Z{ [N;;l* W; ] (3)
i *:t
The ranking of Ri in between 0 and 10 can be obtained by:
T1
Ranking, RRi = Ri * 10. /SORTz{ [Ri"*2]] (4)

It may be noted that j= 1, indicates the critical index, i=2,3,4 and 5 denote time, cost, issues and value
indexes, respectively.

ln order to apply the above computation, ABC flowcharter program was linked through DDL (dynamic
data linking of Microsoft Windows) to EXCEL - a Microsoft spreadsheet program. The activities from the
ABC flow-charter program were pipe-lined as inputs to the first two columns' fields of EXCEL program,
meaning they were not entered but automatically extracted from the ABC flow-charter shape lists. A
sample result of this analysis is presented for illustration in Table 2. Note the values entered for each of the
five indexes is fictionalize. The results do, however, illustrate how to obtain a combined rating from the five
individual indexes.

The last column in Table 2 - the ranking column outlines the relative rankings of the activities. The activities
which are closer to '10 mean that those are quite important. The activities which are closer to 0 and less
than 5 represent weak links. They represent candidates for possible modifications or eliminations.

CONCLUDING REMARKS

IFLOW methodology showed the power of "visual" techniques to process improvements. The different
types of shape-icons for the activities and line-icons for connectivity helped us to capture the current (as-
is) process very easily. lt also allowed us to communicate the information through simple iconic means
than through plain words. By depicting the activities into suitable icons, team members were able to visu-
alize easily the entire process without much needs for explanations and tutorials. The charts were easy to
follow and quite self-explanatory. lt also provided a medium of discussions for process improvements in
many areas of our productivity concerns. The implementations of the change process as a result of those
improvements presented very little or no resistance, from our employees. The members were also seen
forthcoming in suggesting new improvements in their own areas and compromising on many barriers that

71
DIE-PRON.XLS

A B C D E F G H I .I
I
2 MATH DATA SYNTHESIS MATRIX Individual Basis of Ratings & Specilications Overall Rankings
3 Activitv No. Activity Description Does it f How lon Whats'e lnfluenc Does thi Relative Severity Index
4 rype Critical lime Cost Issues Value Weishted Value
5 Index Index Index Index Index Value Ranks

6 Weiehtine Factor 2OVo 307o 20% 207o to%


7 Minimum V lue 0 0 I 0 0 0.03
8 Maximum V lue 4 I I o.29
9 Normalized alue 4.58 8.74 5.74 1.60 3.91 0.78
0 Standard Dev iation 0.49 1.23 0.00 0.27 0.47 0.07
I SI Advanced Prints (if any) 0 0 0 0.06 I
) S2 Plant Information (press information) 0 0 0 0.06 I

3 S3 Receive work order from SME for Die Processing or I o o.24 8

4 S4 Work Order and assign iob to Die-processor I 0 0. l4 4

5 S6 Start of Die Processing or Die Pre-Processing Phas 0 0 I 0 0 0.03 0


6 S7 Review Work Order and advanced prints if available 0 4 I 0 0.5 0. l8 6

7 S8 Up Load latest Product Data from CGS I I I 0 0.l4 4

l8 s9 f this is a new nart? I 0 I 0 o 0.08 2


r9 s 0 Product Released Databank (comorate) o 0 I 0 0 0.03 0
20 S Is the released data same as Dreprocessed? I 0.5 I 0 0 0. l0 2
2l S 2 Are there any maior changes since last process? I 0.25 I 0.33 0 0.13 4
'r7 S 3 lnnut from Formabili:v Platform SME 0 0.5 I 0 I 0.08 2

23 S 4 fip Part (based on user Exp.) I 2 I 0 I 0. l7 5

24 S 6 Determine Die-tvne & Number of Die Operations I 4 I 0 I o.24 8

25 s 7 Versatec tipped plots ( full size or scaled down) 0 0.25 I 0 0 0.04 0

26 s 8 Is any Product Change required? I o.25 I 0 0.5 0.10 3

27 sl9 Ch"ne9!gq!g't !lg!"d off_by SME 0 o.25 0 0 o.04 0


28 s_?0 DES User Files (M/F) 0 0 0 I o.06
29 s2l s this Acceptable by SME & Design Staffl I 0.5 o o.22
30 s22 Write Die Line-up (minus blank Die) Info. I 3 0.67 I o.29 l0
3I s23 Preliminarv Die Line-uo Review I I o 0.5 0. l3 4

Table 2: Relative Ranking of the Activities (Die-processing Example)


became vividly clear when the entire process was flow-charted.

ACKNOWLEDGMENTS

The authors wish to acknowledge R, Ayres, C. Gerrish, B. Hoye, R. Pakizer, l. Zahoor and M. Zibble for
their invaluable contributions in developing and applying the methodology to the creation of an "As-is"
model as described in this paper. ln addition, R. Brady, A. Lambertz, D. Porembiak, Ed Stepanski and
others in their departments are acknowledged for providing the details of the existing process at the Die
Engineering Services. Thanks are also due to Bill Happel, R. Pierson, M. Zuzelski and T. Zwickl for their
support and encouragement.

REFEBENCES

t1] ABC Flowcharter, Users'Guide, Version 2.2, Document Number FCU-ENG1, Micrografx lnc.,
Richardson, TX, 1993.
I2l Materials Laboratory, AFWAL, WPAFB, "lntegrated Computer-Aided Manufacturing (ICAM) Func-
tion Modeling Manual (lDEFO)', June 1981.
t3l Yuhwei, Yang, "lDEFl X Style Guide For PDES Users", Draft, August 31 , 1989.
t4l B. Prasad, J. MacDonald, D. Auxier, "Door Design Methodology: Requirements Specifications",
Report Number EDS/CPC/AVE/DOOR5-003, Jan. 1 991.
t5] Kulkarni, H.T., Prasad 8., and Emerson, J.F.," Generic Modeling Procedure for Complex Compo-
nent Design", SAE Paper 811320, Fourth lnternational Conference on Vehicle Structural Mechan-
ics, Detroit, 1981.
16] Prasad, B., Rajesh Shah, Kulathu Subramanium, RajAbraham, Dave Dalling and Jim Copeland,
"Generic Requirements for Developing a Parametric-Based Application", Report Number VAPD-
DR-003, EDS/TSD Report, August, 1987
171 Shin Taguchi, "Taguchi Methods and QFD: Hows and Whys for Management," American Supplier
Institute, A.S.l. Press, Dearborn, Ml.
t8l Bob King, "Better Designs in Half the Time - implementing QFD Quality Function Deployment in
America," 1987, GOAL/OPC, Methnen, Massachusetts.
[9] John S. Oakland, lnternational Conference on Total Quality Management (2nd: 1989: London),
Total Quality Management - Proceedings, 1989.
[10] William R. Carey, "Toolslql Today's Engineer Strategylql Achieving Engineering Excellence":
Section 1: Quality Function Deployment, SP-g13, SAE International Congress and Exposition,
Detroit, Michigan, USA, February 24,1992 - February 28,1992, SAE 920040.
[11] E. Ezop, T. Jacoby, and L. Leach, "Continuous lmprovementWith Quality Function Development,"
SME-1Q890405, SME Technical Papers, 1989, 10P.
[12] Donald E. Freeze, and Howard B. Aaron, "Customer Requirements Planning Process [CRPIl]
(Beyond OFD), SME Technical Papers, 1990, 30P, SME-MS90003.
[13] W. Edward Deming, Out of Crisis, 2nd Edition, Cambridge , MA, MIT Center for Advanced Engi-
neering Study, 1986.
[14] Taguchi, G., lntroduction l9 Quality Engineerino, Asian Productivity Organization, Tokyo, Japan,
1 986.
[15] R.J. Dika and R.L. Begley, " Concept Development Through Teamwork - Working for Quality,
Cost, Weight and lnvestment", SAE Paper # 910212,1nt Congress and Exposition, SAE, Feb. 25-
March 1,1991 ,pp.1-12,1991, Detroit, Ml.

73
QUALITY CONCEPTS'93
WORLD CLASS MANUFACTURING

Working Together for Quality


E

t!4 tlrruF
a.Zr_t
{ til -
T+

General Motors Technical Center


Warren, Michigan October 4-6, 1993
Sponsored by:
ESD The Engineering Society
-
American Society for Quality Control (ASQC) Automotive Division
-
View publication stats

You might also like