d:\scanned\advanced.

tif Page 1
 300 Techniques o f Value Analysis and Engineering

The approach to this system for intense study of functions and function
interactions lies in asking three questions and then diagramming the
results. The questions asked about each function that is performed by
the product or service are: "Why?" "How?" and "When?"
Answers to the question "Why?" lead back to what is called the higher-
level function. They answer the question "Why is this function being per-
formed anyhow?" Answers to the question "How?" lead ahead to the
specific functions that must be performed in order to accomplish this one.
All functions so identified are charted from left to right, with the
highest-level functions, which were identified by the question "Why?" on
the left and the functions identified by the question "How?" on the right.
When this has been properly done, each function in the entire series is
illuminated on the left by "Why it is done?" and on the right by "How
it is done?"
Next the question "When?" is asked of each function, and it is dia-
gramed in time sequence in relation to the other functions so that ex-
amination of the diagram vertically illuminates the matter of which
functions are to be accomplished at the same time and the sequential
relationships of all others.

How to Use FAST2

This technique is useful in determining the function interrelation in

analyzing an entire system or a major portion of a system and gives a
better understanding of the interaction of function and cost.

Function Determination Logic

This system provides us with the determination logic to determine the
basic function and the higher- and lower-level functions, as well as the
supporting systems or equipment. The use of this system requires the
construction of a FAST diagram utilizing the determination logic ques-
tions "How?", "Why?", and "When?"
The steps necessary to construct the FAST diagram are as follows:
1. Prepare a list of all the functions by assembly or system using the
verb and noun technique of identification of function.
2. Write each function on a small card. Select a card with the function
that you consider to be the basic function. Determine the position of the

2 Appreciation is expressed to the Xerox Corp., Rochester, N.Y., for permission to

use this "How to Use" material, and to Francis Xavier Wojciechowski, value engineer
for Xerox Corp., who prepared it.
 Advanced Techniques 301

next higher and lower function cards by specifically answering the fol-
lowing logic questions :
"How?": How is this function accomplished?
"Why?": Why is this function performed?
"When?": When is this function performed?
Using one of the functions selected, apply the logic questions to deter-
mine the functions to the right and left of it, as shown in Figure 17-1.

ASK W H Y ?
+UNCTION
SELECTED
FA~I HOW ? Fig. 17-1 The "Why"
"How" logic.
and

Let us look at the example of an overhead transparency projector. The

function selected was to "show diagram." By asking the logic questions
"Why do we show diagram?" we answer, "Teach students." How do we
"show diagram"? "By reflecting image." How do you "reflect image"?
"By projecting image," as shown in Figure 17-2. At the same time, the

question "Why?" must be answered. Why do we "project image"? "To

reflect image." Why "reflect image"? To "show diagram," etc. By this
method we can check the proper order of the functions in the diagram
and also determine the completeness of the diagram by assuring that all
functions are accounted for. -
In the next step, we ask, "How is the project image function per-
formed?" The reply is, "Illuminating transparency." The next question,
"How do we illuminate transparency?" gives two answers: (1) by pro-
viding a light source and ( 2 ) by supporting the transparency in a suit-
able fashion. Therefore, both "provide light" and "support transparency"
are needed to answer the question "How?" What about the question
"Why?" The answer to both questions, "Why do we support transpar-
ency?" and "Why do we provide light?" is "To illuminate transparency."
The logic questions are satisfied, and we have the blocks on the FAST
diagram shown in Figure 17-3.

TRANSPARENCY

Fig. 17-3 Basic FAST diagram.

302 Techniques of Value Analysis and Engineering

The determination logic questions are then applied to each function

and the FAST diagram continued until we arrive at a function that is an
accepted interface function for the scope of the problem. This will be
discussed in more detail under Determination of Scope. '

Critical Path Determination

When all the determination logic questions "How?" and "Why?" are
answered for each function, we have established the relation between the
functions at higher and lower levels that are required to perform the
highest-level basic function. The arrangement of these functions, as shown
by the FAST diagram, establishes the critical path. The critical path
identifies the functions that are the result of other functions to be
performed.

Supporting Function Determination

After the critical path functions have been selected and positioned on the
FAST diagram, the remaining functions from the list in step 1are placed
on the diagram by applying the question "When?"
On the accompanying complete FAST diagram for the overhead pro-
jector, the function of "cool lens" is performed when?-At the same time
as the "illuminate transparency" function is performed, and therefore it
is placed in the same vertical line. Note, also, that there is a dotted line
connecting the "cool lens" function with the "project image" function.
The dotted line indicates a secondary or supporting path. The secondary
function paths are usually the result of specific methods chosen to imple-
ment the function. In this particular case, the "cool lens" function re-
sulted from a light source, which "produces heat" (an unwanted func-
tion), and a Fresnel lens, which requires cooling.
If hardware details have been or are being designed, the parts that
perform a function can also be added to the FAST diagram and are
placed directly below the function that they perform.

Determination of Scope
In order to limit the FAST diagram to a specifically determined problem,
the scope of the problem is defined and outlined on the FAST diagram
by the scope lines (vertical broken lines). The left scope line is placed
between the basic function under consideration and the highest-order
basic function. In other words, the function on the left of the scope line
is the next higher-order function, which will not be completely satisfied
by the solution of this problem. The function on the right is the basic
 Advanced Techniques 303

HOW- I WHEN I -WHY

I I I
I
I I
I MOUNT PROVIDE 1
I EQUIPMENT CASE
I
'
I I I
I SCOPE L I N E I Case I

I ,CRITICAL

p
G
7m
STUDENT I DIAGRAM SCREEN
i M o u n t i n g and Screen
I
I
i
I
Lamp

Fig. 17-4 FAST diagram of overhead transparency projector.

function, which must be satisfied by this product or service. In the ex-

ample considered here, the overhead projector must satisfactorily "show
diagram," but it will only partially perform the "teach student" function.
The right scope line is drawn to the left of a function that is an accept-
able interface to the product or service under consideration-a suitable
input to the system.
In our example, the "provide power" interface function is the lowest-
level function we wish to consider in discussing the overhead projector.
The manufacturer of the overhead projector will expect the user to pro-
vide an ac outlet and the transparencies. The choice of the scope is
arbitrary, and it is often necessary to reevaluate the scope at the comple-
tion of the FAST diagram.
This completes the mechanics of constructing the FAST diagram. A
careful analysis of the details of the diagram will reveal interesting and
useful relations. The function to the right of a selected function tells how
the function is performed. The function to the left indicates why the
function is performed. The functions above and below show which
functions are happening at the same time. Finally, if parts are available,
the diagram shows what functions are performed by which parts. See
Figure 17-4.
In the innovation phase, we generate alternate ways to perform func-
tions. By examining those functions on the FAST diagram that are
adjacent to each other or are performed at the same time, many ways to
combine, modify, or otherwise perform these functions can be developed.
This often leads to innovations or better ways to do the required task.
304 Techniques of Value Analysis and Engineering

FAST Diagram Cost Allocation

If the function classification has been performed by the construction of
the FAST diagram, the evaluation may be performed utilizing the dia-
gram. To perform this cost allocation, the function blocks on the FAST
diagram are numbered. These functions are used as column headings
under the appropriate number of the function-cost breakdown sheet. The
individual part names and total costs are placed in the two left-hand
columns. The functions of the part are analyzed and the cost allocated
actually or proportionately to each of the functions performed by the
part. The totals at the bottom of the page provide the visibility of cost
for each function as well as the total costs for basic and secondary func-
tions. It is sometimes desirable to transfer these totals to the blocks on
the FAST diagram to give added visibility to the lateral buildup of
costs that provide the basic function. This method makes immediately
apparent where the costs are distributed and the total costs involved in
performing the functions on the critical path.

Case Study
FAST DIAGRAM USED; PARTS REDUCED FROM
NINETEEN TO FOUR

The "How?", "Why?", and "When?" were diagramed covering each of the
functions (and parts) of a timing mechanism. The developmental design,
shown in Figure 17-5, contained nineteen parts, eighteen of which are
shown in Figure 17-6.

Fig. 17-5 Developmental de-

sign, timing mechanism.
 Advanced Techniques 305

Fig. 17-6 Eighteen of the nineteen mechanism parts studied.

The FAST diagram shown in Figure 17-7 organized the logic and
formed the basis for developing improvement thinking. The diagram
was then rearranged somewhat, and the names of the parts that accom-
plish each function were added. See Figure 17-8.
Functions were evaluated, function constraints were clearly understood,
and the creative problem setting and solving of the job plan were used,
with the result that it was determined that the functions of sixteen parts
could be performed by one, as shown in Figure 17-9. The resulting pro-
duction design is shown in rear view in Figure 17-10. Of course, produc-
tion costs were correspondingly reduced to less than one-fourth of what
they would have been had the original developmental design been manu-
factured in quantity.

17-2 Quantitative Evaluation of Ideas3

As value analysis and engineering progresses it has become clear that the
original method of evaluating ideas by using the good/bad T chart has
serious limitations. For one thing, it precludes the use of value analysis and
engineering for such things as marketing trends or feasibility studies. If
the problem is putting men on the moon or building sophisticated radar
equipment for a strategic strike aircraft, cost may be a secondary factor,
the dominant factor being the possibility of being able to satisfy the re-
Written by Arthur Garratt, M.B.E., B.Sc., F. Inst. P., FRSA. Mr. Garratt and
Sonia Withers, M.A. of Value Engineering Ltd., London, England, devised the sys-
tem.
 - - --
I

DETONATE*
PRIMER
2
RELEASE
3
RELEASE
FIRING * COCKING
PIN SHAFT
- 4
RELEASE
SEAR
LEVER
5

~~~~~ -
* ",pip,':
RELEASE
*
ROTATE

TEEG
+
8
START
9

Sizp'-
PLUNGER
+
10
PIVOT
ARMING
LEVER
+
II
RELEASE
ARMING
PIN
*
I2
REMOVE
ARMING
WIRE

\ I
I \
\ \ \
\ I
\ I
\
i
FIRING WIND
CLOCK ARMING
SPRING PIN

I \
I \
ROTATE
1 \ 26
ROTATE
COCKING I \
\ TIMING
SHAFT I \ DlSC
COCK
I
FIRING I \'
I 27
POSITION 4 POSITION
SEAR I ARMING
LEVER LEVER

I STOP
19 21 23 PLUNGER 28
LATCH
TIMING TIMING * CLOCK + CLOCK
LEVER DISC ESSEMBLY HOUSING

\
-
\
MOUNT
CLOCK COVER
HOUSING PLATE

Fig. 17-7 FAST diagram of timing mechanism.

 Advanced Techniques 307

CRITICAL RELEASE
PATH DETONATE FIRING
FUNCTIONS. PRIMER SHAFT

I.FIRING
PIN
3. COCKING
SHAFT
CzEF
I LEVER
S.TIMING
LEVER
IO.TIMINGi
LEVER I
COMPONENTS REQUIRED
TO PERFORM FUNCTIONS: 2 SPRING 4. RELEASE I. STUD I
PIN I
5. PLUG

-
6
RELEASE
7
ROTATE
8 iI 9
RELEASE
ilo
PIVOT I RELEASE
l2
REMOVE
-t TIMING
LEVER
* TIMING
DISC
--t
START

T15.ARMING
START-STOP
PLUNGER
+ ARMING
LEVER tI ARMING
PIN
-+ ARMING
WIRE

b. TIMING
DISC
14. CLOCK ASSEMBLY
. CLOCK c. START-STOPI LEVER
PLUNGER I I6.PIVOT
PIN
I
I
I T ARMING '1
PIN
I8.ACTIVATION
1, 21. ARMING

u
2 2 . O-RING

Fig. 17-8 FAST diagram associating names of parts with the various functions.

Fig. 17-9 Sixteen parts ( A ) replaced by one ( B ).

308 Techniaues o f Value Analysis and Engineering

Fig. .10 Under side view

resu production pro,duct.

quirement. Of course, once satisfaction has been accomplished, value

analysis and engineering comes into its own to reduce costs and improve
performance. Another disadvantage of the simple evaluation system is that
at a value analysis and engineering seminar the ideas that are not de-
veloped are lost, and it often happens that a team has not enough knowl-
edge of company policy to choose the optimum solution, which may not
always be the lowest-cost solution. A further disadvantage is that there
may be two possible solutions, one costing, say, $25 and the other $26.
The second may be much superior from the performance point of view,
and yet if the first can be made to satisfy the functional requirements
partially,
. . the second will be lost.
It was with these considerations in mind that a more sophisticated
method of evaluation was devised. It was decided that it was necessary to
find a numerical method of rating performance and other features that
would also allow the users to separate their variables and judge how well
an idea satisfied each function separately. This led to a form of matrix
evaluation. The idea of matrix analvsis is not new and has been used
before for decision making. It has the advantage that numerical scores
can be worked out, with due weight factors, thus providing a figure of
merit for each possible decision. Perhaps of even greater importance is the
separation of variables. Many psychologists have shown that man has
serious limitations in his ability to combine factors in subjective decision
problems.4
4 See, for example, G. A. Miller, "The Magical Number Seven, Plus or Minus Two,"

Psychol. Reu, 63:81-97, 1956; and R. N. Shepard, "On Subjectively Optimum Selec-
tions Among Multi-attribute Alternatives," in M. W. Shelley and G. L. Bryan (eds.),
H u m a n Judgments and Optimality, John Wiley & Sons, Inc., New York, 1964, pp.
257-81.
 Advanced Techniques 309

The conventional method of matrix analysis is to set up a grid so that

different decisions can be evaluated against the factors that have to be
satisfied, with suitable weighting factors inserted. Such matrices have
certain disadvantages when applied to value analysis and engineering
evaluation. The particular objection is in the application of weighting
factors to the various functions. It was found that users generally con-
sidered three or four functions essential and gave them equal weight and
then degraded other functions out of all proportion to their importance.
Various modifications were tried until a method was found that is very
satisfactory in practice. The modified matrix is set up in the following
way.
The first step is to write down all the functions and characteristics that
must be satisfied. These are put in a positive form; e.g., for an electric
motor one writes down "low noise" as being desirable not "high noise" as
being undesirable. I t is possible to write the latter and then give it a
negative sign in the evaluation process, but it has been found that this
leads to some confusion and using the positive form throughout is better.
The functions and characteristics must now be arranged in order of
importance. Often this can be done by inspection. If not, a preliminary
scale matrix is used to sort out the order of importance of each function.
This checks against all the others in turn. The process is simple. On a
matrix sheet similar to that shown in Figure 17-11 all the functions and
characteristics are written down the left-hand side in random order and
numbered A to Z. Across the top of the matrix the same letters are also
written. A diagonal is drawn across the dead squares where a function is
compared with itself.
Starting with function A, go to the second square in the top row and
ask the question, "Is function A more important than function B?" If so,
fill in 1 in this square; if not, fill in 0. At the same time fill in the opposite
sign in the second square down, which is the complementary square and
shows that function B is less or more important than function A. Going
on step by step, the matrix is filled in. Then sum the numbers across the
matrix line by line and write the totals in the extreme right-hand column.
The function with the highest score is the most important; the rest are
graded accordingly. The one with a score of zero comes last and is the
least important. If the scores are not all different, a mistake in logic has
been made somewhere; this provides a check on the use of the system.
Let us take as an example Figure 17-12. Having put the functions and
characteristics into order, the main matrix can be started. This is written

FUNCTION 1 2 3 4 5 6 7 8 9 RATING
1. DISTRIBUTE AIR
2. CONTROL FLOW
3. DIRECT AIR
4 RESIST HEAT
5. RESIST CORROSION
6. LOOK GOOD
7. RESIST DAMAGE
8. EASY TO ASSEMBLE
9. PROVIDE RIGIDITY

Fig. 17-11 Function rating grid.

310 Techniques of Value Analysis and Engineering

MATRIX EVALUATION CHART

Fig. 17-12 Matrix evaluation chart. (Colt Heating and Ventilation Ltd. )

on a prepared sheet. (These can be purchased ready printed.) The func-

tions are written in their correct order across the top, with the most im-
portant function at the right-hand side and the functions numbered from
1 to n from left to right in the row marked "Function Rank Number."
Below this is another row, marked "Function Rating Number (+) ." This
enables a weighting differential to be applied to the functions while pre-
serving their overall order. Weighting is already built in by the rank
number, so that the most important function has the highest number, and
so on downward. But often a device may have, say, seven functions of
which four are very important and the other three, like "looks good," may
 Advanced Techniques 311

be much less important. The function rating number enhances the im-
portance of the really significant functions by suitably degrading the
importance of the minor functions. The rules of the game are that the
function rating number may be less than or equal to but may never exceed
the rank number, and so positive weighting is applied automatically. The
function rating number also allows two or more functions to be given
equal weight where their difference of importance is marginal.
All this may sound complicated, but in practice it is a straightforward
operation that sets out clearly criteria that must be satisfied. Now one can
proceed to the next stage, and the ideas that have been generated in the
creative session are filled in on the left-hand side of the matrix sheet.
Against each idea fill in an estimated cost in the extreme right-hand
column in the same way as done in the good/bad T chart.
Next, examine each idea on the basis of how well it satisfies each and
every function or characteristic listed across the top. These satisfaction
factors (s) scored 0 to 10 are filled in under each function. If any s fac-
tor is low, creative ability can be used to try and improve it, but remem-
ber that this may well alter the cost, which must be modified appro-
priately. In this way the ideas are under protracted scrutiny on both
performance and a cost basis. If any s factor is zero and cannot be raised,
the idea is rejected forthwith as being unworkable because it fails to
satisfy one function and no further time is spent on it. When all the
figures have been filled in, s x + is worked out for all functions and
summed across for each idea; it is filled in on the right-hand side under
Z+s. This gives a numerical value for the performance of each idea, duly
weighted for the importance of each function or characteristic and along-
side an estimated cost.
From the last two columns, an intelligent appraisal of each idea can
easily be made. If the problem is performance at any price, the idea with
the largest performance score can be selected. Any idea that has a lower
score and higher price than another idea can be jettisoned. Then appraisal
can be made on the basis of the market. It could be the supply of Christ-
mas novelties, where price is more important than performance. In this
case, the lowest cost idea is taken, providing that no s factor is zero. On
the other hand, if the target is a medium-quality market, one can select
the idea that is the best compromise between performance and price.
The matrix that has been described is a typical value analysis and
engineering exercise on a product or part of a product. But it is equally
valid for any management decision once the functions and characteristics
have been established.
It is obvious that performance is not the only criterion that can be
applied. A second matrix might be filled in, after a value analysis and
engineering session attended by designers and production men, by mar-
keting, using a new set of s factors on the basis of marketing. A third
matrix can be prepared on the basis of quality and reliability, and so on,
as required. Then an intelligent management decision can be reached to
give an optimum answer to several mutually conflicting needs.
312 Techniques of Value Analysis and Engineering

A particular feature of the matrix is that a team can prepare one during
a value analysis and engineering session, and this can later by analyzed by,
for example, a company executive who has before him all the ideas that
were generated, costed, and rated.
This system of matrix evaluation has been tried out at many seminars in
Europe, with outstanding success. At first, as a control, both systems were
used side by side, but it was soon found that the old good/bad T chart
led to confusion and teams using the matrix system performed better
evaluation much faster. Latest work has combined the matrix with the
good/bad T chart in reverse order: initial assessment is done on the matrix,
and problems of detail are then tackled on the T chart. The marriage
of the two techniques is then a happy one, with lusty offspring in the form
of new concepts. A lot of the sting can be taken out of the summation of
the product of c$ and s by using a desk calculating machine on which the
products are calculated without zeroing the total so that the final total is
Z d X s.
Readers are strongly recommended to try the system for themselves.
"Like capital punishment, it clears the mind wonderfully."