Ta.~Uc..

h\ '5

'/

/1

A 5 IlL-?
,&""1

,J)J?.

Ph nosofhX

Loss
. .' .Fundion
.
.. .
~

Taguchi has defined qualityas the loss imparted to society from the lime a product is
ship~.Societallo!lse8 include fuilure LO 'incch customer requirements, failure 10 meet
iQeal ~~onn~U'ice,; and harmful side eff~ts~ Many practitiorters liave i~luded lhe losses
due pr~ducii6n;' such as
m:iterial,!energy, and labor cClilsuinCd on unusable prod
u~ts'or toxic Jjy~prOqucts."" :<;' ';',.; .:,; ,; 'f ,,',i;II';),Ihl:, " ,"
, ; The 10ss-to..soietY concept can be ilI!lstrated byanexarhple isSOcWcd wilb lhe pr0
duction of large vinyl covers to protect' materials' from the elements. FiguR 2Q.. t .ahows
three stages in the evolution 'of vInyJ thic~ness. At (l):-~e' ~
is just apable of
I,;. '... "

raw

to

..!

~"

.,'

."".

Total loss = Producer's loss + Customer's loss

~
l_~

!i

LSL

USL

't

Vinyllhickncss

Figure 20-1 Loss to Society

Reproduced, with permission, from Taguchi Methods: Introduction to Quality Engineering (Allen Park, Mich.: American Sup
plier Institute, Inc" 1991).

l'

, meeting the specifications' (USL and LSL); howe~er, it is on the target tau, "C. I After con
'siderableeffort, the production prpcess was improved by reducing the variability about
the target, 'as shown at (2): In an' effort to reduce its production costs, the organization
, decided' tosh~ft the target closer to the LSL,as 'shown at (3). This action resulted in a
, substantial improvement by lowering tIle cost to the organization; however, the vinyl
covers were not as 'strong as before. When farmers used the covers to protect wheat from
the elements, they tore and a substanti4.! loss occurred to the farmers. I~ addition, the
cost of wheat increased as a result of supply-and-demand factors, thereby causing an in- ,
crease in wheat prices and a further loss to society. The company's reputation suffered,
which created a loss of market share with its unfavorable loss aspects.
Assuming the target is correct, losses of concern are those caused by a product's crit
" " ical perf0~ance. chara~teristics, deviating, fro~ the. target The importance of concen..
_, tratin~ on "hiHi~g th,e tar$ef~ is d~ume~ted by, Sony~ I~ ~pite: ()f the 'fact that the design
, ' . ,anq .Jsp,ecifict;ltiQns
w~re
identical,
U.S. c;ustorri~rs' preferred thec,olordensity of shipped
;
'. .
,'"
i
,.
.
, ,.,,;':! TV, ~~ts, produc~d, by,Sony:-:-JaP!1,n over~ose pt:Od~ced by Sony.~~SA. Investigation of
this situation revealed that the frequency distributions were markedly different, as
",' ,',;' .: shownin Figure 20-2. Eyen
though SOI)y-Japan had n.'3%outside' the specifications, the
.'!.
!
. f'" i:',:.::j(,.'i'.~ "!;~ ~;;).''1 .: .-,
;
,

t.

.',.

.,.;

'.

l'

".

,.,j

""

"

"

:' 11.,

'J.""

j"

\,

'I

' :

'

. ' , .'

"'~'.

I.

,.,'

,'J!

.r

' . t'
I

'.

~'

'"

i :.'

'. ~., .,: '.: '.'

. :'.1 "

Taguchi uses the symbol m fOr the target

"

V'
~/,

'

" .

'I :.r
1,

~ .t

.'

.~,

USL

"..i. .

t Gt.

If' i

"",I"

Sony-Japan

to

Sony-USA
.1

it.

~,

'
Color
density
-5

l'

Figure 20-2 Distribution of Color Density for Sony-USA and Sony-Japan

Source: The Asahi, April 17, 1979.
.

'

,f

:J.I~

.~
;

t'
~:

,/," .

i~

distribution was normal and centered on the target. The distribution of the Sony-USA
was uniform between the specifications with no vl:llues outside specifications. It was
clear that customers perceived quality as meeting the target (Japan) rather than just meet
ing the specifications (USA). Ford Motor Company had a similar experience with
tmnsmissions.
"'
Out of specification is the common measure of quality Joss. Although this concept
, may be appropriate for accounting. it is a poor concept for all other areas. It implies that
\"
.:
all products that meet specifications are good. whereas those that do not are bad. From
'.
the customer's point of view,the product that barely meets specification is as good (or
bad) as the product that is barely out of specification. It appears the wrong measuring
system is being used. The loss function corrects for the deficiency described above by
'.!
combining cost, target, and variation into one metric,/, '
-

,.'

.",

',:',
",,~_.

' ...

... r

'1

I.,

t,
"1

l. il'! r

j:;i l
I !:j ~
" ~ 1~

l
~J
1, ,
I;

I,

i
I
, 1n. !ff'
i!
1'1

, ,1:!

'l'

'II
{It

:'.ll[.,t

'!

.,,~

(,

..