Jou nal o Indus ial Enginee ing and Managemen
JIEM, 2014 – 7(5): 1385-1396 – Online ISSN: 2013-0953 – P in ISSN: 2013-8423
h p://dx.doi.o g/10.3926/jiem.1184
A Quan i a i e Analysis o Cellula Manu ac u ing in Appa el Indus y
by Using Simula ion
Guoqiang Pan
College o Mechanical Enginee ing, Zhejiang Uni e si y o Technology (China)
Depa men o Mechano onics and A ia ion, Zhejiang Ins i u e o Communica ions (China)
panmik[email p o ec ed]
Recei ed: May 2014
Accep ed: No embe 2014
Abs ac :
Pu pose:
This esea ch, by using he SIMIO simula ion pla o m, p o ides a quan i a i e and
compa a i e analysis o how he e iciency o ou di e en cell p oduc ion modes is a ec ed.
I is hoped ha he esul s will be o some help o ga men ac o ies o op imize hei
p oduc ion lines.
Design/me hodology/app oach:
The SIMIO simula ion pla o m was employed in he
esea ch and compa isons we e made o he simula ion es esul s abou he ou popula
cellula manu ac u ing modes.
Findings:
The ope a ion mode, numbe o ope a o s, and numbe o bu e a eas a e key
ac o s a ec ing he p oduc ion line e iciency, and need o be easonably se o achie e he
highes e iciency.
O iginali y/ alue:
As mos esea ch li e a u e so a is quali a i e, his esea ch p o ided a
simula ion-based quan i a i e analysis o he p oduc ion e iciency unde di e en cell
p oduc ion modes.
Keywo ds:
cell p oduc ion, SIMIO, simula ion, appa el indus y
-1385-
Jou nal o Indus ial Enginee ing and Managemen – h p://dx.doi.o g/10.3926/jiem.1184
1. In oduc ion
Cell p oduc ion e e s o a kind o p oduc ion mode in which he p oduc ion line is a anged as
an in eg a ed uni o equipmen and wo ks a ions based on he ope a ion p ocedu es and in a
pa icula o de ( o example U-shape layou ). The Ame ican Appa el Manu ac u ing
Associa ion has de ined cell manu ac u ing as “a con ained manageable wo k uni o 5 o 17
people pe o ming a measu able ask. The ope a o s a e in e changeable among asks wi hin
he g oup o he ex en p ac ical and incen i e compensa ion is based on he eam’s ou pu o
i s quali y ou pu ”. In a cell sys em, p ocesses a e g ouped in o a module ins ead o being
di ided in o hei smalles componen s, and ewe numbe s o mul i- unc ional ope a o s s and
s ill, o mo e coun e clockwise wi hin he pa icula p oduc ion a ea o pe o m he assigned
asks. Compa ed wi h he adi ionally specialized low-line p oduc ion me hod, cell p oduc ion
can swi ly adap o a iable o de ype and quan i y. When app op ia ely applied, i ensu es
lexible p oduc ion capaci y, lowe labo and equipmen cos , imp o ed p oduc ion e iciency. I
he e o e becomes a signi ican componen o he lean p oduc ion whe e JIT is pu sued, and is
widely adop ed by many indus ies including appa el manu ac u ing, elec onics assembly and
machining in Japan, Eu ope and Ame ica.
The e e imp o ing li ing s anda ds and apidly changing ashion ends a e pushing appa el
manu ac u e s o espond as as as possible o model and quan i y changes and o p oduce
high-quali y, low-cos p oduc s. Gi en his challenge, new p oduc ion sys ems we e adop ed
and expe imen ed in he appa el indus y, and he cell p oduc ion is widely a o ed.
2. Re iew o he li e a u e
Resea ches abou cell p oduc ion mainly ocused on wo aspec s: one was abou he
ad an ages o cell p oduc ion, while he o he was abou pe o mance o a ious cell
p oduc ion modes, hough mos esea ches include hese wo aspec s a he same ime.
Fa ma and Canan (2007) in oduced ha he cell sys em was i s implemen ed a Toyo a in
1978 as pa o JIT, and was known in he 1980s in he Wes as he Toyo a Sewing Sys em
(TSS). They ci ed o he esea che s’ esea ch indings and lis ed ad an ages o cell p oduc ions
modes. Bischak (1996) examined he pe o mance o a manu ac u ing module wi h mo ing
wo ke s and ound ha he p oduc ion sys em helped o achie e quali y p oduc s a lowe
cos s and educed lead ime. Besides, i could easily adap o high employee mobili y.
Resea che s a o ed simula ion in s udying pe o mances o di e en cell p oduc ion modes.
Wang and Ziemke (1991) simula ed a sys em based on he Toyo a Sewing Sys em (TTS) and
ound ha he sys em showed high pe o mance, e en hough he ope a ion imes o each
s a ion di e ed conside ably wi h each o he . They also no iced ha a e some ime o
-1386-
Jou nal o Indus ial Enginee ing and Managemen – h p://dx.doi.o g/10.3926/jiem.1184
ope a ion, he wo ke s would o m hei own wo k pa e ns. Adams and Sch oe (1999)
es ablished a simula ion based on he Rabbi Chase (RC) mode o clo h manu ac u ing, and
wo ked ou he law be ween ou pu olume and ime consumed. Black and Sch oe (1993) pu
o wa d he Linked Cell (LC) p oduc ion mode which used decouple s o connec , balance and
bu e elemen s wi hin he cell so as o achie e high e iciency. Wang and Tang (2009) s udied
he cell managemen me hod o manu ac u ing sys ems whe e pa s we e p ocessed ac oss
di e en cells. They cons uc ed a model and algo i hm, and used a simula ion o e i y he
e ec i eness o i s me hod. Schlo s like Ne land (2013), Wal e and Tubin (2013), Ka im and
A i -Uz-Zaman (2013) and Amin and Ka im (2013) s udied he assembly line layou s o cell
p oduc ion sys ems and e alua ed hei pe o mances. These esea ch inds a e o g ea alues
o u he s udies.
Pas esea ches a e o limi ed alue o he eal manu ac u ing wo ld, as hey mainly alida ed
he e ec i eness o cell p oduc ion sys em o ocused on pa icula cell p oduc ion modes.
The e was a lack o quan i a i e compa ison be ween hese di e en cell p oduc ion modes.
Besides, o a pa icula cell p oduc ion mode, impo an elemen s ha a ec he o e all
pe o mance o he sys em we e no adequa ely s udied. This pape aims, by using simula ion,
o compa e ou majo cell p oduc ion modes ha a e mos equen ly adop ed in he clo hing
making indus y, and y o ind ou he key ac o s ha impac he o e all pe o mances o
hese di e en cell p oduc ion modes.
3. Popula cell p oduc ion modes
The e a e many di e en cell p oduc ion modes being used in he p oduc ion indus y, howe e
he Sub Cell, Sha ed Cell, Rabbi Chase and Toyo a Sewing Sys em a e mos equen ly applied
in he appa el indus y. This chap e will gi e a desc ip ion o he ope a ion o he abo e ou
cell p oduc ion modes.
3.1. Sub Cell (SC)
Jus as Figu e 1 illus a es, a p oduc ion cell is di ided in o se e al sub-cells, which collabo a e
wi h each o he o inish he whole p ocess o wo k. Ope a o s a e only esponsible o he
wo k o hei own cells. Some imes he wo k o downs eam and ups eam s a ions can also be
included in he neighbo ing cells. Tha is o say, ope a o s a e esponsible o many asks in
bo h sides along he U-shape line so as o balance he whole p oduc ion p ocedu es. Fo
example, Op1 can be esponsible o he wo k in S a ion 1, 2, 12 and 13.
Labo di ision is adop ed in he Sub Cell p oduc ion mode, which has low equi emen s o
ope a o s’ comp ehensi e labo skills. They a e only equi ed o inish he wo k in hei own
-1387-
Jou nal o Indus ial Enginee ing and Managemen – h p://dx.doi.o g/10.3926/jiem.1184
cells. While enjoying less in es men o equipmen , Sub Cell mode equi es labo di ision o
balance he p oduc ion line. When necessa y, a p oduc bu e a ea in he sub-cell is necessa y
so as o educe he loss when he p oduc ion line is no in balance. The quan i y o wo k-in-
p og ess in he bu e a ea depends on he demands o p oduc ion balance, and he e a e
usually much wo k-in-p og ess.
Figu e 1. Sub-Cell (SC)
3.2. Sha ed Cell (SHC)
Jus as Figu e 2 illus a es, o educe bu e a ea and capaci y (namely, o educe he numbe
o wo k-in-p og ess), he i s and las s a ion in each sub-cell is designed as a “sha ed
posi ion” ( he s a ion labeled wi h S in Figu e 2). Neighbo ing ope a o s hand o e wo k in he
sha ed s a ion o in he bu e a ea.
In Figu e 2, he e is an “in e sec ion poin ” be o e and a e each sha ed s a ion and a “bu e
a ea” a e he i s s a ion in each sub-cell (excep he i s cell). Each ope a o can decide he
sha ed s a ion o wo k hando e depending on whe he he neighbo ing ope a o is busy o
no . Fo ins ance, a e he comple ion o he wo k o S a ion 1, Op1 will hand o e wo k o
Op2 in he “in e sec ion poin ” a e he S a ion 2 when Op2 is wai ing; ne e heless, when
Op2 is busy wi h wo k o o he s a ions, Op1 will con inue o inish he wo k o S a ion 3, hen
hand o e wo k in he in e sec ion poin a e S a ion 3 o pu he p oduc s in he bu e a ea
a e S a ion 3, and inally e u n o S a ion 1 o pe o m nex cycle o wo k. In he Sha ed Cell
mode, wo k hando e be ween neighbo ing p ocedu es should be pe o med in he in e sec ion
poin o bu e a ea a e one p ocedu e is inished, ins ead o be ween ope a o s di ec ly in
he p ocessing p ocedu e. In he Sha ed Cell mode, he e may be some wo k-in-p og ess in
he bu e a ea.
-1388-
Jou nal o Indus ial Enginee ing and Managemen – h p://dx.doi.o g/10.3926/jiem.1184
Figu e 2. Sha ed Cell (SHC)
3.3. Rabbi Chase (RC)
Jus as Figu e 3 illus a es, ins ead o labo di ision, ope a o s sha e he whole p oduc ion line
and a e espec i ely esponsible o manu ac u ing one p oduc in an o de ly way ( om S a ion
1 o S a ion 13). Di e en om Sub Cell and Sha ed Cell modes which ha e no pa icula
demands o he p oduc ion line layou , Rabbi Chase p oduc ion cell needs a U-shape layou
and ope a o s who a e p o icien in he wo k o all s a ions. Ope a o s pe o m hei wo k in a
“chase manne ”, hus line jam may a ise when some ope a o s inish hei wo k oo slowly.
Acco ding o Black and Sch oe (1993), jam is no likely when he numbe o ope a o s is less
han 3. Bu as he numbe g ows, a jam is mo e likely. In he Rabbi Chase mode, he e is no
bu e a ea be ween di e en s a ions and he e is no ex a wo k-in-p ocess.
Figu e 3. Rabbi Chase (RC)
3.4. Toyo a Sewing Sys em (TSS)
TSS can also be called Bucke B igade o Expand Rabbi Chase. Jus as Figu e 4 illus a es,
ope a o s sha e he whole p oduc ion line. Each ope a o ecei es he wo k-in-p og ess a he
s a ing poin o om he o me one, p ocesses i ill he end o he p oduc ion line o hands i
o e o he nex ope a o , hen e u ns o he ini ial s a ion. In his mode, he hando e o
wo k can ake place a any ime and any place. As shown in Figu e 4, Op1 and Op2 hand o e
-1389-
Jou nal o Indus ial Enginee ing and Managemen – h p://dx.doi.o g/10.3926/jiem.1184
wo k a S a ion 3, bu hey may hand o e wo k a S a ion 2 o 4 nex ime; a e he
hando e , Op1 e u ns o he s a ing poin while Op2 pe o ms he wo k coun e clockwise ill
he hando e o wo k o Op3. The Toyo a Sewing Sys em cell p oduc ion mode is a sel -
balancing sys em. Only h ough se e al ials, in e sec ion posi ions will end o be s able, and
he whole p oduc ion line will s ay s able and balanced. In his mode, each ope a o is mul i-
skilled. Ope a o s will be a ed acco ding o hei wo king p o iciency, and hose who a e he
mos p o icien a e always pu a he ends o he line in o de o imp o e he whole p oduc ion
e iciency. Like RC mode, he e is no bu e a ea be ween di e en s a ions and he e is no
ex a wo k-in-p og ess in TSS.
Figu e 4. Toyo a Sewing Sys em (TSS)
4. Simula ion es ablishmen
Fo he p oduc ion p ocess o a ce ain ga men which include 9 wo k s eps as shown in Table
1, we es ablish he SIMIO simula ion model. Mos o he p ocess in ha p oduc ion cell is he
sewing ope a ion accomplished by manually ope a ed sewing machines (di e en ypes) and
he ope a ion ime has ce ain andom luc ua ions.
4.1. Da a collec ion
We use a s opwa ch o conduc he da a acquisi ion o en-day ope a ion ime du ing he
pe iod o 2013.01.07-2013.01.11 and o 2013.01.14-2013.01.18. We employ he con inuous
eco ding me hod o conduc consecu i e ime measu emen 50 imes (10 ial obse a ions)
a 8:00(s a -up ime in he mo ning), 9:30, 11:30 (hal an hou be o e going o wo k), 13:00
(s a -up ime in he a e noon), 14:30 and 16:30 (hal an hou be o e going o wo k)
espec i ely. The da a acquisi ion equency o each s a ion is 3000 imes (300 imes each
day). Then a e we employ he iple s anda d di e ence me hod ( he absolu e e o is ±3 ％
and he eliabili y is 95%) o emo e he ou lie , we use Mini ab o ca y ou he da a i ing
analysis and ind ha manual ope a ions basically acco d wi h he iangula dis ibu ion o
ce ain pa ame e s shown in Table 1.
-1390-
Jou nal o Indus ial Enginee ing and Managemen – h p://dx.doi.o g/10.3926/jiem.1184
S ep Ope a ion Ope a ion ime/s S ep Ope a ion Ope a ion ime/s
1 Seam pocke s T(48.1,50,51.5) 8 Sew elas ic bands T(27.8,30.1,33)
2 Seam pocke s in o pan s T(68.2,70.1,72) 9 Seam bu ons T(14.3,15.2,16.5)
3 Seam(connec ) wo legs T(57.6,60,64.1) 10 Make bu on placke s T(8.9,10.2,12.1)
4 Sew wo legs T(37.2,40.2,44) 11 Seam ubbe s i ch T(12.5,15.3,17)
5 Sew pan s opening T(91.4,95.8,100.3) 12 Tes T(17.2,20,35.6)
6 Flip pan s T(15.4,18,23.3) 13 Package T(18.1,20,24.2)
7 Flange T(37.8,40.2,42.1)
Table 1. Ope a ion p ocess and ime measu emen
4.2. Basic SIMIO model
Es ablish he basic model o cell p oduc ion unde he SIMIO en i onmen . As shown in Figu e
5, each se e in he model s ands o a wo ks a ion and co esponds o he equipmen M and
each Node ep esen s he p ocedu e hando e poin o bu e a ea.
Figu e 5. The basic model unde SIMIO en i onmen
4.3. Assump ions and cons ain s
Make he ollowing assump ions and cons ain s in he SIMIO model:
1. he e a e no ob ious di e ences in skill le el among ope a o s and each ope a o has
iden ical e iciency o each p ocedu e;
2. he p ocessing ime o each p ocedu e complies wi h he s anda d ope a ion ime shown
in Table 1;
3. assume ha equipmen ailu es a e no aken in o conside a ion;
4. bu e a eas be ween cells can accommoda e mul iple a icles being p ocessed;
5. he walking ime be ween di e en de ices by ope a o s is uni ied o be se o 2
seconds. Time in Table 1 does no include walking ime;
-1391-
Jou nal o Indus ial Enginee ing and Managemen – h p://dx.doi.o g/10.3926/jiem.1184
6. he e a e no sho age p oblems in he i s s a ion;
7. all ma e ials and a icles being p oduced low uni-di ec ionally in cells acco ding o he
FIFO p inciple. Non-con o ming p oduc s a e only disco e ed and ea ed acco dingly in
he es p ocess.
In he SIMIO model, he e iciency o di e en models is es ed by con olling he numbe o
ope a o s and he olume o pe mi ed s anda d a icles being p ocessed. We need o ca y ou
sub-uni di isions unde he SC and SHC model. When he e a e 3, 4 and 5 ope a o s, he sub-
uni di ision esul s a e shown in Table 2. Unde he SHC model, each ope a ion uni has
dynamically sha ed s a ions.
Acco ding o he da a in Table 1 and Table 2 and ele an assump ions and on he basis o base
models, we es ablish ou simula ion models including he SC (sepa a ed ype), SHC (s a ion-
sha ing ype), RC ( abbi - chasing ype) and TSS (Toyo a ga men ype) and compa e and
analyze he a ia ions in p oduc ion capaci y and bu e capaci y unde hese ou models by
changing he numbe o ope a o s (con olling ak ime) and he numbe o ope a ing
s anda d a icles being p ocessed.
Numbe o cells Numbe o ope a o s SC model
Tasks
SHC model
Tasks Sha ed s a ions
3
1 1,2,3 1,2,3,4 3,4
2 4,5,6,7 3,4,5,6,7,8 3,4,7,8
3 8,9,10,11,12,13 7,8,9,10,11,12,13 7,8
4
1 1,2,9,10 1,2,8,9,10,11 2,8,9,10,11
2 3,4,8 2,3,4,7,8,9 2,3,4,7,8,9
3 5,6,7 4,5,6,7,8 4,7,8
4 11,12,13 10,11,12,13 10,11
5
1 1,2 1,2,3 2,3
2 3,4 2,3,4,5 2,3,4,5
3 5,6 4,5,6,7 4,5,6,7
4 7,8,9,10,11 7,8,9,10,11,12 6,7,11
5 12,13 11,12,13 11,12
Table 2. Sub-uni di isions unde SC and SHC model
5. Tes and esul analysis
Acco ding o he ac ual ope a ion o p oduc ion line, we se 3~5 ope a o s, 0~3 bu e capaci y
unde he SC model, 100-day ope a ion ime o SIMIO simula ion model wi h 8 hou s each day
and 10-day o Wa m up ime pa ame e s. Then we ope a e he simula ion model and conduc
he ou pu olume and bu e capaci y analysis.
-1392-
Jou nal o Indus ial Enginee ing and Managemen – h p://dx.doi.o g/10.3926/jiem.1184
(1) The analysis esul o ou pu olume is shown in Table 3 ( he numbe a e SC and SCH
s ands o he bu e capaci y. Fo example, SC-1 means ha he bu e capaci y is 1. The
bu e a ea is no se up in p oduc ion cells unde he RC and TSS model).
Ope a ion
model
Daily ou pu Daily ou pu pe capi a
3
ope a o s
4
ope a o s
5
ope a o s
3
ope a o s
4
ope a o s
5
ope a o s
SC-0 154 197 258 51.33 49.25 51.6
SC-1 161 203 278 52.67 50.75 55.6
SC-2 163 203 278 52.67 50.75 55.6
SC-3 159 203 278 53 50.75 55.6
SHC-0 146 178 213 48.67 44.5 42.6
SHC-1 168 203 272 56 50.75 54.6
SHC-2 168 201 273 56 50.25 54.6
SHC-3 168 201 273 56 50.25 54.6
RC 170 225 276 56.67 56.25 55.2
TSS 171 225 282 57 56.25 56.4
Table 3. Tes esul o ou pu olume
F om Table 3, we know ha :
•when he e a e 3 ope a o s: SC-3 has he highes ou pu while SC-0 has he lowes
ou pu unde he SC model, which indica es ha highe ou pu can be gained when he
maximum designed bu e capaci y is 2; SCH-1, SCH-2 and SCH-3 ha e iden ical
ou pu unde he SHC model. We can see ha he inc ease o bu e capaci y will no
lead o mo e ou pu ; he ou pu s unde he RC and TSS model a e basically he same,
bo h abo e he highes ou pu unde he SC and SHC model.
•when he e a e 4 ope a o s: SC-1, SC-2 and SC-3 ha e he same ou pu and he
inc ease o bu e capaci y will no lead o mo e ou pu unde he SC model; SCH-1 has
he highes ou pu unde he SHC model, which indica es ha maximum ou pu can be
ob ained when he bu e capaci y is 1; he ou pu s unde he RC and TSS model a e
basically he same, bo h abo e he highes ou pu unde he SC and SHC model.
•when he e a e 5 ope a o s: maximum ou pu can be acqui ed when he bu e
capaci y is 1 unde he SC model; maximum ou pu can be acqui ed when he bu e
capaci y is 2 unde he SHC model; he ou pu unde he TSS model is highe han he
highes ou pu s in he o he h ee models.
•when he e a e 5 ope a o s unde he SC model, he p oduc ion cell pe capi a has he
highes ou pu ; when he e a e 3 ope a o s unde he SHC model, he p oduc ion cell
-1393-
