00 ETHODS FO TOTAL QUAITY ANAGEENT
100
METHODS FOR TOTAL QUAITY MANAGEMENT Gopal K. Kanji and Mike Asher
Lodo
P •
Touad Oa New De
© Gop K. Kn n Mk h 996 t puh 996 ht v. o pt o th pucton my pouc to n tv ytm, tnmtt o ut n ny om o y ny mn ctonc mchnc photocopyn con o othw wthout pmon n wtn om th Pulh. SGE Pucton Lt 6 Bonhl Stt Lonon U SG Pulcton nc 255 Tl Ro Thoun Ok lon 932 S Pucton n Pvt Lt 32 MBlock Mkt Gt Kh I w Dlh 1 1 48
Bitis Liba atalin in blicatin data ctou co o th ook v om th Bth Ly ISB 839 776 8 ISB 839 777 6 (pk
Liba f nss catal cd availabl Typt y Photopnt Toquy Dvon Pnt n Gt Btn t th nvty P m
CONTENTS
Preface
vi
Understanding Total Quality Management
The Role of TQM Methods
List of Methods (by Category)
Purpose of Methods Aphabetical ist)
3
Management Methods
Analytical Methods
79
Idea Generation
3
Data Collection, Anaysis and Display
54
References
34
Index
36
PREFACE
A eea par o e deveopme o a oa quay maageme (TM) proce oud be e educao ad rag o everyoe e orgazao Te m a objecve oud be o provde ormao o e prcpe ad poopy o TM ad rag e meod o ep e orgaz o mpeme oa quay maageme a yema c way Oe o e ma purpoe o boo o ep a empoyee o uderad e proper ue o e oa quay maageme meod requred or e aceveme o er orgazao quay goa w ao provde e educaor ad pracoer area w a compree ve e o TM meod Toa quay maageme ecompae cera bac prcpe To mpeme ad prace ee prcpe eceary o uderad e wor g o varou meod o our oacaegore quay maageme Tee meod are caed boo w
anagemen mehod or exampe, Dem g wee (Meod 10) Analycal mehod or examp e aure mode ad e ec aay (Me od 3)
dea generaon: or exampe, brao rmg (Me od 51 ) . Daa collecon analy and dlay or exampe, ay car (Meod
96), ogram (Meod 8) ad pe car (Meod 88) epecvey. A o meod by c aeg ory, gve o p . 0; a apabe ca o all meod (w a bre de crpo o er purpoe or ue) gve o p 13 W eac caegory, eac TM meod expaed mpy uder e oowg eadg • • • • •
purpoe we o ue ow o ue bee exampe
We pu o pracce, approprae qua y meod ca rap dy gve re o quay mproveme . Coo g e rg meod or e deveopme o a TM proce oe o e va roe o maageme ad e degree o ucce obaed w deped upo maagera Te oa quay maageme proce compex ad e ue o ome o ee meod requre car eu coderao ad cear uderadg.
NDERSTANDING TOTAL QALITY MANAGEMENT tal quality management principles
To udertad te proce of tota quaity maagemet (TM) we wi foow Kaji ad Aer (993) were a wor i ee a proce ad tota quaity maagemet i a cotiuou proce of improvemet for idivi dua , group of peope ad woe orgaiatio Wat mae tota quaity maagemet differet from oter maagemet procee i te cocetrated focu o cotiuou improvemet. Tota quaity maage met i ot a quic maageme t x; it i about cagig te way tig a re doe w iti te orgaiatio i fetime To improve te proce terefo re, peope mut ow wat to d o , ow to do it, ave te rigt met od to do it, ad be abe to meaure te improvemet of te proce ad te curret eve of quaity acievemet Tota maagemet ecompae a e t of four pricipe ad eigt core cocept Te four guidig pricipe are • deigt t e cutomer • maagemet by fact • peope baed maageme t • cotiuou improvemet Eac of te pricipe ca be ued to drive te improvemet proce However, to acieve ti eac pricipe i expreed wit te ep of two core cocept to mae te pricipe worabe Te eigt core coc ept are give i Tabe 1 Delight the ustomer
Ti focue o extera cutomer ad a 'Wat woud deigt tem Ti impie a rea eed to udertad te product or ervice, agree requiremet ad fu tem 'Deigt mea beig bet at wat reay able 1
Priniples and ore oneptso TQM
Prncpls
or concpts
Dht th customr
utomr stscton Intrnl customrs r r work s procss Msurmnt mwork Pop mk qut ontnuous mprovmnt cc Prvnton
Mnmnt b ct Poplbsd mnmnt ontnuous mprovmnt
100 METHODS FOR OA QALIY MANAGEMEN
matte r mot to the cutomer and thi can change ov er tim e Bei ng in touch with thee change and alway atifying the cutomer are an integra part of tota quaity management Management by act
Knowing the current tandard product or ervice your cutomer hand i thequaity rt tage of be ofingthe abe to improve Youincan ony meaure your improvement if you know the bae you are tarting from. Having the fact neceary to manage the buine at a eve and giving that i nformation to everyone o that deciion are baed upon fac t are an eentia apect of continuou impr ovement Peoplebased management
f peope unde rtand what to do how to do it and obtain feedback on their performance they can be encouraged to take repo nibii ty for the quaity of their own wor k The more peope fee invoved the greater wi be their commitment to cutomer atifaction Sytem tandard and technoogy themeve wi not provide qua it The roe of peope i extremey important in the continuou improvement of quaity within an organization ontinuous impovement
Tota quait y management i not a hortte rm activity tha t wi nih when a et target ha been achieved t i not a programme or a proect . t i a mana geme nt proce that recognize that ho wever much we may improve our competitor wi continue to improve and our cutomer wi expect more fom u Here continuou impovement i an incrementa change and not a major breakthrough which houd be the aim of a who wih to undertake the tota quaity management journey re cncepts fr imprvement
Each of the eight core concept given n Tabe 1 can be ued to drive the proce of continuou improvement and to deveop a framework for quaity improvement over many year ustomer satisaction
Many companie when they tart the quaity journey become very introverted and dea with their own nterna probem negecting their externa cu tomer A better way i for companie to ue thei r cutomer t o earn what i important to them and then meaure their own performance againt cutomer expectation. Aking your cutomer to et cutomer atifaction goa i a cear ign of an outward ooking company To fu cutomer atifa ction Federa Expre an American compan y urveye d thei r cutom er to iden tify the top te n caue of ag gravation The aggravation point were weighted according to cutomer view of how important thee were A comp ete check wa made of a occurren ce and a
NDERSTANDNG TOAL QALTY MANAGEMENT
eople
rocess
rodcts
qpment
or
or
ateral
ysem
evces
3
ethods nvronment
Fg
1
Pocess
weekly satsacton nde comed Ts allowed te comany to kee a weekly montor o customer satsacton as measured by te customer Inteal customes ae eal
anj and Asers ( 1993) denton o qualty 'satsyng agreed cus tomers requrements relates to nternal customers a s well eternal ones Many eole also reer to te customer suler ca n We beleve tat t s necessary to aceve successul nternal workng reatons n order to sats y te needs o te eternal customer Weter you are sulyng roduct s or a servce, te eole you suly nternally are as rea as your eternal customers Tey also requre seed, ecency or accurate measurement, but acevng a qualty servce between nternal customers can sometmes be t meconsumng One way to deal wt ts s to assess oor qualty n nancal terms Measurng te actual cost o oor quaty, and te way tat amount s made u, can rovde an metus or management to olow te qualty mrovement at. n ts way, you can use te dea o te terna customer as a ocus or mrovement All wok is pocess
Anoter ossble ocus or mrovement s tat o busness rocesses A rocess s a combnaton o metods, materals, manower and macnes (see Fgure 1) tat, taken togeter, roduce a roduct or servce All rocesses contan nerent varab lty an d one aroac to quaty mro vement s rogressvely to reduce varaton Ts can be done, rst, by removng varaton due to secal causes and, secondly, by drvng down te common cause o varaton, tus brngng te rocess under contro and tem mrovng ts caablty Measuement
Ts core concet o total qualty management suggests tat, n order to mrove, we must rst o al measure ow we are dong at resent By measurng our resent stuaton, we can ocus bot nterna customer
4
1 00 MHOS OR OAL QUAY MANAGMN
satisfaction and xtrna customrs rquirmnts ntrna quaity masurmnt of roduction migt incud • • • • • • • • •
brac of romis rformanc to standard rct v accidnts rocss in contro yid/scra tim cost du to non-avaiab raw matria numbr of cang s to works ordr cost of quait y
eamwok
Tamwork can rovid a ra oortunity for o to work togtr to aciv quaity i mrovmnt Po w o work on tr own or n a sma grou oftn av a ictur of tir organization and t work tat i t dos wc s vry comartmntazd Ty ar oft n unfamiiar wit t work tat is don vn by o wo work quit nar to tm as a rs ut ty ar unawar of t consquncs of oor quaty n t work ty tmsvs do Bringng o togtr in tams, wit t common goa of quaity imrovmnt aids communication btwn dartmnta or functiona actviti s Tamwork sowy braks down t communication barrir s and acts as a atform for cang ommunicaton is art of t cmnt tat ods tog tr t bricks of t tota quait y managmnt rocss su ort ing t rinc of o basd man agmnt . To communcat rory , i t is ncssary to foc us on t rcvr of t mssag ommunication is vry muc a twoway rocss Managrs o ftn tak about t 'midd managmnt song into wic information sms to buton outtof wc i nformat ionng com Part of t a acka of go focus nds no of tos rc ivi t isnformato n robm Figur sdicts comany wit oor communication For succssfu communicaton, you nd to bud crdibiity nto t mssag and n t rson gving t mssag Anyting tat dtracts from tis dos damag to bot Tamwork aso nabs a grou of o to work as a task forc, ooking at crossfunctona robms or as an action tam soving oca robms, in ordr to idntify and adot nw ways of doing tings People ma ke uality
Most of t quaity robms witin an organization ar not normay witin t contro of t individua moy As many as 80 r cnt of ts robms ar causd by t way t comany is organzd and mana gd T systm oft n gts in t way of moys wo ar tryng to do a good job n suc a situation it s difcut t o sov t robm by sm y ti ng t moys to do bttr n ts circ umstanc s moy
NDRSTANDNG TOTA QALTY MANAGMNT
5
io maageme t
Fgre
2
Poor communication in an organization
motvato n alone cann ot work. It requres real ractcal effor ts on te art of managers to remove te barrers to qualty mrove ment Te role of manage rs wt n an organzaton s to ensure tat everytng necessar y s n lace to allow eole t o make qual ty Ts n turn begns to create te envron ment were eole are wl lng to take res onsblty for te qual ty of ter own work. Releasng te talents of everyone wtn te organzaton n t s way can create a culture for qualty mroveme nt ontinuous improement cycle
Te contnuous cycle of establsng customers requrements, meetng tese requrements measurng success and keeng on mrovng can be used bot eternally and nternally to fuel te engne of contnuous mrovement By contnually ceckng customer requrements a comany can kee ndng areas n wc mrovements can be made Ts contnual suly of oortunte s can be used to kee qual ty mrovement lans u to date and to renforce te dea tat te tota qualty journey s neverendng Preention
Ts concet s centra l to total qua lt y manageme nt and rovdes a ostve aroac to acevng contnuous mrovemen t Prevent on means seekng to ensure tat falures wll not occur Te contnual rocess of removng te roblems and falures out of te system wll create a culture of contnuous mrovement
6
METHOD FOR TOTL QLTY MNGEMENT
There are severa methods whch are wdely used or ths urose Falure ode and eect analyss Method 3) s a wellknown method assocated th both desgn an d roce ss analy ss. yramid mdel and leadership
From the ou tset t he total qu aty manage ment ar oach has th e vson that concentrated management acton can mrove the qualty o servce or roducts o an organzaton at a very comettve cost satsyng cus tomers needs and ncreasng market share Ths ncreased market share wll be stable because t has been earned wth the hel o sod customer goodwl and not by gmmcky advertsng. anj and Asher (993) suggested a model whch llustrates the rnc les o TM as a yramd The base o the yramd s occued by the our rncles o TM and two core concets corresond to each sde o the yramd. Although n anj and Ashers model the leadersh o to management s central to the creaton o a TM organ zaton ths s not emhaszed ther o dagram. We thereore roduce oo the yram dnmodel TM (see Fgure 3) by smly a moded etendngverson the base the yramd Here the organzaton has to be guded through the TM rncles a nd core concets by to management eadersh
UNDERSANDING OAL QUALIY MANAGEMEN
T Q M elht he csomer eadershp Figr 3 Pyamid model o TQM Souce Ka a Ahr 3
THE ROLE OF TQM METHODS Total qualty anagnt can b lntd by uttng nto ractc su tabl TQM to ds. Howr, adot ng t rgt knd o tod s on o t ost ortant jobs o snor anagnt and t dgr o succss obtand wll dnd on tr knowldg and undrstandng o ts tods. TQM tods ar unlkly to b usul not usd rorly Wn TQM s lntd n an organzaton , t narably starts wt a sl rocdur t sttng u o tas to sol artcular robls. Howr total qualty anagnt dals wt qualty cultur, wc s all about cultur cang basd on a dsr to satsy t custor and lnat stng robls ranntly. Educaton and tranng ar ky actors n total qualty anagnt, ncludng t rocss o larnng TQM tods tas start to look at qual ty anagnt robl s wtout ro r tranng ty wll los tr way and bco dsartnd t qualty robl s not dntd accuratly and t TQM tod slctd o r soluton basd onl y on data analyss, tn t robl wl l not b lnatd orr n addton, t qua lt y ront rocss nds to b anagd by an ct lad r to nsur tat ror lntaton s ac d . By alyng TQM tods rorly and not uctuatng ro on st to anotr bor colton, t qualty ta wll a a uc gratr canc o coltng t task succssully T robl-solng rocss s a natural and logcal squnc or orcong qualty robls and rong t standard o dcson akng t s also a gud or dntyng wc total qualty anagnt tods to b al d Probls no attr wat tr sz or c ol t y, can b st b sold by roc dng troug a squnc o sts. T s nsurs tat rytng ossbl wll b don by alyng t aalabl TQM tods n t ost ct annr. t also gs t oortunty to consdr a nubr o otons and to slct t bst solutons . Many q ualty robls , on t sura c, a ar to b sl to sol, and t s asy to la to t rst aalabl soluton How r, n t long tr, or any robls t s unl kly t at t bst solutons wll b ound n ts asy way. t s also ossbl tat so sd-cts wll b gnratd, causng robls n otr aras. n total qualty anagnt all work s a rocss and t robl sol ng rocss s a contnuous cycl o onng your nd to a wd rang o ossbl solutons and tn dcdng on t ost asbl oton. t s ts contnuous aroac and t narrowng o t otons tat aks t TQM rocss so owrul.
THE ROLE OF TM METHOS
9
Te basic rol e of TM met ods in r oblmso lving for qual ity im rove ment is to el meet customer requirements. Te metods also el to generate ossible root causes and otential solutions, and to use data and information to select te best otions for managing quality To imlement total quality management it will be necessary to aly te metods in every asec t of business life
LIST OF METHODS (BY CATEGORY) anagement methds
1 Ap abl qualiy lvl AQL ) Afny dagram 3 Arrow dagram 4 Bnmarkng 5 Connu raing 6 Coningny planning Cobn analyi 8 Criria ng 9 Cuomr onngny abl 10 Dming wl PDCA) 11 Dparmnal purpo analyi DPA)
20
1 Error Forproong analyi pokayok)
43 44
14 Gann ar 15 SO 9000 16 u n im ) 1 Kaizn 18 Myry opping 19 Objv rankng 0 Paro analyi Ponial problm analy PPA) Problm prvnon plan 3 Pro dion programm ar 4 Programm valuaion and rviw PER) niqu 5 Qualy irl 6 Quaiy funon dpoymn QFD) Rlaion diagram 8 amwork 9 Toal produiv mainnan 30 Wyow aring 31 Zro df
23 25
30 32 33 35
37 39
41
46
48 50
51 52 54
56 59 61
63 65
6 69
74 75 77
78
nalytical methds
3 Cau and ff analyi
9 8]
33 Crial pa analyi 34 Dparmnal o of CPA) qualy 35 Domanal mapping 36 Evoluonary opraion EVOP)
85 87 89
LIST OF METHOS
37 38 39 4 41
ailure mode and effe analyi MEA) aul ree analyi ore eld analyi Minue analyi Paired omparion
4 43 44 45 46 47 48 49 5
Parameer deign Proe o of qualiy Reliabiliy Robu deign off-line qualiy onrol) Soluion effe analyi Sraiaion Syem deign agui meod olerane deign
9 96 98 10
03 105 107
1 112
13 115 17
19 122
ea generatin
51 5 53 54 55 56 57 58 59 6 61 6 63 64 65 66 67
Brainorming Brainwriing Breaking e Buzz group dea wriing magineering mprove inernal proe P) plan Laeral inking Li reduion Mind mapping Morpologial fore d onneion Mulivoing Nominal group enique Opporuniy analyi Ri piure Snowballing Suggei on eme
3 15 17 9 13 13 134 136 138 14 14 144 145 47 149 15 153
ata cllectin analysis an ispay
68 69 7 71
Bar ar Bai aii Bo and wiker plo ar
Cekee diagram 7 Conenraion 73 74 Cuum ar 75 Do plo
154 56 59 16 64 66 68 7
12
76 77 78 7 8 81 8 83 84 8 86 87 88 8 1 3 4 6 7 8 1
100
MEHODS FOR OAL QUALIY MANAGEMEN
Flowars Geomer movng average Hsograms Hosn kanr (qualy poly deploymen) s/s no mar
173 17 177 18 184
Mar daa analyss Mar dagam Movng average Mul -var ars NP ar Payner ar P ar Pe ars Proess analyss Proess apabl y Samplng Saer dagrams Spder web dagrams Sasal proess onrol (SPC ) Sem and lea f dagram Tally ars Tree dagrams U ar X movng range (X) ar ar
186 188 1 1 14 17 1 4 8 1 1 14 16 1 1 3 8 31
PRPOSE OF METHODS (APHABETICA IST) Method
o povide a uue of ampling plan ik and inpeion aegie o enue a e uome eeive e qualiy a e upplie a onaed o delive. Anity diagram o oganize lage amoun of daa in goup aoding o ome fom of naual afniy. Arrow diagram o ow e ime equied fo olving a poblem and wi iem an be done in paallel Bar charts o diplay diee daa olleed by ekee o a paen an b e dioveed. Basic statistics e mean median mode ange and
Acceptable quality level (AQL)
andad deviaion ae e wayi of ummaizi ng and deibi lage volume of daa ee ae meaue of ng loaion e la wo ae meaue of pead. Bencmarking o idenify and ll gap in pefomane by puing in plae be paie eeby eabliing upeio pef omane Box and whisker plots o povide a imple way of dawing e bai ape of e diibuion of a e of daa. Brainstorming o geneae a many idea a poible wiou aeing ei value Branwrtng o geneae a many dea a poble Breaking set o oveome blok in inking by geneaing new idea. i paiulaly ueful in pomping a goup o be moe eepive o new uggeion Buzz groups A way of geing e immedia e eaion of a goup o a new idea o poble m . Cause and eect analysis o eamine effe o poblem o nd ou e poible aue and o poin ou poible aea wee daa an be olleed chart o idenify wen e numbe of defe in a ample of onan iz e i anging ove im e. Checksheets o olle daa wen e numbe of ime a defe o value ou i impoan. Concentration diagrams o olle daa wen e loaion of areaching defe o poblem impoan Consensus o give aieam a meodial way of eamining alenaive o ea a olleive onluion wi all eam membe an aep.
Page
5
68
5
69
56
7
7
59
5
5 5
5 7
54
9
79
7
6
7
6
7
66
5
0
ME OS FOR OA QUAL Y MANAG EM E N
o avo gng an wa of ou by lannng fo onngn n omlon of a oj Co-bene analyi o ma al o an bn of a oj un onaon
6
3
7
33
Cieia Tein
8
35
33
81
9
37
74
168
1
39
34
85
11
41
35
87
75
171
1
43
36
89
37
91
38
96
76
173
13
44
Coninency plannin
o by valua oluon o a oblm ang an moma agan aalnav l of a Ciical pah analyi A oj lannng nqu w aa wok o b on n o lmn allowng ky lmn a aff ovall oj o b n Cuome' coninency able o unan n of bo nnal an nal uom fo fullmn of uom a fa on Cuum cha o nfy wn man valu angng ov m Demin wheel (PDCA) A managmn on o af y qualy qum n of uom by ung yl lan o k an aon. Depamenal co of qualiy o ov a nanal mau of qu aly foma n of an oganzaon Depamenal pupoe analyi (DPA) o vw nnal uom ul la on. Domainal mappin o a n naon of nnal uom an n Do plo A ml ga v w n obvaon a o on a ozonal al Eo poon (pokayoke) o gn an oaon n u a w ay a o a vn fo m aung majo oblm o uom. A qunal
Evoluionay opeaion (EVOP)
mnal ou fo ollng nfomaon ung onln ouon o mov a o wou ubng ouu. Failue moe an eec analyi (FMEA) o a n fooloong of a gn o a o Faul ee analyi o fom a quanav a wll a qualav analy of a oml ym Flowcha o gna a u of ow wok g on by lnkng og all akn n a o. Foce analyi o nfy nal an nnal fo a wok wn vlong a onngny lan Foce el analyi Allow you o nfy o fo a bo l an n you n long ga bwn w you a now an w you wan o b
39
98
PURPOSE OF M EOD S
15
Fo plann ng te tep neceay to mplement qualty mpovement. Gemeric mving average o dentf y tend n mal l cange n te poce mean e geometc movng aveage ometme called te eponenally wegted
14
46
77
175
movng ave age Hisgrams oEWMA) dplay contnuou data collected by ceckeet o tat any patten can be dcoveed Hshin kanri (quaiy picy depymen) o delg t te cutome toug te ma nufactung and evcng poce by mplementng te qualty goal of te oganzaton Idea riing o bng all patcpant nto goupwok Imagineering o at a company to dentfy aea of oppotunty by concent atng on te d eal outcome ten wokng back fom t Imprve inerna prcess (I) pan o povde te tuct ue to develop wok plan detal fo a tak ung vaou facto uc a meauable eponble eouce tme and pevou tak owne Is/is n marix o dentfy patten n obeved caac tetc by a tuctued fom of tatcaton ISO 9 o demontate to youelf you cutome and an ndependent aement body tat you ave an effectve qualty manageme nt ytem n pl ace Jus in ime (JI) o delve te aw mateal o component to te poducton lne to ave jut n tme wen tey ae needed Kaizen A apanee tem meann g can ge fo te bette te concept mple a contnuou mpovement
78
177
79
18
55 56
1
57
14
8
184
15
48
16
5
17
51
58
16
59
18
81
186
8
188
6
14
Gann chars
n all company functon at all level aera hinking A way o f tanfeng fo m on e fame of efeence to anote enablng you to beak down bae wc nbt ceatve tougt Lis reducin o educe a l t of dea t o one of manageable ze arix daa anaysis o povde a pctue of numecal data fom a mat dagam n an efcent way arix diagram o povde nfomaton about te elatonp and mpotance of tak and metod element of te ubject. ind mapping A way of geneatng and ecodng dea ndvdually ate tan n a goup. Mnd mappng make u e of wod aocaton encouagng you to follow you own tougt patten weeve
16
100 METHODS FOR TOTAL QUALITY MANAGEMENT
y ad ao provid a wrin rord of ida gnrad Minute analysis o ima urviva priod of a pariuar produ uni undr rain ondiion uing a imuad primna nvironmn.
4
1
Morphological forced connections
o gnra nw ida or way of appr oaing p robm . ombin i of aribu and for nw onnion bwn m o riggring nw opion. Moving average o idnify rnd in daa wn orrm variaion or yia parn ar onfuing ongrrm piur Multi-vari charts o ow diprion n a pro ovr or and ong rm uing a grapi onro ar. Multi-voting o mo popuar or imporan im from a i.
61
14
83
19
84
19
6
144
Mystery shopping A niqu invoving ooking your buin from ouid and mauring a
18
5
6
45
85
194
19
54
64
14
41
1
4
15
56
86
19
finy of your own ky pro from uomr viwpoin. Nominal group technique A way of gnraing ida from a group and dnfying v of uppo r wii n group fo r o ida. N chart o idnfy wn numbr of dfiv im in a amp of onan iz anging ovr im Objective ranking p o p a your urr n aiviy in prpiv and nab you o undrand purpo of your ffor. Opportunity analysis Offr opporuniy o va ua quky a ong i of opion again d ird goa and avaiab rour. aired comparisons o p a group o quanfy prfrn of i m mbr . arameter design o drmin w faor ar imporan in manufauring pro and o nd opimum of working ondiion areto analysis o para mo imporan au of a probm from many rivia . A o o idnify mo imporan probm for a am o work on aynter charts o dipay informaion ovr m in a way a aow ang in parn of faiur o b diovrd. Paynr ar wi ow wn on faiur
PURPOSE OF M EHODS
17
mode takes over rom anoter in terms o importance or wen te overall ailure rate is canging over time car o identiy wen te percentage o deective items in a s ample o variable size is canging over time
8
99
i cars way o pictorially pie carts are an Aeective mean s orepresenting sowing te data, relative
88
59
6
89
4
9
8
43
3
63
4
65
comple and multi-level projects Qaliy circls A special type o small group activity wic orms a veicle or te development o
5
6
individuals
6
69
2
44
65
49
size o te individual parts to te total o eamine plans to identiy wat can go wrong wit tem so tat preventi ve action can be taken rblm prvnin plan o anticipate wat can go wrong and plan to prevent problems rcss analysis Ena ble s a group to loo k or opportunities to improve processes t can also be used to identiy standards and measures or critical parts o processes rcss capabiliy o demonstrate w eter a process is capable o meetin g a specication and to calculate an inde to sow tis capability rcss cs f qaliy o provide a nancial measure o te quality perorma nce o an organization rcss dcisin prgramm car o ocus on possible sequences to elp lead to a desirable result and contingency planning nial prblm analysis (A
rgramm valain and rviw (ER cniq o establis a plan ning tecnique or
A tecnique or discipline or optimizing te process o developing and producing new products on te basis o customer need Rlain diagram o illustrate te relationsip between pro blems and ideas in comple situations Also to identiy meaningul categories rom a mass o ideas wen relationsips are difcult to determine Rliabiliy o n d te cause o ailur es and try to eliminate tem and to reduce te eects or consequen ces o ailure Ric picrs o allow a group to capture all ideas developed witout judgement or analysis in a pictorial orm tat a llows te strengt o te ideas to be recorded Qaliy fncin dplymn (QFD
18
1 00 MEHODS FOR OAL QUALIY MANAGEMEN
o acieve te proper functioning of a component even wen affected by interfering factor, weter eternal, internal or manufacturing variation. Smpling A metod by wic a mal number of
Robust design oline qulity control)
item te ample) drawn from a larger number of item te poplation) in order to draw a concluion about te popuation baed upon information from te ampe Sctter digrms o alow te reation ip between caue and effect to be etablied. Snowblling Sometime called pyramiding, nowballing i a tecnique for gatering information or idea Solution eect nlysis o eamine olution to problem to nd out weter tere are any detrimental conequence and to pan te mpementaton of te olution Spider web digrms
o ow performance againt a target wen everal criteria are being et Sttisticl process control SC) o identify wen procee are cangng over time . Stem nd lef digrm o preent raw data and to ow teir ditribution viualy Strtiction o ait in te denition of a probem by identifying were it doe and doe not occur Suggestion schemes o generate idea for improvement System design o appy pecia cientic and engineering knowledge to produce a baic functional prototype model, aving urveyed te reevant tecnoogy, manufacturing environment and cutomer need Tguchi methods A tecnique for te optimization of product or procee, aguci involve a twotage eperimental deign tat give te benet of robut ne and efc iency wit te m ini mum number of eperiment. Tlly chrts o colect data wen te vaue of a defect or probem i im portant Temwork o organize activity wic requre a number of peope to colaborate and work togeter for a common goal Tolernce design o nd out by eperiment were te variability in a proce product) occur and were adutm ent can be made
45
11
91
1
9
11
66
151
46
113
93
4
94
6
95
19
47
115
67
53
48
117
49
9
96
8
74
5
1
PURPOSE OF MEHODS o elp a proce wic am at makng te mot eective and ecient ue o eting producto n tructure . Tree iarams o identiy te tak and metod needed to olve a problem and reac a goal. Toa procive mainenance
9
75
97
3
98 3
5
99
8
3
3
78
char
o i denty t e numbe deect n a ample o varable iz wen e i cangng ove rr o time . Whyhow charin Wen tnkng n bot abtract and concr ete term and needi ng to move betwee n te two wyow carting enable a goal to be tranlated into acton. X movin rane X) char o dentiy wen a value i canging over time. char o identiy wen te mean va lue or range n a ample o contant ize i canging over time. Zero efecs o allow team to eper ence te ucce nvolved n meetng ever more demanding target witout demotivatng tem by not aceving abolute ucce at once.
77
MAN AGEMEN T M ETHODS M eth od 1
Acceptabe q ua ity leve (AQL)
urpse
To provde a ru ure of amplng pla n rk and npeo n raege o enure a e uomer reeve e qualy a e uppler a onraed o delver. When t use
Wen amplng by arbue o a ea em npeed laed a aepable or una epable . A QL an be ued fo r 'defe ve un were e un'defe laed defeve e.g norre or fauly whol boe); or for e.g a error on nvoe or mark form on panwork) w t use
Before any ba of work ampled eenal o ave era n pee d nformaon abou e qualy andard o be me ogeer w bak gound nfomaon abou e paula ba of work o be amped. Scic nformaon Te agr eed aepa ble qual y level AQL ). T dened a e w or qualy w an be ondered aepable a e average 'per en defeve or defe per 1 un of a proe AQL are agreed beween e uppler and eernal uomer or beween nernal uppler and uomer AQL range from 1 o 1 defe per 1 un and are gven n BS 61 Te npeon level. T allow ome laude n ample ze. Tere are r ee npe on level normal; ge ned; and re dued. An AQ L e borderlne bewe en aepable and unaepable qualy Wen an AQL a been agreed e deal uaon o ave all bae beer an e AQL aeped and all oe wore rejeed Sne mpoble AQL ue omeng alled an 'operang araer urve w ell you e rk a you are akng and effevely allow e level of rk o be oen o ree your knowledge of e urren uaon and wa mporan o you 1
•
Normal incion degned o proe e uppler agan a g level of rejeon n effe e uppler gven e bene of any doub. Te uomer reeve proe on from a e of rule a allow
MANAGEMEN MEHOS
• • •
21
wcng o gened npecon wen e uppler perfor mance own o ave become le afac ory. ighn incon make le lkely a a fauly bac wll be paed. Ruc incion occur wen e uppler performance a conenly been beer an e from AL agreed. S 6] gve rule for wcng one n pecon level o an oer and e conequence of .
3 Knowledge of weer defecve un or defec are beng npeced for.
acgroun nformaon 4 Wa e ze of e bac o be npeced? 5 e bac • • •
par of a ere of bace wc ave already been np eced? e r bac of a new ere? a 'one-off or olaed ac?
6 e mo advanageou meod of amplng •
•
• •
Singl amling a ngle ample of a gven ze aken and a dec on made o accep rejec, or ample e wole bac f more an a gven number of defec found. Doubl amling a ample aken of a maller ze an e ngle ample and en a decon aken o accep, rejec or ake a econd ample. Mulil amling an eenon of double amplng, allowng up o even ample o be aken before a decon made . Squnial amling no ed ample ze; e ample are accumu laed unl enoug nformaon avalable o make a decon
Benefits
AL and BS 6 gve a e of rue on accepance and rejecon of ample, baed upon e ample ze and number of defec found n e ample, a can be muually agreed beween e cuomer and uppler ereby reducn g dp ue. xample
An organzaon purcang conaner e crcal par of e conaner e nde damee r of e curve d corner. A pecca on a been drawn up for e radu of e corner and an AL of 5 per cen oude peccaon agreed w e uppler.
100 MEHODS FOR OAL QUALIY MANAGEMEN
22
ab 1 AQL smping inomion npcton v
o tt
Samp
G
ab
K
8
ccpt 2 5
Rjct 3 6
25
7
8
32
ccptanc ct a
Opeing eisi uves o smping pns G J nd
cmtanc
K K
G
Qalty o mtt atch tt than QL
5% chanc o jcton
25% chanc o jcton
.5% chanc o jcton
Qaty a a a 5% ctv
8% chanc o accptanc
78% chanc o accptanc
7% chanc o accptanc
Qaty a a a 1% ctv
37% chanc o accptanc
9% chanc o accptanc
7% chanc o accptanc
e uomer will be inpeing onainer o a defeive uni are o be inpeed e onainer i eier rig or wrong. e normal inpeion level a been eled e ba ize i onainer and e ba i par of a erie Referene o 61 give e informaion in able 1 e impliaion in erm of ampling rik of eleing o ne or oer of ee level an be found by a udy of e operaing araerii urve of plan G and K able ) mu be remembered a a lower inpeion level doe no imply a lower leve l of aepable qual iy bu an inreae in e ri k aoiaed wi e aepane of a ingle ba n a erie of ample we ave o ake ino aoun e fa a ome wole bae will be inpeed and rejeed 1 per en and i will affe e ougoing quali y i allow you o make an in formed judgemen abou wi ampling plan i be fo r your ue.
Rfrnc Gd s BS Smpg Pd
23
MANAGMNT MTHODS
Me th od 2
Ain ity di ag ram
urpse
o organize arge amount of data in group aording to ome form of natural anity When t use
When a team i trying to ae rtain utomer need with the intention of tranlati ng them into deig n requ irement. w t use
ah team member tart by wtg idea about utomer need on eparat e le ar d Then l aying the a rd on a table withou t onver ation to inuene them the team member hould arrange them into the natural group they an ide ntify. dea whih have an afnity for eah other hou ld be grouped together. A n eample of an afnity diagram i hown in Figure 1
Verbal data headn or B finity
eain for or
eadin B for or
ffinity
finity
Fgr Anity diagam
24
1 00 MT HODS FO R TOTAL QUALTY MANAG M NT
Beneits
Organzng data n te form of natural afnty can llustrate te assoca tons rater tan te strctly logcal connectons between customer needs. xample
1 Dene te subject tat s to be consdere d e g. custome r requrements for te product Arrange te data generated by te team under subordnate eadngs (see Fgure 1) • • • • •
Workng n slen ce arrange two c ards wc are related n some way Repeat ts step. Dfferent opnons about te relatonsp between dffe rent data wll be dscovered Complete te work wen all data ave been organzed acc ordn g to a lmted number of groups and dfferent opnons ave been resolved Fnd a eadng for eac data group.
Rfrnc B. Bman and B Klesjo (99 ew Yok MGaw-Hl
Qy: fom Csom Nds o Csom Ssfon
MANAGMNT MTHODS
25
Me thod 3 A ow dia gam urpse
o sow e me requred for solvng a problem and wc ems can be done n parallel When t use
For dao-da projec and producon plannng and o ensure e mos suable mescale for ceran ask s w t use
e arro w dagram s a smpl ed crcal pa meod of plann ng o sow e opmum scedule for fulllng a projec and rackng s progress Benefits
o ensure e mos suable me pla nning for ceran ask s
teror walls
ondaton ramework cafodng
nal lmbng nspecton
antng lectrcal work
-----
mont o tm e for each operaton eaton of work wthot tme for each operaton Fgre
rrow diagam
nteror walls
Completon
26
00 MTH ODS FOR TOTAL QUALTY MANAG M N T
xampe
1
Figure ow an arrow diagram ued to plan te contruction of a oue identifying • • •
amount o f time for eac operaton relation of work wtout time for eac operaton eac pecic operation.
MANAGEMENT METHODS
Me thod 4
27
Benc hma rking
urpse
o idenif and ll gaps in performance b puing in place bes pracce ereb esablis ing superior perfor mance When t use
As par of a oal quali process wen aking an ndependen look a perform ance b compa ring wi a of oers w t use
ere are ree dsinc pes of bencmarkng wc can be used b an organiza ion progr essivel o simulae e improvemen proc ess
1 nal bnchmaring s s e comparison beween funcions deparmens or a smilar organzaon as a means of improvng performance e usua am s o opi mize process perf ormance b e removal of errors 2 Comiiv bnchmaring is is a cross-comparison wiin one indusr secor aimed a esablsng bes pracice roug e iden caion of gaps beween o ur own and our compeors perf ormance. s can be done on produc funcional deparmenal or on a companwide bass 3 Comaraiv bnchmaring is s e comparison acros s al business secors aim ed a esablsi ng bes prac ice in all areas of ope raon . e rou e o bencmarking is as follo ws: 1 Dene e business misson e miion amn las down e business a ou are in and can be ranslaed ino roles, goas and obecives for eample: 'o be one of e worlds premier suppliers of parmaceuical producs dsincive and successful n wa we do. denif e crical success facors (CSs) a mus be acieved o full e misson ac CS mus be necessar and ogeer mus be sufcien o acieve e mission e CSs are e absolue minimum se of ams o acc ompls s ample s for a parmaceuic al compan mig be: • • •
ecnical ecellence n new produc s ecellen suppliers well-qualied saff
28
00 MHODS FOR O A QUAY MANAGM N •
sound gulaton
3 dntfy t ky businss pocsss to b undtakn to aciv t CSFs Pocsss dscib ow t wok tat is don in t busnss actually gts don. list sould b bot ncssay and sufcint to acv t CSFs. ampls migt b • • •
manag t gulato manag aw matal suppls manag patnt potct on
4 Flowcat t pocsss idntfyng any gaps dad nds o duplica tions. All intnal customs and supplis must b dntd 5 St standads fo t pocsss idntifyng t lvl of pfomanc tat must b mt s sould b as concs unambiguous and masuabl as possbl. 6 Dcd ow t pocss pfomanc can b masud 7 Plan ow to idntfy ow ots pfom t sam pocss A typcal fomat fo ong ts us t followng adngs • • • •
dscpton o pocss stp standad masu qustons
Benefits
Bnc mak ng povid s an ntoduction tt ida of masumnt , lps to focu s on t mission an d to idn tfy masus o ta gts fo ky businss pocss s Comp ans tat av pvously bn sy of m asumnt nd tat toug t ntoduction of bncmaking, it coms natually Bncm ak ngtnally lps oganzations to clos movto away bng ntospct towads bng focusd and tifom makts.
v
xample
A suppl to t moto ndusty usd bncmakng to dntfy ts ky businss pocsss. t tn dvidd ts into a numb of subpocsss and st up mpovmnt tams o owca t and st standads and masus fo t pocsss O n pocss idntid was nt oducng nw poduc ts. s bok down into nn sub-pocsss as follows • • • • •
dvloping a custom quimnt stmatng costs autoizng nw poducts dntfyng suppls dvlopng nw toolng
MANAGEMEN MEHODS • • • •
29
purcasing new supples developing a bill of maerials wring new specicaion s developng a new producion plan
e eam drew ouline owcars for eac of e sub-processe s and en began o draw deailed owcars quickly became apparen a all produc designs were beng reaed equally weer imporan o e company or no. Te consequence was a minor designs were being forced roug e oal process and were makng i difcul for maor designs o progress. Te amoun of variabiliy i n e process made i very difcul o esmae imes and keep o scedules. Te organza on agreed on wo separae proc esses depending upon e imporance of e desgn and se sandards on process seps n erms of bo me and compleen ess Havng fo rmalzed eir own process es ey en began o compare em w oers ouside e organizaon.
Rrnc RC Camp 989) Bhmak h Sah fo Ids Bs Pas ha Lad o So Pfoma Mlakee ASQC Pre . B. Kaf ad S blm 1994) Bhmak a Sos o Ex Qa ad Pod Ne Yk Wiey.
30
00 MTHODS FOR TOTAL QUALTY MANAGMNT
Method
5
Consensu s e achin g
urpse
To give a team a methodial way of examining alternatives to reah a olletive onlusion whih all team members an aept When use
When a team is examining different ou rses of ation or hoosing possible solutions to a problem w use
There are six steps involved
1 All team members think individually what the options are and eah makes a list of his or her own ideas
2 Partiipants are invited in turn to read out their suggestions These are reorded on a ip hart. Partiipants are not allowed to disuss the ideas at this stage 3 After everyone has listed their suggestion s group members are all owed to add to the list any new ideas as they think of them The group leader then heks that all suggestions are understood and seeks lariation i f neessary 5 The ideas are ounted. Everyone individually assigns them points f are 12now ideas starttheir with own point s for the mo are st import ant. on a 12 top Partiipants reveal hoies whih reoded 6 there ip hart. The disussion then starts and the ideas that the maority hoose are debated until onsensus is reahed f neessary the proed ure is repeated with a redued number of ideas Benefis
People often ling to their own ideas and are relutant to onsider the suggestions of others when making a deision The tehnique permits a deision to be made without taking a nal vote that an leave some team mem bers feeling isolated. The proe dure allows a group to reah the best solution not a ompro mise soluti on and to harness al membe rs ommitmen to making the solution work The tehnique works best in groups of 812 people.
MANAGMN MHODS
3
ample
An organization was in a serious asow ris is and te mana gemen t team was seeking ways rst to stop f urter damage and seondly reverse te situation Te problem was ompounded by different professional and personal interests te witin te group. Consensus reaing was used as were a means of getting groups agreemen t to a series of measures wi unpalatabe but neessary.
32
00 MTHODS FOR TO TA QUA TY MANAG M N T
Method 6 Contingency pann in g urpse
To avoid reghting and waste of reso ures by pl anning for ontingenies in the ompletion of a projet When t use
When planning to implement a projet w t use
There are six key steps involved 1 List in a logial sequene the steps to be taken to ahieve suessful ompletion of the pr ojet. 2 Examine eah step and deide whih are the most ritial areas and where it is most likely that a problem will our These are for example where deadlines are tight where other departments are involved or where approval is needed Use brainstorming to disover the likely ause of these problems. Remember that you are looking at the serious problems not all problems 4 When you hav e identi ed the potential li kely auses list them learly 5 Deide where preventive ation an be taken: how an the problem b removed or minimized? 6 Modify the srcinal plan to take the preventive ation into aount Benefits
By taki ng into aount possible problems bef ore the y ari se the i mplemen tation of plans an be made smoother and more ertain xample
An organiation is planning to register a trag ourse before a set deadline and is drawing up a ontingeny plan n order to be suessful with the projet the projet manager examined the most ritial areas where problems might our ie tight deadlines involvement of other departments where appraisals are needed et He then modied his plan for suessful registration of the training ourse
MANAGMNT MTHODS
Method
7
33
Costbenefit analysis
urpe
o eime he el o nd bene o poje nde onideion. When t ue
A poblem-oling em wold e o-bene nlyi o nd o i olion i pil in em o o w t ue
hi imple ehniqe inole eling ll he o wih ed implemening pil poje n d omping hemoied wih he epe bene he elion lly oe hee o ey e pe iod. he ehniqe n lo be ed wiho l o ge b ing weighing . hi i pilly pplible when deling wih ognizi onl o hmn poblem. Beneit
Some olion o poblem e e lie ly ey o el e b oen e h o be mde o implemening pil olion. Co-bene nlyi enble em o mke e o olion being doped xample
A poblem-o lin g gop hd dioeed h ke y pnh eo w he mjo e o inoe d eny hey ideni ed io wy o ping hi igh inding he phe o new piee o hdwe oing £5, he o-bene nlyi looked like hi:
Co
Mhine Re-wiing nd inllion
£ 5, ,5
Co o e-ining Lo ime o
, ,
ol o
8,
34
00 METHODS FOR TO TA L QUAL TY MANA GE ME NT
nt ar
Rework redued by 2% Redued reoniliation osts
375 25
Total ost
25
Rework redued by 2% Redued reoniliation osts
nt ar
375 25
Total ost
25
Comparing the osts and benets over two years shows
Year Year 2 Tota
Cot 87 87
nt 25 25 25
Prot 245 25 38
n two years the new equipment will re-pay the ost of purhase and generate additional savings.
MANAGMNT MHODS
35
Method 8 Criteria testi ng urpse
To evaluate and ompae altenative solutions to against a list of iteia
a poblem by ating them
When t use
When you need to deide whih ideas to hoose fom a shotlist. The tehniqu e all ows ideas o solu tion s to be anked so that ompaisons a n be made
w t use
Thee ae ve steps involved.
1 State the iteia equiements of a goo d s olution . 2 List the iteia in ode of impotane then plae them aoss the top of a ip hat.
3 List the alte nativ e solutio ns o ideas down the leftha nd side of the ip hat When this is omplete, fom a mati.
4 o eah option in tun estimate how well it meets eah iteion. A sale of 1 an be used elates to the best lowest ost fastest
S
souion e f a gou of eoe is invoved, as eah eson to estimate individually and take the aveage soe of the goup. o eah option, add togethe the soes The pefeed solution is the one with the lowest total soe
If, duing step 4 the soes ae vey diffeent stop and hek that eveyone in the goup has the same undestanding of the iteia t might emege that someone has a hidden agenda and is ating aodingly. Altenatively new infomation might aise at this stage. An altenative to using a -point sale is to ank the options on eah iteion the best hoie soing et The seleted option is again the one with the lowest total soe
1
Benefits
Simply voting on ideas o foing a weighting system an be vey divisive. Citeia testing allows a goup to make a deision with a ommon set of
36
00 MTHO DS FOR T OTA QUA TY MANAGM NT
ptns By addtnal Cntract t Change wrkng practice nthng
Fg
nn me pprval Cst asy ard w ast 0 gh 0 lw 0 ard 0 asy
tal
0
0
0
2
iteia testing
aumption. Done thi way, the commitment to the olution will be greater. xample
A company had a problem with low utilization of dektop publihing equipment and examine d alternative way of increaing throug hput igure 1 how a completed criteria teting matr for the four dented opton. he option elected wa to change wor king pract ice to all ow thoe carrying out the tak more ay in the cheduling of work.
MANAGMNT MTHODS
Method 9
3
Customers' contin gen cy tabe
urpose
To undesand he needs of boh inenal and exenal usomes fo he fullmen of usome saisfaion When to use
When a eam is yi ng o lean he level of usome saisfa ion of bo h inenal and exenal usomes ailue o saisfy he usome may pu he ompany ou of business ow to use
Lis boh you inenal and exenal usomes and hei needs ie. wha hey equie wha hey expe and wha would exeed hei ex peaions This an be genea ed wih he help of bains oming, min d mapping o a ombinaion of boh
xtena C ntenal
V C bove ust xpectation expectation asy to ode 2hou delivey eceive dded vaue poduct ode ee shipping elpful service eceive coect infomation
asy to deal with the pocess o be teated like a valuable custome
Fg
ustomes' ontingeny table
ull coopeation
38
00 MHODS FOR O AL QUALY MANAG M N
Benefits
t helps you to understand the assoaton between your ustomers and ther level of requrements xample
A ustomers ontngeny table s shown n gure 1 Here mu s the minimum ustomer requrement. alng n ths wll ause dssatsfaton.
39
MANA GE ME NT METHODS
Method 10
Deming wheel (PDCA)
urpe
A management onept suggested b Demng to satis te qualit requirements o te ustomer b usng te le pan do ek and ation. When t ue
For te development o a new ustomer
produt based on t e requirements o te
w t ue
Develop teamwork between te ompans various untions ie. produt deveopment m anuatu ring sales and market re sear. se te plan (P) do (D) ek (C) and ation (A) le as sown n Figure 1 Disuss detais o ea stage o te le Te le or weel mus t be onstantl rotating
Plan (Podct development)
ction (aket research)
o (anfactre)
Check (ales) Fgre 1
Deming weel
40
00 MT HOD S FOR TOTA QUATY MANAG M NT
Plan When a probl em is deteted in produt develo pmen t, rst nd the auses of the problem The deision to make hanges must be based on fats, and the refore it is neessary to obtai n the data to detet the auses of error and variation 2 Do When the ause of a problem is deteted a quality improvement team willproblem take responsibili ty for arrying thro ugh the neessar y steps to solve the 3 Chc When proper steps have been taken to solve the problem an investigation will take plae to hek whether the improvement proess was suessful 4 Action f the step s taken were suessful the n ew and bette r quality level should be aept ed f the steps taken were not su essful then the PDCA yle should be repeated Benefis
t will help to ensure that the ustomer is always fully satised. xample
A ompany implementing the M proess used the Deming wheel for ahieving ontinuous improvement of the various business proesses in order to develop quality improvement of the whole organization
Rfrnc WE Dm
( \ 98) O ii Cmd Mchu MT P
MANAGMNT MTHODS
4
Me thod 1 1 Departmen tal pu rpo se analysis (DPA) urpse
To review the internal ustome rsupplier relationship When t use
When you want to understand the nature and ost of quality n the department ie. to improve interdepartmental quality w t use
Ask pertinent questions about the department suh as What is the role of the department? Does my boss agree? 3 Why is this department doing this ativity? 4 s it being done the way the 'ustomer department would want it? 5 What i mpat does the ativity have on the prime objetive of satisfying the requirements of the eternal ustomer? Beneits
By establi shin g the above informatio n both within departm ents and a ross departmental barriers DPA helps managers address improvements in interdepartmental quality for the bene t of the org anizati on xamples
Eamples of key tasks and ativities involved in DA are as follows
K ta • • • • • • •
oordinate total quality manage ompany quality system imple ment vendor assessment system maintain ustomer retur ns system maintain srap reporting system maintain nished produt audit system maintain quarantine store
42
00 METHODS FOR TOTAL QUALTY MANAGEMENT
ppliers
Fgre 1
eqirements eqirements rganization • ission Cstomers • esponsibilty • ctivity npt tpt eedback Voice of the Cstomer easrement
Depatmental purpose analyss
Aciviie • •
updte projet pn orgnize Bord steering group
• • • • •
mnge onsultnt interfe identify trining (qulity) needs gree mesures gree progres s reporting system orgnize triner trining
Deprtmentl purpose nlysis fouses on the ustomer When the requirements of the ustomer re seen s the responsibility of the deprtment n effetive mesurement system n be dopted (s given in igure 1
Reference D B P. MBrd ad G. Wil Hma.
994) Qy Ld Buttth
MANAGMNT MTHODS
Method 12
43
Error proofing (pokayoke)
urpse
To design an operation in suh a way that spei errors are prevented from ausing major problems to the ustomer When t use
t an be used when defets our and require immediate feedbak and ation at the • soure of raw materi als an d omponents • start of the prodution proess • prodution points where an error may our
1 per ent inspetion
w t use
n a pokayoke system arry out automati ontinuous inspetion and if abnormalities our then provide feedbak and take ation However before proessing an begin one has to halt the proess obtain feedbak and exeute neessary ation Benefits
nlike statistial qualit y ontro systems in whih a fairly long time el apses between the 'hek stage and the exeution of feedbak and ation pokayoke minim izes defets by ar rying out feedbak and ation imm ediately at a low ost xamples
A large stee press is automatially monitored for wear f the thikness beomes less than a speied amount an al arm sounds and ation has to be taken to retify the error A seond example involves a ar manufaturer whih was onerned to prevent omission of ar door pokets in prodution The operation in whih pokets were mounted on the door trim involved three speia tions and workers oasionally negleted to mount the pokets or mounted them inorr etl y aordin g to their attenti on to work in pro gress To improve the proess door trims were exposed to a detetor whih reognized whether pokets were missing f right and left pokets had been reversed or if a poket had not been mounted a buzzer sounded air stopped owing to the srew-tightening lok and the operation oud not proeed n this way instanes of poket omission were eliminated
44
1 00 M HOD S FOR O AL QUALY MANAG M N
Method 1 3
orc e a nalysis
urpose
To identify etenal and intenal foes at wok wen developing a ontingeny plan When o use
Use tis metod befoe developing a ist of potential poblems fo yor poblem peventon plan ow o use
1 Daw a ile. 2 Wte down te etenal foes tying to keep te plan fom ong
otsde te ile
eam wok
0
Figure 1 Fore analysis
MANAGEMENT METHODS
3 Write down the interna l fores trying to aomplish the plan irle 4 Denote the strength of the fore by the length of an arrow.
inside the
Benefits
Powerful me thod for helpi ng to develop a ontin geny plan b y providing a graphi illustration of external and i nternal proesses at wor k. xample
External and internal fores for developing a business plan are shown in igure 1 Here fore analysis he lps to id entify the for es whih are helping you to aomplis h your purpose and those whih are resis ting your e fforts n this situation ustomer requirement an be met by management ommitment. However external fores suh as modern tehnology and internal fores suh as ultural hange might resist the development of the business plan.
100 METHODS FOR TOTAL QUAITY MAN
46
Method 1 4
AGEMENT
Ga nnt chart s
Purpose
For planning the steps necessary to implem ent quality improvement When to use
When a team ha s deci ded upon a project and is pannin g its implementation, Gannt charts are usefu l for scheduling the events necessary to complete the improvement activity How to use
Gannt charts are very easy to use 1 2 3
4
Break down the implementation plan into achievable tasks and activities Estimate how long each task will take and then se t a realistic completion date. Break down the step s into a logical sequ ence Line s denote wh en a task is due to commence and end. The relationship over time between each task is immediat ely visible. Assess each step individualy, identifying: •
•
sa any issue that stops you completing a stated task (this is noted a key issue) any dependent task that must be completed before another task is begun. onths to complete
Task o. escription 1 2 3 4 5 6 7 8 9
1
2 3 4 5 6
7 8 9 10 11 2
Identi fy aea for improvement efine project anagement revews Taining povided ata analysis Formulate hypothesis Identify causes uggest quality solutons Implement quality soutions
Figure 1
Gannt chart for the implementation ofa quality improvement project
MANAGMNT MTHODS
4
Benefits
The viual epeentaton of tak help identify the key iue and bing into the open the tep needed fo ucce ful completion t alo make it eay to e e when de adine ip and change to the plan have to be made xample
n the exampe h own in igue 1 an oganization i uing a Gannt chat to plan the impementa tion of a quai ty impovement po ject The timecae i given in month t can be een that tak 3 and 4 cannot be defeed without delay to the total plan Tak 5 i a key iue in tha t, without data analyi, the quality impovement poce cannot be implemented
4
00 MTH ODS FOR O AL QUALY MANAG M N
Method
5
ISO 9000
urpse
To demonstate to youself you ustomes and an independent assess ment body that you have an effetive quality management system in plae When t use
When t ying to fomalie opeat ions to ensue onsisteny of appoah w t use
Thee ae steps as soiate d wth bulding a quality management system MS) 1 Obtain management undestanding of and ommitment to the quality management appoah. Dene the sope of the ativities to be inluded in the MS 3 Dene the oganizational stutue and esponsibilities of those within the sope of the MS 4 Audi t the exis ting systems and poedues against the equi ements of the standad S Develop a plan to wite the neessay p oedues Tain sufient pesonnel to wite thei own poedues. 7 Daft and edit the poedues and gain ageement to them 8 Compile a daft quality manual. 9 mplement the system on a tial basis 1 Tain intenal auditos to ay out audits of the system and its opeation. 1 1 Revise the opeation of the system in light of the esults of audits and othe infomation 1 Apply fo egistation (sometimes alled thid-paty appoval) fom an aedited body 13 Maintain th e syst em by intena l audit, usin g it a s an oppot unit y to impove. Benefit
By establishing a onsstent appoah it beomes easie to ensue that tasks ae aied out in the same way whoeve does them. This in tun ensues onsistent quality fo ustomes
MNGEM ENT METHODS
49
xample
A QMS normally ontans four levels of doumentat on . These are: A qualiy licy manual ths s a poly doument a statement of ntent about wha you ntend to do 2 A qualiy rcedure manual ths s a more detaled seres of dou ments detalng hw you wll arry out the 'whats of the qualty poly manual. 3 Qualiy recrd these are the proofs that the 'hows have been arred out. 4 Wr inrucin these are the small detals that explan how spe tasks are arred out Eah proedure may have several work nstru tons for dfferent produ ts or serv es
Reference BS EN SO 9000 Son nd Us. BS N SO 900 QA Dsin.
50
00 METOS FOR TOTL QLTY MNGEMENT
Me thod 6 Jus t i n time (JIT) urpse
To delive the aw matei al o component to the poduction lin e to aive jut in time when they ae needed. When t use
When you want to minimize o eliminate tock in ode to pevent the oganization fom incuing unpoductive cot w t use
The podu ction poce ue the pull ytem wheeby the mateial i not eceived fom the upplie o equeted fom the peceding poduction phae unti l it i needed to utain podu ction . Beneits
t help to eliminate tock in ode to pevent unpoductive tock cot. The benet to fou companie of uing J T can be een in Table 1 xample
JT wa deve loped by Toyota Moto Compan y n 984 Toyota ageed to eopen Geneal Moto old plant in Califonia a a join t ventue A new tamping plant wa built next to GM old plant o that vaiou tamping could be poduc ed in m all batche when neede d Thee component had peviouly been made atonothe factoie which meant aembly wa dependent lage GM amount of component beingthat ent ca by tain fom othe factoie. The adoption of Toyota poduction ytem (JT in GM facto y alo meant that thee two ytem wee inepaable an d poduc tivity and quality at th i old facto y wee accep table ab Bee o JT Rc invntoy (% )
Ruc a tims ( % )
Ruc wok (% )
Rc pac (% )
A B
94 8 75
5 9
50 5 7
40 70 58
94
70
75
40
ompy
MANAGEMENT METHODS
Method
7
Kaizen
urpse
A Japanese term meanng 'change for the better the concept mples a contnuous mprovement n all company functons at all levels When t use
The term s so common n Japan that t s used n all aspects of lfe w t use
The aizen concept s based on peoples
commtment and partc paton by
usng ther knowledge experence schemes and can ttherefore be establshed through qualty crcles and and suggeston can be used n both manufacturng and busness actvtes Benefits
Gven the same basc technology aizen can help to lead to a hgher productvty r ate and h gh-qualty products. xample
Dahlgaard et al 199 surveyed companes n Japan, Korea and Den mark and found that the number of companes wth qualty crcles was about 97 per cent n Japan and Korea but onl y 2 per cen t n Denmar k. About 78 per cent of emplo yees n Japan and Korea too k part n qualt y crcles, but only 1 per c ent n Denmark Suggeston schemes wth or wthout a reward system can be found n more tha n 95 per cent of compan es n Japa n and Korea The corre spond ng gure for Denmark s only 45 per cent n Japan aizen actvtes are n addton t o mantenance of the proc ess and they can therefore mprove the qualty of servce and products wth lmted nvestment.
Reference J J Dh G K Kj K Krst s 99 A compr tiv study o qulity co tol mthos pcipls i Jp Ko Dmk l Ql Mngmn . M. mi (1986 Kzn: Ky Jpn mp Sw York Rdom Hous.
52
00 MHODS FOR OAL QUALIY MANAGMN
Me thod 1 8
My st er y shopping
urpse
A tehnque nvolvng lookng at your busness from the outsde and measurng the efeny of your own key proesses from the ustomers vewpont When t use
When measurng ustomer satsfaton Mystery shoppng s often used durng be nhmarkng eers es or as part of a motvaton progra mme
w t us e
An org anzaton rst develop s an unders tandng of ts own key proess es and behavours and then delberately makes a omparson wth the ompetton Alternatvel y a ompany an o mpare the performane of ts own staff Beneits
Many employee of the month shemes are based on 't must be Daves turn rather than on any objetve measure Mystery shoppn g bestows an understood and gves redbltymeasure to suh to shemes etern ors al ly mysterymeasure shoppng gves a rsthand omp Used are ompett wth your ow n perfor mane . xamples
An advertsng ompany dedes that one of ts key proesses s takng a bref from a lent The ompany develops a set of measures that t wshes to use These measures ould be • •
speed of nt a response understandng of bref
• •
speed to bref qualtyofofreply respon se The mystery shoppng s then done n several smple steps:
MANAGEMENT METHODS
2 3 4 5
53
The company usng an outsde agent, bulds up a typcal bref from a clent The company decdes whch compettors to nclude n the study The outsde agent approaches both the company and the compettors and presentng t as though t were a real bref asks for the same response from each Each company s rated aganst all the measure The results are reported and an acton plan agreed
A second example concerns a motor trade dstrbutor who wshes to compare the perfor manc e of ts own sale s force across al l ts ou tlets as part of a motvaton programme The company rst develops a set of successful sellng behavours that t expects ts sales team to exhbt These could be • • • •
welcome on enter ng the showroom understandng customer needs sendng nformaton as requested follow-up telephone call
The company ng a outsde agentteam vsts of ts own the showrooms posng as a real uscustomer The sales areeach rated aganst agreed crtera and the results used to plan tranng and to recognze desred behavour
Reerence Sarah Ck 1992) omer r Ne York gan Page
1 00 MHODS FOR OTA QUAITY MANAGMN
54
Me thod 19
Obje ct iv e ra nki ng
urpse
Helps to ple your urrent tvty in perspetive nd enble you to understnd the purpose of your efforts When t use
When you wnt to know where to strt w t use
Collet list of potentil objetives Write down eh obetve on rd 3 Eliminte ny objetive whih s not pproprite
ncreased market share bjective to be achieved
liminate customer complaints
denti the customer Fige 1
Objetive ranking
ncrease maket share
elight the customer
mprove customer care
Povide the best seice
MANAGEMENT METODS
55
Rank obetive l ist aording to highest or lowest purpose.
5 Chek the list with ustomer needs. ind the mnmum ation that ould be used to satsfy ustomer requirements.
7 Pk the optimum obetive whh an be ahieved with the available 8 resoures. Pik up the obetive from the list that meets ustomer requirements and an be ahieved by the group. Benefits
It gves a wider understanding of varous ativties step into a proess of produtive hange
when transforming eah
xample
Suppose that the obetive to be aomplished is to delight the ustomer. Obetve rankng should onnet one obetive to another supportin g the next higher one as shown n gure 1 .
56
1 00 METHOS FOR TOTAL QUALITY MANAGEMENT
Method 20
Pa re to a naysis
urpse
To seprt e the most import nt uses o f probem fro m the m ny trivi Aso to identify the most importnt probems for tem to work on Preto nysis ws rst use d by Wi fredo Pr eto n tin eo nomist When t use
When tem is nysing dt reting to probem to deide whih re the most importnt fto rs to be tked rst to hve the most impt on the probem.
able 1 Pareo aaly Errr deripi
Errr de
Wrg mpe Miig mpe N ldered her lderPB prblem Wrgly iered Shr-lder ide Elerial dee Wrg plaeme
33
Shr aed by mpe Te eqipme . . N rrely iered PB al Elerial lerae Mehaial dee SM al igm e Shr-mpe ide Prri legh me prgr al Addiial mpe SM mbeee Miig label ir Brke lder ji
56 94
Ball-haped lder ji Cmpe damaged Tal
51 7 92 54 77
2 53
al
Cml
33 254 3 2 1307 1 14 2 31 722 67
22.95 1.59 101 956 .35 60 52 503
2295 41 54 5165 61 .2 6956 75.64 0.92 5.95
649 523 237 17 3 7
90.70 9452 9625 9756 9.6 9.73 9917 99.54 997 99.5 99.90 9995 99.97 99.9 99.99 99.99 10000
C
95
51 23 19 6 3
22 76
1
475 33 .73 1 .30 0.6 057 043 037 07 014 006 004 002 0.01 00
2
0
000 0.01
1 3 67
10 0.00
6 9 15 11 5 7 46
3 52 59
41
59
% 00 0 0
.
.. ..
• • •••
• • . •
% of total
0
0 0
• . . . . . . . . . . . . . . . . . . . . . . . . . . • • • • • • ••••
Cmlative %
"
"" """ 0 "" "" 20
l Z l Z I o
0
0 rror code Fgure 1
areo dagram
-
58
00 METHODS FOR TOAL QUALITY MANAGEMENT
w t use
Pareto analyss s sometmes alle d te 0/ 0 rule . Ts mea ns tat 0 per ent of te problems are ause d by 0 per ent of te atvtes and t s ts mportant 0 per ent tat soul d be onentrated on Tere are s smple steps nvolved: 1 Lst te atvtes or auses n a table and ount te number of tmes ea ours. Plae tese n desendng order of magntude n te table 3 Calulate te total for te wole lst Calulate te perentage of te total tat ea ause represents. 5 Draw a Pareto dagram wt te vertal as sowng te perentage and te orzontal a s te atvty or ause Te umulat ve urve an be drawn to sow te umulatve perentage from a auses nterp ret t e results Benefits
Wen workng n teams t an be dfult to rea agre ement w en people wt dfferent opnons want to follow dfferen t ourses o f aton. Pareto ana lyss brngs te fa ts to te attenton of all members of te team to ad deson-makng ample
Followng manufature of a prnted rut board te board s tested to dentfy any faults Table 1 sows te error desrpton error ode and ount of te number of errors Te perentage of te total and te umulatve perentage s also gven Fgure 1 gves te error ode on te orzontal as and te perentage on te vertal as. Te umulatve perentage urve s sown as a dotted l ne Te Pareto dagram sows learly t at s out of 5 error types ( per ent) aount for nearly 0 per ent of te total number of errors Te errors tat must be redu ed to ave a maor mpat on te overal s tuaton are learly sown. t would also be possble to draw a Pareto dagram sowng te ost of errors or te mportane of dfferent errors f tese fators were more mportant tan smple ourrene aone.
Rfrnc P. Ha. Spny 992)
Word a Prforman rog Toa QaLndn Chapman and
MANAGEME NT MEHO DS
Method 21
59
Pot entia l prob lem a na lysis (PPA)
urpse
To examine plans to identify what an go wrong with them so that preventive ation an be taken When t use
Potential problem analysi s (PPA ) is used when pla ns are rst drawn up and subsequently at planning reviews to antiipate future problems and pan ontingeny ations. w t use
PPA is a very simple tehnique for a team to examine pans There are eight steps 1
Draw up the plan in time order.
2 lowhart the plan identifying all the key steps where spei outputs are needed
3 At eah key step brainstorm the problems that ould our. Rate the potential problems using t he folowing sheme • •
5
7
Lielihd 10 (very likely) to 1 (very unlikely) Severiy 10 (atastrophi) to 1 (mild)
Multiply the likelihood by the severity to get the potentia problem risk (PPR) number or eah pote ntia l probl em ide ntify the li ke ly auses This is done rst for problems with a high PPR numb er. Al l problems with a PPR abov e 50 or a li keli hood or severit y above 7 must be prevented or eah ause brainsto rm the ourse s of ation that ould be tak en to prevent it happening. or probems th at an be prevented tak e the neessary steps to remov e the potentia ause or problem s that annot be prevented draw up ontingen y pans to retify t he problem if it our s
Beneits
By antiipating problems with plans before they our and either remov ing them by prevention or drawing up ontingeny plans, smooth imple mentation of projet s an b e ahieved.
60
00 METHODS FOR TOTA QUAITY MAAGEMET
xample
An organization was planning to hold its AGM in a Central London loation and used PP A to assist with the pl annin g of the event to ensre that no problems oured
MANAGEMENT METHODS
Method 22
61
Problem prevention plan
Purpose
To anticipate what can go wrong and plan to prevent problems. When to use
When you need to aalyse potential p roblems and their causes. How to use
1 2 3 4
S
For each specic task write down a list of potential problems For each problem, rte the chances of it happening and the seriousness if it did happ en . For most serious probems write down the potential causes and chance of occ urrence For most lik ely cause s, decide wh at prevention strategies ca n be put in place Integrate prevention strategies into your plan.
Benefits
It helps preven t potential problem s from a ctually ta king place
Potential problem
Chances igh edium ow
Person in charge il Computer out of order
X
X
X
Figure 1
Problem prevention plan
X
X
adly produced document Program bug
eriousness igh edium Low
X
X
62
00 METHODS OR TOTAL QALITY MANAGEMENT
xampe
Potenti pobems in delivey nd ode doumenttion e ted in igue 1 Hee the like lihood of se ious poble lies in bdly podu ed doumen t Tone oud be use of this poble m Hene egul hek of the tone wi l be p of the poble m pevention p ln
MANAGEMENT METHODS
Method 23
63
Process decision programme chart
urpse
To focus on possible sequences to help lead to a desirable resut and cont ingen cy p lanning . When t use
When designing a new pan to achieve a desired resu t, and to avoid certain undesired outcomes. w t use
A process decision programme chart can be used to design a plan to achieve your desired objective to deal with problems encountered while imple mentin g the pan and to mak e correct decisions to enhance the plan thereby achieving the objective. n addition, the chart can be used to conceiv e counter- measures to avoid an undesirabe situatio n b y simulating a process of events leading to an undesirable result. Benefits
t helps to plan to obtain a desirabe outcome xample
The proce ss decision programme chart given i n igure 1 s hows the proce ss which can hep to secure a contract
64
1 00 METHODS FOR TOTAL QUALITY MANAGEMENT
rder request from a colege
Could not fix a date
rder lost
Price not competitie
Price competitie Bid not agreed
rder lost Fgu 1
Process decso programme char
rder secure
MANAGEMEN MEHODS
65
Me thod 24 rogram me evau ation an d re vi ew (ER) te chnique urpse
To estabsh a pannng technque or compex an mut-eve projects When t use
hen you nee to meet a eane to compete a compex task w t use
Descrbe the task to be compete Estabsh the prevous task to be perorme n orer to compete he present task 3 ork backwars rom the competon ate (see Fgure 1), ncatng the task wthn a crce Present the task whch precee t n crces an ncate the requre competon ate S Draw an arrow rom the prevous task to the present task
ach
ask to be comp leted Fgue 1
PER echnque
ach
2 ay
66
METHODS OR OA QUALTY MANAGEMEN
Contnue untl all the tasks have been taken nto account 7 evew the agram an establsh the crtcal path whch requres the most amount o tme to compet e the pr oject See also Critical ath analyi (Meth o 33) . Beneits
It heps to match elvery an orer ocumentaton to complete the scheul e or a project xample
Fgure 1 ma tches elve ry an other ocumentaton by PE pro cess
Reference R Ands Sny Sc. ND.. Yk Wst and T.A. Wiiams
(994 A Irodo o Mgm
MANAGMNT MTHODS
Method 25
67
Qua lity circles
urpse
A specal type o small group actvty wc orms a vecle or te evelopment o nvuals When t use
Qualty crcles are especally useul n te later stages o a total qualty process wen n vuals n te r own work areas begn to tackle ter own problems. Tey lea to sel-regulaton n work groups w t use
Qualty crcles are small groups o bet ween tree an 1 people wo o te same or sml ar work, voluntarly meetng togeter regularly or about an our per week n pa tme. Usually uner te leaersp o ter own superv sor or manager, tey are trane to enty, analys e an solve some o te problems n ter own work, werever possble mplementng te solutons temselves Crcle leaers an members soul be trane n te ollowng tec nques • • •
branstormng tally carts concentraton agrams
• • • •
Pareto analyss stograms cause an eect analyss control carts
Qualty crcles are very erent rom acton teams an task orces, wc are ntate by management to solve a specc problem an are sbane wen tat problem as been solve. Qualty crcles are orme an trane tey ten enty ter own p roblem s en tose problem s ave been solve, te crcle remans n place an entes urter problems to solve Benefits
Qualty crcles gve str ucture an ocus to mprovement an allow t to occur n a planne way, wereas smply askng or suggestons coul generate too many problems wt no means to anle tem.
68
100 METHODS FOR TOTA QATY MANAGEMENT
Thee ae anges n usng qualty ccles too ealy n the pocess they may be seen as anothe management a o as manages seekng to abogate esponsblty xampe
A company n the Pottees set up quaty ccles t o allo w all employees to contbute the own ea s to the ben et o the whole company In t he st yea moe than 5 ccle pesentatons wee mae an successully mplemente
Reference Dad Htchns (985 Qua Cirle Hadbook London: Pitman
MANAGEMENT METHODS
Method 26
69
Quaity function depoyment (QF)
urpose
A echnque or scpne or opmzng he process o eveopng an proucng new proucs on he bass o cusomer nee When to use
Durng esgn commssonng or pos-commssnng o ransae cus omer requremens no company requremens The echnque can be use n research prouc eveopmen engneerng manuacurng markeng an srbuon reas. ow to use
QFD s a vesage process ha akes a esgn rom cusomer requremens no a pan an scheue (se e Bosser 199 ; Day 1993) . 1 The rs sage s en yng cusomer nees or wans These are usuay characerscs recy arbuabe o he prouc or servce such as wha ooks ke how ees how ong ass how compa res wh he compeon. A hs sage he requremens are no usuay measurabe These requr emens ar e hen ransae no echnca speccaons hrough he use o echnca expers A hs sage he requremens become measurabe 3 The echnca speccaons a re hen urne no en-prouc specca ons These are cae 'crca par characerscs Taken ogeher hey are boh necessary an su ce n o ea o an en prouc meen g he cusomer specc aon. Th e ourh sage s o esgn he process o ever he prouc or servce In oher wors ece how o urn he esgn no reay. 5 The na sage s o pan he acv es necessary o pro uce he requre oupu Fgure 1 usraes a ypca QFD marx oen cae he 'House o Quay se oThe he cusomer requremens as escrbe The n (1)eabove opmarx o hecomprses marx shows he organzaons requremens as escrbe n (2) The rgh se o he marx gves he pannng conseraons o prouce he oupu. The roo o he house
00 METODS FOR TOTAL QUALTY MAAGEMET
0
nterrelationship between technical descriptors
Customer requirements (voice o the customer)
Technica l descripto rs (voice o the company)
Proritized customer requirements mportance Competitive analysis arket potential
elationship between requirements and descriptors
Fure
1
Prioritized technica descriptor s
F : H
shows th ourth stag whr csons ar ma to turn th sgn nto raty. h cntra part o th matrx comprss th ratonshp btwn customr an sgn rqurmnts hs s whr th rqurmnts bcom a ns spccaton It s usua to us th oowng schma to show ratonshps n th matrx
[ D
Som ratonshp Strong ratonshp ak ratonshp No ratonshp
Benefits
Quaty ts uncton poymnt ncourags organzatons to stabshng ocus on th procss rathr than ust on th prouct or srvc By corratons btwn what s want an how t s to b vr, th vta aspcts bcom mor vsb, ang csonmakng
MANAMN MHODS
7
xample
The moel shown n Fgure 1 gves only the basc elements o QF. The House o Qualty as the process o ea generaton to evelop new proucts on the bass o customer nee
72
00 METHODS FOR TOTA QALTY MANAGEMENT
Method 27
Re ation di ag ra m
urpse
o iustrat th rationship btwn probms an ias in compx situations. Aso to intiy maningu catgoris rom a mass o ias whn rationships ar icut to trmin When t use
hn a topic is so compicat that rations btwn irnt ias cannot b stabish through convntiona rasoning an th probm in qustion is xcusivy a symptom o a mor unamnta unrying probm w t use
h vopmnt o th ration iagram shou b conuct in tams. h tam writs ach ia in a circ an cust rs th circs in proximity to ach othr. It thn intis which ia strongy inuncs anothr an uss arrows to inicat th irction o inu nc h rsuts ar vauat by intiying ias that hav th most arrows ntring or xiting Custo mer focus
esponsblty not cea
Fgu 1
Relation diagram
MANAGEMENT METHODS
73
Benefits
It entes relatonshp between problems an eas n complex stuatons xample
Fgure 1 presents the results o a team branstormng sesson whch ente ten major ssues nvolve n evelopng an organzatons qualty plan.
74
00 METHODS FOR TOTA QATY MANAGEMENT
Method 28
Teamwo rk
urpo
To oganze actvty wh ch eues a numbe o people to coll aboate an wok togethe o a comm on goa \ Whn to u
hen people wth eent but complementay esponsbltes know lege o abltes can contbute to the evelopment o a stategy o soluton to a poblem mpotant to the company. ow to u
Teamwo k eu es the coopeatve e ot an contbuton o a numb e o people wth a common goal o objectve. Bnfit
It helps to evelop a ualty stategy o to acheve solutons to a poblem vtal to the company xampl
Qualty mpovement teams unvesally aopte by evey company engage n ualty mpovement ae an example o teamwok. Qualit irle (Metho 5 ae anothe way o usng the benets o teamwok. Teamwok can pove a ocus o a goup o people n a task oce lookng at cossunctonal poblems. One oganzaton whch ha pe vously ha an autocatc style o management, whee manages tol eveyone what to o an then scplne them when they not ollow nstuctons, always epeence culty n ntoucng change. Thee was an atmosphee o suspcon, an most sgncant change ha to be negotate thou gh the powe ul tae uno n By makng sgncant change pat o a TQM pocess an movng towas coss-epatmental teams to ene an solve poblems the oganzaton was able to use the powe o the team to suggest change an then to mplement changes at a pace that woul pevously have been mpossble.
MANAGEME N M EHODS
Method 29
5
Total productive maintenance
upose
To e lp a process wic aim s at maki ng te mo st eective an ecient use o existing prouction structures When to use
en an organization nees to improve its maintenance system an to eucat e opera tors in mai ntenance tecniq ues. o to use
Use all te companys unctions an personnel to conribute o mainen ance troug bot in iviual eort an teamwork . Benefits
It provies eective an ecien use o existing prouction metos. It estabises a toroug system o preventive maintenance, conition monitoring etc. or te equipments enire iespan xample
A typica proce ure or total prouct ion main tenance i s as olows 1
Develop an agree te mainte nanc e strategy an ien tiy prioriti es or te key items o pla nt an equi pment in terms o tei r contribution o operating pans ake accoun o te companys manuacuring pilosopy an its manuacturing organization. 2 Establis target level s o mainten ance epenent on plant an equipment availability an ieniy tose items wic eman close attention an control. 3 Review te maintenance organization, especially in its reationsip wit prouction panning an sceuling an prouction itsel, aressing, or exam ple, te avantages o, an barriers o, combin ing responsibilites or maintenance, particulary were proucion sta can contriute to te provision o inormation on operating icul ties or iagnos tic routines
76
METH O FOR OAL QUAL Y MANAGE ME N
4 Investgate t e appcaton o eevant tecnoogy e g conton montong an ts possbe utue ntegaton wt any exstng statstca pocess cont o . Estabs a amewok o a mantenance pannng an nomaton system an ts ntea ce wt , o exampe te companys pouct on pannng an conto systems Det em ne te eeva nce scope an speccaton o a compute-base panne mantenance system 7 Specy te eeence ponts (e cuent stuaton) appopate atos yastcks an epotng system to measue mantenance value fr mney bung upon any cuent contos aeay mpemente Ienty tanng equements o bot opeatos an te mantenance epatment 9 Estmate te bottom ne gan te company can expect as a esut o any mpovement pog amme 10 Specy t e spaes suppot uncton ( e . povs onn g stockng pocuement systems an poceu es) 1 1 Pu a o te above togete as a mantenance mpovement pan wc ceay entes te objectves actvtes tmescae an esouces eque to mpement an aceve stanas o goo pactce
Reference Sichi akamima 9 Pss.
Introdtion to TPM. Cambridg, Massachsts Podciy
MANAGEMENT METHODS
77
Me thod 30 Why ho w cha ing urpose
hen thnkng n both abst ract an concrete terms, an nee ng to move between the two, whyhow chartng enables a goal to be translate nto acton. When to use
Ether nvua lly or as part o a group, the metho can be use to • map the mplcatons o a goal • generate alternatve statements o a problem • evelop alternatve solutons • present nngs an stmulate scusson •
begn the plannng process
ow to use
There are three smple stages: rte the goal or problem n the centre o a p chart. Ask a seres o why quest ons an pl ot the answers on the page above the goal. 3 Ask a seres o ho w quest ons an plot the answers on the pa ge below the goal Benefts
By brgng betwenge een strategc an tactrevealng cal thnnew kngways the technque t possble to chall goals , perhaps o meetmakes ng them whle smultaneously plannng to meet them. xample
In the example below the problem s that the managng rector overloae. The esre goal s to reuce the loang. Why? To generate new busness Why? To evote more tme to marketng Gal euce the MDs loang Hw? enty what can be elegate Hw? Dscuss wth sta ther roles
8
100 MEHODS OR OA QAY MANAGEMEN
Method 3 1
Zero defe cts
urpose
o allow teams to experence the success nvolve n meetng ever more emanng targets wthout emotvatng them by not achevng absolute success at once When to use
As part o qualty mproveme nt teams when settng targets an measurng mprovement ow to use
Zero eects s a very smple concept escrbe by Crosby (1984 The process s easy ollow thelevel teamoapproach: 1 Ienty antoagree theUsng current perormance Agree a target an tme-s cale or the target to be met wth mle sto nes 3 Mont or perormance aganst the target Publs h the resul ts 4 I the target s met on tme recognze the teams perormance an repeat r om step Benefits
The use o targets rsks emotvatng people they are set n a way that puns hes alure rath er than recognzes success Zero eects gves a way o settng tougher targets an recognzng the success o teams that meet them so encouragng uture eorts xmple
An organzaton nvolve n the evelopment o sotware s experencng problems wt h sotware bu gs an s b egnnng an mprovement pro ject to enty the causes o bugs so that the number present n new sotware releases can be reuce The current level o bugs per 1 000 lnes o coe s 2 The team sets a zero eects target o 15 ater the rst year. hen ths s successully met the ach evement s recognz e an t he zero ee cts targ et or the e o the secon year s set as 8. The target s revse every year an the team s recognze or the achevement o the target rather than amonshe or the presence o bugs The targets get tougher n each successve year
Reference B Crosby 94
Quty wtout e ew Yok: MGaw Hl
ANALYTICAL METHODS Metod 32
Cause and eect analysis
urpos
To examne eec ts or proble ms to n ou t te possbl e causes an to pont out possble areas were ata can be collecte Whn to us
en a team s tryng to n potental solutons to a problem an s look ng or te root cause o to us
Tere are our steps to constructng a cause an eect agram 1
Branstorm all possble causes o te problem or eect selecte or analyss Classy te major causes uner te eangs mater als metos macnery an manpower 3 Draw a cause an eect agram (see Fgure 1) 4 rte te eects on te agram uner te classcatons cosen Bnfits
en a proble m or eect s beng an aly se t can be temp tng to look or a temporary soluton or quck x tat oes not solve te problem at all but smply gets roun t Cause an eect analyss allows te problem to be consere ully an all optons consere. It also ponts to possble areas or ata collecton xamp
Te example sown n Fgure 1 examnes te possble causes o soler eects on a reow solerng lne Te group looke at all te possble causes an classe tem uner te man eangs as sown Te cause an eect root cause agram was ten use to plan ata collecton to scover te
ype of sode p a
:
e
Components Shape
Q
echnology
hickness
Component packaging
ul efow Batch size Comp nent densty ine ptch
Peventve maintenance
Fgu 1
Cause and effect dagram
$
ainng peato kil Knowledge elfsuficiency hfts chedule Weekend
{{
oos ven
{
{
ype of sceen eede
peed empeatue
RLO DCT
0 e J
» r » r
$ » » $ Z
ANALI AL M ETHODS
Me th od 33
8
Critical path a na lysis (CPA)
urpose
A project plannng technque whch separates the work to be done nto dscrete element s allow ng the key el ements tha t aect the overall projec t to be dented When to use
At the project desgn stage and then at all subsequent major project revews ow to use
CPA s a very smple process but t can be very tmeconsumng o overcome ths , there are ma ny sotware packages aval able to carry out the computa tons Each s de rent but the basc steps ar e the sam e 1 3 4
S
Branstorm all o the actvtes to be carred out and label them For each act vty , record any constrants such as tm e or order Draw the owchart o the actvtes n tme order Estma te the tme or ea ch actvty and ns ert these tmes n brackets on the approprate labelled brackets o the network Usng the avalable sotware, calculate the crtcal path and the dle (or oat) tme let the crtcal path ndcates that the projet not
possble n the tme avalable reexamne the assumptons made and mody the nsh date necessary 6 Montor progress recalc ulatng at each revew stage, snce the changes caused by actual perormance can change the crtcal path Beneits
CPA allows the eects o derent courses o acto to be determned at the plannng stage allowng the best overall approach to be decded xmpe
Smartwa re Systems L td s desg nng and nstall ng a comput er system or a cle nt n the brewng ndustr y he tasks nvolve d and the estmated duraton, together wth det al s o precedng actvtes whch m ust be completed beore
1 00 METHODS OR TOTAL QUALITY MANAGEMENT
82
able Tasks equied to install compute system Actity A B C D E F G H
Seect compter otware Int al oftware Tet otware Deeop data bae I nt al ofce ne twork Train employee Tet oce network Impement ytem
(
Preceding actty
Dration week
A B B A C,D E FG
2 3 4 9 2 1 6 2
)
a gvn act vty can start ar shown n ab 1 Smartwar n to stmat th mnmum tm takn to compt th projct an whch actvts any hav any xb ty o r sppa g wthout ayng nstaat on .
Uing the netwr Ntwo rk agrams (Fgur 1) ar us t o contro th xcut on o actv ts Som actvts w b crtca (n that any ay xtns th ngth o th projct) wh othrs ar not In bg projcts thr c ou b a vry a rg numbr o actvts to consr o ncat what spar tm thr s btwn actvts w show on th agram th oowng n ormaton th vnt numbr or asy r rnc (top numbr wth crc) th arst start tm or any actvty avng ths vn (numbr on wh r) h as nsh tm or any actvty ntrng ths vnt (numbr on rght wthn crc) h procur s as oows 1 Nu mbr a th vn ts v th start vnt th numbr 1 an gv th nxt nu mbr to any unnumbr vnt whos prc ssor vnts ar aray numbr Dtrmn th arst start tm (ES) or ach vnt (a) St th ars t start tm or th start vnt as zro. (b) For othr vnts consr th ES o ach mm aty prng vnt an a th uraton o th connctng actvty h ES o ths nw vnt s th larget o ths vaus 3 Dtrmn th atst nsh tm (LF) or ach vnt hs ar cacuat by workng backwars rom th nsh vnt (a) St th atst n sh tm o th n sh vn t qua to ts arst start (b)
tmothr hs vnts s th targt compton projct For consr th Ltm o or achth mm at y oowng vnt an subtract th uraton o th connctng act v ty Sct th mallet o ths vaus
ANALA L M ETHODS
u y ay
Citia l path
Fur
Nw dagam ccal pah analys
Calculang he a fr an acv Th oa t or a gvn actvty s th xtnt to whch t can b ay wthout ayng th projct as a who (assumng a othr actvts proc as schu) It s cacuat as oows atst nsh tm o uraton Foat oo wng vnt
arst start tm o prcng vnt
84
1 00 MTHODS FOR TOTA QUAITY MANAGMNT
I any actv ty has a oat o s ay 3 ays th en t took up to 3 ays onger than panne t wou not aect the project uraton However a eay onger than ths wuld eay the projects competon In our exampe the oat or each actvty s as oows Actvty Foat
A
B
C 5
D
E 5
F
5
H
I an actvty has a oat o zero t s a crtca actvty Any eay n a crtca actvty w eay the overa project Paths through the network compose ony o crtca actvtes are know n as ciical ah The crtca path s thereore A�B�D�F�H
Reeence .R Anderon Sweene Sene ew.. York Wet and TA Wiam
994) An Inodon o nement
ANALYAL MEODS
Method 34
Departmental cost of qualty
rpose
To prove a nanc al measure o the qua lty perormance o an organzaton When to se
Ether at the begnnng o a total qualty process to establsh the nee or change or l ater when entyng oppo rtuntes or mprov emen t ow to se
Departmental cost o qua lty e ntes thre e specc cost areas: 1
Preventin The cost assocate wth pl ann ng, tranng an wrtng
proceures assocat e wth o ng t rght rst t me Araial The cost o checkng an test ng to n out whether t has been one rght rst tme 3 Failure The cost, nternal or external , assocate wth alure to o t rght rst tme. These costs are calculate n two stnct ways Frst, the accounts, costs n each category
hard, or
Preventin e g tranng co urses, preventve mantenance Araial e.g eprecaton o test equpment an nspecton contracts. 3 Failure e g. scrap re-work, wa rranty Al l o these costs are rectly ente ro m accounts n ormaton Seconly , the t, or people , costs n each category: 1 Preventin e g percentage o peoples tme spent on tran ng, wrtng proceures, plannng etc Araial e g percentage o peoples tme spent on checkng, testng etc 3 Failure eg percentage o peoples tme spent on rework, hanlng alure n all ts orms All o the se ar e estmate ethe r rectl y or as part o an estma te nclu ng normal work (see Table 1 ) . These estmates are then converte to cost by usng average epartmental cost gures rom accounts Benefits
Ientyng the epartmental cost o qualty n an organzaton can have several benets:
METHOS FOR TOTAL QUALTY MANAGEMENT
86
Tabe 1 o o ual emae Cos of u y
Poon of m Wok dscpon cp nd bnkn of chus Posn csh cps o ss d Pocss cd nos fom Cs cvd fo und oods, pc djusmns nw dy c Chs pymn of ovdu Ds usn phonfx c Chck smns nd of monh nd hhh ovdu sums, xpo cn pymns c vw pckn nos c Cs on hod nd nfom supp of suon D wh cusom n us copy nvocs Now dv nd suppy h ud nfo To
Tm Nom spn wok
P
F
Nom wok
5
5
5
15
P
F
5
5
5
5
5
5
35
5
7
P, pvnon; , pps; fu
S Kn nd sh 99
It provies a benchmark or uture perorma nce. It bui ls ini viu al an companywi e awareness o the importance o 3 4 5
quality It ienties improvement projec ts or action. It ienties areas or investment in quality. It allows epartments to chart progress in cost terms.
xampe
he cost o quality was estimate in an accounts epartment where estimates were mae o normal work, prevention, appraisal an ailure (see Table
Reference BS 6 Gd h Enm f Ql GW Pk 99 Ahn C Efn Qly. ondon: Gow
ANALIAL METHODS
et hod 35
87
Dom ai al mappi g
rpose
To assist in the ientication o interna customers an their nees When to se
Either uring the iagnostic phase o a tota quaity process when conucting interna interviews or as part o a epartmenta purpose anaysis when reviewing the interna custome rsuppier reationshi p. ow to se
Begin the process by rawing a circe an pacing yourse or your epartment in the centre o the circe Spokes are rawn out rom the centre represent ing ierent interna customers an a circe is rawn at the en o eac h sp oke. The names o the inter na customers ar e written in each circe.
uality plans one month befoe fist poduction n spection of goods outine sevice no backlog etc
Figu 1
Domainal mapping
88
00 METHODS FOR TOTA QATY MAAGEMET
For each nterna customer, wrte own on the spoke what you beeve that ther requrements are . hen ths s compete vst eac h customer n turn an agree or ater the requrements Then agre e how the requrement w be measure an how eeback o perormance w be gven Beneits
Doman a m appng gves a pctora represen taton that aows you to thnk through nterna reatonshps an the requrements an measure s that go wth them mple
Fgure 1 shows a omana map rawn by a quaty manager estabshng hs n ks wth other epartments n the organzato n The Quaty Depart ment entes ts requrements e goos nspecton routne servce an no backog etc. rom the Prouct on Department The requrement s w be m easure an eeback o perormance w be gven n orer to u the nees o nterna customers
ANALI AL M ETHODS
Method 36
89
Evolutonay opeaton (EVOP)
urpose
A sequenta expermenta proceure for coectng nformaton urng on ne proucton to m prove a process wthout sturbng outp ut Wen to use
hen of f-ne expermentaton s expensve an you wou ke to use on n e process contro It shou be run not on y to prouc e prouct but aso to prove nformaton on how to mprove the process an prouct. ow to use
It requres a smpe twostage proce ss usng smpe statstca concepts It s run urng norma routne proucton by pant personne. At each stage a smpe esgn s use to estmate the recton of ncrease or ecrease of the ye as requre by the experment hrough the panne ntroucton of mnor varants nto the process the operatng contons are mae. Proceeng a certan stepength n the requre recton ay own a secon expement to e-estmate the ncrease unt thee s no further mproveme nt. Consta nt repetton of ths programme w ea to contnua mprovement of the pocess Benefits
It proves sequen ta sea rchng of the esgn space for the mpro vement of the proce ss wthout sturbng prouct on . xampe
In an nus tra process two temperatures are e nt e as havng an ef fect on the percentage of parts wth no fauts he two temperatures are surfa ce temperature () an base temperature () he rst experment gave the foowng r esuts
90
00 METHODS FOR TOTA QUAITY MANAGEMENT
t ° 74 75 74 75 745
t 5 7 7 5
Yeld n ercentage 5 5 71 7 8
To stmat t rcton o ncras t tt lnar mo o t procss rom t rst xprmnt was:
Y 8 + .95 t745 55 t 75 75 an t
T cntr or t scon xprmnt was takn at rsults wr
t 7 77 7 77 75
t 7 8 8 7 75
Yeld n ercentage 8 97 95 9 94
T abov normato n n cats tat t m provmnt on t prcnt ag yl can b obtan by usng 77 an 8 ptton o ts prodr wll lad to t o ptmm mprovmnt of t procss.
2°
Reference homas Ba (994) Qity y Epimn DgNw o Mac
ANALYTA METHODS
ethod 37 (FEA)
9
Failure _mode and eect analysis
urpose
To assist in t oolproong o a sign or a procss When to use
n invstigating a procss to intiy poss ibl causs o ailur o r wn xamini ng a prouct or srvic to look or wat can go wrong n t lattr cas, FMEA taks plac at t sign stag to allow prvntion to b 'pla nn in ow to use
FMEA o rs a structu r or tinking tro ug t lik lioo , srious nss an probability o tction o potntial problms Tr is a simpl procss to b ollow: 1 Brainstorm wat ca n go wrong. A lis t o potntial problms is gnrat , oring as many icultis as possibl For ac potntia l problm stimat ow lik ly it is to b o un i it is wrong Tis is gra on a scal o 0 as ollows
cale
nterretation
] 3 4 5 6 8 9
Vry ig Vry ig Hig Hig Morat Morat Low Low Vry low mot
Probability detection %of 86100 685 665 5665 4655 365 35 65 65 5
3 For ac potntial problm , stimat ow costly it is lik ly to b Tis is gra on a scal o as ollows
92
00 METHODS FOR TOTAL QUALITY MANAGEMENT
Scale 3 4 56 8 9 10
nterretation Mnor ow ow Moerate Moerate Moerate Hgh Hgh Very hgh Catastrophc
4 Estmate how lk ely t s that each potental alure w ll happen A scale o 110 s use as beore
Scale 1
3
4 5 6 8 9 10 5
nterretation
Lielihood
emote Very low ow Moerate Moerate Moerate Hgh Hgh Very hgh Very hgh
Eectvely 0 1 n 0000 1 n 0000 1 n ,000 1 n 1,000 1 n 00 1 n 00 1 n 0 1 n 10 1 n
Multply the outputs rom stages ", 3 an 4 together to generate a rsk prorty number (PN he PN wl l le between 1 an 1 0 000 I each stage gave a result o 5 the PN woul be 15
6 potental ank the alure potental the an PN wo arethan then00; use Frst nono canbyhave PN o rules greater seconly nvual output rom stages , 3 or 4 can excee hese rules show whch potental alures must be nvestgate Bnits
Falure moe an eect analyss gves a structure way o assessng possble alures an whch areas to nvestgate rst A major avantage s that t o es ths at an early stage n both proucts an proc esses xmp
A company mak ng plastc moulngs was examnng a new pro cess the hot water annea lng o nylon n jecton-moule components Two pos sbe
So � SupieUnt names) MOUING/NNEAN Cmnen decrtion VAROS NYLON INJCTION MI) N
DE
te:
FMA date (ogina): PIL 92
Occunce Hgh nely w noUnkely
(Le revin)
evety
rect NI.N ANNAN(;
Deen
tL ilH P0
rces functon I we nein of nyn injecio
J'oil f"ilue m e
Eet(s)f falure
, 1(
aue(s) f ailure
urren! contrls
Hig n.Hgh sevey Lw n Lw sevety nlkely w n.Ikely t ee Reve cnns
J;
y numbe
Recommened acn and statu
N
Renbe AeEngineer r ecve acon
aken
Rik rrty number PN
: :
mold
cmnent
m
ner
ne v n
)
(U) I'Url
�hlc
I
brittle break dive nstallaton () a al prematurely n use (d Pas exbe r nallan r d
v u wh
time cause by a) nadequate contr or ou dte nnation 0 Oeraor ang star/sop ess im k nti r s-r<ng clk
Fgu Faure mode and effec anasis Source: Kanj and Ash 13
Wk tucn le mark X
Iegular pedi s uf control daa
28
Verbal neracn cnro ata vl er ttnn
N cn
a) range or weekJy upda isue of control daa reeval of obsolete fle (b) Supevsor chec tal. log eo nM git clck aube Mrm Mhing lght b) nsM Mutmatc l releaser bke lft pen
1
:
r
neee
M
aa processng
S eaon
Mntnae
eekly bach un f control da mplemented Aug Hourly supeor ceck 1 book nsalled Au, 2 N umec cc . sireig ed Aug 92 lck a n li clse 0 A o bake reeae te Se
2
r s o
t W
Sht � � SupperUnt nae( : D/AG DET Component descrpton VAOUS NYLON
ME ae (oginal R 92 e igh no.key Occun w nonlikey
(tes revision)
MULNGS Site SINSOK
roect NYLON ANNAN
Severy eteton
ndon
em no
Contd
Pocess functon
Potenta faure mode Under annea
c
0
§ @
Eec(s} f aiure A page
Caue(s f faile (ii) Incoet water temperatre b () mmeson pat nu of o prouct ping wae temp (b) old wter rHs eyeJ ve fo wate at ng tnk tem. (v) xeive e (ii)&h anneang tmes ue (v) y g nt B sae aequaey (v Poy ag Mt' secon puncresd to of poy bg sha prodts lamnae
Fgue 1
Contnued
Hgh noHigh evety w no- w seve Hih no-nkeyodec w no.-key to detect
Revied onditions sk y numer (RIN)
Remmen aon() and sttus
20
Lie wate mp.
jk poy numbe (N)
espnse Area/ngne for coective acon
: o o s m I o
2
12
32
cntr st ital lock upon rechng rret tem Fi we u u+ emp cne t t k ! e () &
emp co oer nked t igit clck Aug 92
40
u ht waer tpu fcity e e 92
0
M"nenn
g/et 92
esgn ng
New heay dy laminate
N an Upgrae poly ag laminae speaon
9
pefed Jue
r p C l r
s l Z l m s m Z
ANALYT AL M ETHODS
95
alure moes were ente as uner/over-annealng an the eects o the alure were ente. As a result o the stuy (Fgure 1) areas wth hgh PNs were exam ne an actons to prevent these occurrng w ere put nto place The PNs were recalculate uner revse contons an showe a ramatc all.
Reerence G McAndw and S O'Suian 993 Ths.
FMA: Mngr Hndbook Lndn Stanly
96
1 00 METHDS FR TTAL QUALITY MANAGEMENT
Method 38
Fau lt tree a na lysis
urpose
To perorm a quanttatve as we as quatatve anayss o a compex system Whe to use
en you want to ustrate te connecton between a nonesre occurrence on te system eve an te causes o ts occurr ence on a ower system eve ow to use
Desgn o te aut tree begns b y specyng te nonesre occu rrence. Te mmeate cause o ts event ten as to be connecte wt te necessary sequence. Ts proceure s repeate unt a basc aut occur rence ev e s reace. Beefits
It eps to ncrease unerstanng o te ormaton an reabty o te system. It aso eps to etect crtca aure moes even wen te component ata are not avaabe
�
ght whee B o bake actio
Cabe boke
et bake pad wo
Cabe boke
Fault tree aalysis
ght bake pad wo
ANALAL MEHODS
97
xampe
In a hanbrake system conser the alures brake pa worn an cable broken wth the ollowng notaton A Complete : no acton brake acton wheel:system no brake B ght C et wheel no brake acton ght brake pa worn Y et brake pa worn Z Cable broken The ault tree an alys s or the occurrence o actor A s shown n Fgure 1
Reference B Bgman and B. Klfs 994 Nw Yk McGawl.
Quy om uomer Need o uome Scon.
98
MET DS FR T TAL QUALT Y MANAGE M ENT
Method 39
Force field analysis
upose
Aows you to enty those o ces that both hep an hne you n cosng the gap between whee you ae now an whee you want to be When to use
hen a team s pannng to mpemen t a souton o make a majo change ow to use
hee ae seven smpe steps nvove n oce e anayss 1
3 4 5
Ienty t he cuent stuaton hs s key to eect t he pobem statement Pace ths statement n the cente at the top o a page Beow the statement aw a vetca ne to the bottom o the page Ienty wh ee you shou be the ese state hs s pace on the ghth an se o the page at the top gan, a w a vetca ne to the bottom he am s to move the cente ne to the ghthan se, movng o m the cuent stuaton to whee you want to be Banstom a the aspects that hep these oces move the ne to the ght (posve oces Banstom a the oces that hn e these oces move the n e to the et (negatve oces Estmate the eas e o nceasng hepng oc es an eceasng hneng oces on a scae o 15 as oows 5, vey easy; 4 easy; 3 meum ; ,
1 , vey cut Estmate the m pact o hepn g an hneng o ces agan usng a scae 6 cut; o 15 as oows 5, vey s tong; 4, stong; 3, meum; , ow; 1 , vey weak ok out the po ty numbe o each oce by mut p yng (5) by (6 Benefts
Foce e anayss aows a team to thnk though the eect o changes, e nt y whch oces have most mpact, an see whee e ot can be use to max mum eect xampe
An oganzaton n the unome sevces s conseng mpementng a majo cutua change pocess an use oce e anayss to assst n ts
Forces (Helping)
c' 0
"
2
- orces (Hnderng)
» »
Communiaton
ear for future
2
2
Training
ak of trust
2
mployees desire to be involved
Poo oordnation
2
2
2
Compettion/CCT
nionzed managers
2
ak of onfidene
2
2
3» r
$m
I o
Fgue Potve an negatve fore n force el analy < <
00
00 METHODS OR TOTAL QALTY MANAGEMENT
ntroucton Fgure clearly entes te postve an negatve orces at work an eps te management team to pan te ntroucton o te cange e team ece to concentra te eavy on te communcaton o te canges n suc a way as to bu trust n te organzatons uture
Reference R Chang and M. Ndwck 993 Chang
otio Impom oo. Caiona Rchad
ANALAL MTHODS
Me thod 40
0
M nute an alyss
urpose
To estmate th e survva pero o a partcuar prouct unt uner certan contons, usng a smuate expermenta envronment. When to use
he n the uraton o the testng pero s preetermne an some o the test unts cou survve the uraton o the experment In genera, the resutng ata are expresse n the contnuous moe wth the speccaton o a aure or a survva or the whoe testng pero Such an anayss requres the assumpton o the aure strbuton o the unt A metho or mnute anayss oes not requre such an assumpto
n
o to use
In a testng experment, rst break own the test pero nto reguar ntevas o a specc uraton cae mnmum unt mnutes The uraton o the mnmu m u nt epens on the probem uner nvestgat on . An experment was conucte to nvestgate n whch mnmum unt the en o the e o a unt took p ace The ata are use n bnary or m wth (0 ) sgnyng that the test unt w st be uncton ng (not unctonng by the en o a partcuar cyce. The anayss o varance technque s perorme n the usua way approprate o bnary ata
Benefits
It oes not requre the assumpton that the aure strbuton o the unt oows a compcate ebutype probabt y strbuton
xamples
In orer to mprove the urabty o uorescent amps the oowng actors were seecte or stuy
A Atve ( agents types ( types B Fuorescent C Coatng metho ( types Tubewashng metho ( types
02
100 METHODS FOR OA QUALITY MANAGEMEN
For eac o te egt combn ato ns te urab ty o two lamps was teste Te wole testng pero was ten ays Every two ays ow many lamps were stl l unctonng was re core An anal yss o varance table can be construc te usng te ata ro m ts experme nt To etermne te be st operatng contons o r te paper- eeng pase at gwere spee o an oset uplcator, te ollowng our actors at two leves seecte
Decriin
Level
acuum eaer type
Normal Lgt wegt
Fee can type
Normal Smoote
Master cylner can
Smoote Normal
Ar re sellng
Normal Hg
Te ata wll represent te number o paper seets successully e troug te uplcator at eac test Ater statstcal analyss or eac o te egt levels te best operatng contons or pa per eeng can be obtane
Reference N Logohs n d . Wynn 99) Quty troug Deg. Pblatons
Ood Ood Sn
ANALYAL MHODS
Method 41
03
Pared comparsons
urpose
To hep a goup to quan ty the peeence s o ts membe s When to use
At the en o a banstomng sesson when tyng to euce a st to manageabe sze ow to use
Each opton s compae hea to hea wth evey othe opton In each nstance each membe o the goup s aske to vote o one o othe o the optons The numbe o votes o each opton s then totae an the opton wth the most votes s chosen The steps to o ths ae smpe:
Set up a g as shown n Tabe
beow The numbe o possbe compason s epens upon the num be o opton s Each membe has one an ony one vote n each compason 3 Eveyone must vo te n each compason howeve unappeang
Benefts
Pae compaso ns oce a goup to each a sng e concus on Some tme s as a esut the goup s ssatse wth the concuson an votes agan
Tabl Paed ompasos o desable shool eatues Opons A B E
AB 3
A A . A E 7
B
B
BE
6 2
E
E Tos
0
3 3
3
3
2 2 2 2 0
1 04
100 METHODS FO TOTAL QUALITY MANAGEMEN
xmpe
A goup o egh sxhom pupls have bansome a ls o chaac es cs esable n a school. These ae A Unesann g he elevance o lessons B Respec o eaches C oo acles D Fa ules E Exacucula acves The esuls o pae compasons ae gven n Table 1 The mos mpoan eaue om he pupls vewpon s unesanng he elevance o lessons
Reference P Spnly 992 Hall
Wold l Peromnce og ol QliyLondon Chapman and
ANAIA MODS
Method 42
105
P rmeter desig n
urpose
To etermne whch actors are mportant n the manuacturng process an to n the optmum set o workng contons When o use
hen you want to reuce the varablty o a process by changng the varablty-control actors whle mantanng the requre average per ormance through approprate ajustments to the target-control actors ow o use
Suppose that we have three comp onents n our process an tha t we ece to look at two stages o each component e then use all eght possble combnatons an make our runs at each o the eght 2 8) combn atons e can analyse the ata n two way s F rst we take as the r esponse the mean o the sets o observat ons In the secon ana lys s we take as the perormance characterstc the estmate varance at each o the eght combnatons
Benefis
It can prove the means both to reuce costs an mprove qualty By mak ng e ectve use o expermental esgn an statstcal technques one can ent y the settngs o easy-to-control prouct or process parameters xampe
Conser an electrcal power crcut where the caracterstc o nterest s the output voltage wth a target value o Y Assume that the voltage s largely etermne by the gan o a transstor n the crcut whose nomnal value x can be controll e. Supp ose that the ee ct o the transstor gan on the output voltage s non-lnear A transstor wth a gan o woul prouce the requre output voltage o Y The eect o a varaton about the nomnal value on the resultng varaton about Y s ncate by bans stralng the nomnal values. However the crcut esgner chooses a nomnal gan o x, then owng to the nonl nearty o response t can be seen that the varaton about the corr esponng voltage s much
METHODS FOR TOTAL QUAITY MANAGEMENT
reuce Now uppoe that there a retor n the crcut whch ha a near eect on the votage at a eve o trantor gan, then the retan ce o th compo nent can be choen o that th e ere nce between the votage y, an the ere votage emnate The repone then on target an the varab ty n repon e mn mze . Thu trantor gan a varabty-contr o actor an retance a target contro ac tor
Reference PW.M John Wy
(1990) Siicl Meod in ngineeing nd Qliy AnceNw ok
ANALAL METHOS
Method 43
07
Process cost of quality
urpo
To prov a nan ca l ma sur o th qua lt y prormanc o an organza ton Whn o u
Ethr at th bg nnng o a total qu alty procs s to stablsh th n o r chang o r latr whn ntyng opportunts or mpr ovmnt ow o u
Procss cost o qualty nts two spcc cost aras
Cnfrmance Th cost assoc at wth plan nng , tran ng an wrtng procurs assocat wth ong t rght rst tm, togthr wth th cost o chckng an tstng to n out whthr t has bn on rght rst tm Ths ar th costs o opratng th pro css as t s n a w holly ctv mannr th concrn s nt whth r th procss s ncssary or cnt Ths mans that whn oprat as spc, t cannot b on at lowr cost Th Ct f cnfrmances th mnmum cost or th procss as spc 2 Nncnfrmance Th cost, ntrnal or xtrnal, assocat wth 1
ncncy n th procss Ths costs ar calculat n two stnct ways Frst, th accounts, costs n ach catgory
har or
Cnfrmance g tranng courss, prvntv mantnanc, prcaton o tst qupmnt an nspcton contracts, togthr wth th normal costs assocat wth th procss 2 Nncnfrmance g scrap, rwork, warranty 1
Al l o ths costs ar rctly nt rom account s normaton Sconly, th ft or popl, costs n ach catgory
Cnfrmance g prcntag o popls tm spnt on tranng wrtng, plann ng, chckng, tstng tc 2 Nn-cnfrmance g prcntag o popls t m spn t on rwo rk hanlng aur n all ts orms
1
1 08
1 00 METHODS FOR TOTAL QUALTY MANAEMENT
s it iday?
Fgu
es
oha o poe ot o qalt
A o tese areese estmate et are er recty or as partot o anby estmate ncung norma work. estmates ten converte cost usng av erage epartmenta cost gures rom accounts Process cost reucton can come rom two stnct areas reucton n CONC (cost o non-conormance) va operators an reucton n COC (cos t o conorm ance) va proce ss owners Benefits
Ientyng te process cost o quaty n an organzaton can ave severa benets It proves a bencmark or uture perormance It bu s n vua an company-we awarenes s o te m portan ce o 2 quaty an nter-epartmenta ssues 3 It entes mprovement projects or acton. 4 It entes areas or nvestment n quaty. 1
ANALYTIA METHODS
09
xampe
The owchar showng how a sales eparmen receves an processes orers s shown n Fgure 1 . The coss o conormance nclue • • • • •
recevng he orer checkng he orer or correcness an auhoraon upang he conrol regser rasng he elvery noe senng he noe o he warehouse The cos s o nonconorman ce nclue
• • • •
reurnng ncorrec nvoces reurnng ousanng orers answerng elephone queres ealng wh resubme orers
Refeence BS 63 Gd o om of Qaty Qa osgLondon Chpmn nd Hl. B.G l nd Pn 99
00 METHODS FOR TOTAL QUALTY MANAGEMENT
0
Method 44
Re li ab il ity
upose
o n h caus o alurs an ry o lmna hm an o ruc h cs or consquncs o alur. When to use
hn h nrnsc characrsc o an obc sysm or uncon s no prormng as xpc by s usr or h pro o m nn by h sgnr o h objc or sysm. ow to use
Suy a un rano mly chosn rom a manuacu r bach . h probably ha hs un wll work atr h oprang m t can b sma by masurng h man m o aur o h un. hs can b achv by obanng h ms o alur or many uns o h sam kn oprang unr h sam conons. hs probaby rgar as a u ncon o , s call h rlabl y un con an s wr n ( ) . I ( ) = 99 hs mpls ha abou 99 pr cn o h uns o a bach wll survv oprang hours.
Fge
m Reliaili diagram
aur tr
ANALYTAL METHODS
Benefits
I helps o increase he ailure resisance o he prouc an he oerance o he prou c o ailur es xample
In a sress srengh analysi s experimen i was observe ha ma ny ailures happene ue o a weak poin which ha been expose o a srong sress This sress can aris ue o exernal or inernal coniions raual weakening can be cause by aigue corrosion eerioraion or iusion The caasrophic eec is as a rule an immeiae overloa Figure inicaes he e ecs o sress on a uni
Reference B Bgmn nd B. Klsjo Nw ok McGwH
994 Qualiy fm ume Need ume Saifai
2
00 MEHODS OR OAL QUALI MANAGEMEN
Method 45 control)
Robust design (o-line uality
upose
To acheve the proper unctonng o a component even when aecte by nterer ng act ors whether externa l nternal or manuactur ng varaton When to use
hen the nvual unt o a prouct s expose to sturbance an to ocus on reucng the varab lty o the process ow to use
Dve the sturbng ac tors nto the ollowng groups 1 nner diturbance .e wear an tear o the nvual unt ue to ts operaton 2 Outer diturbance e varaton o temperature an other envronmental actors urng usage 3 Manufacturing variatin, e evaton rom the set target Use the esgn to check whether nvual u nts expose t o the above sturbances alter or vary n mportant characterstcs. Benefts
It helps to ocus on reucng the varablty o the process by esgnng qualty nto the process. xample
Statstcans use the term robust to mean nsenstve to epartures rom set con to ns. The sample me an can be pre erre to the samp le mean as an es tmator becaus e t s robust aganst extreme observ atons It s oten possble to choos e a es gn that wl l be more or less senstve to sturbance In an experment the amplcaton o a transstor an the output o voltage can be aopte oll own g the esgn pr ocess By choosng a nomnal amplcaton A nstea o A the voltage can be ajuste by the sprea o actual ampl caton The l evel can then be controlle towar s the target value wth the help o a resstor
Reference B. Bgman and B Kfsjo 99) Nw ok McGawH
Quiy From uomer Need o uomer Sfon
ANAIA HODS
Method 46
Solution eect analysis
urpse
To eamine solutions to problems to nd out whether there are any detrimental conseuences and to plan the implementation of the solution When t use
When a team has found potential solutions to a problem and is eamining them to decid e whic h to i mplement
w t use
There are four steps to constructing a solution effect diagram Brainstorm all possible effects of the solution selected for analysis
2 Classify the eects under the headings materials, methods euipment and peo ple 3 Draw a solutio n effect diagram (s ee Figure 1 ) 4 Write the effects on the diagram under the classications chosen Benefits
A proposed benec ial chang e may have sideeffec ts elsewh ere hese side effects may be as bad as the problem bei ng solv ed Solution effect analysis alows the implementation of change to be planned by identifying and removing any detrimenta sideef fects xmpe
The eample shown in Figure eamines the eff ects of int roducing a new workin g method int o an ofce The group looked at al l the possibe effe cts of a new accounti ng system and classi ed them under the mai n headings as shown The solution effect diagram was then used to plan the introduction and min imi ze the sideefec ts of the new metho d This was do ne by taking each item o the solution efe ct diagram in turn for eamp le remove old
4
00 METHODS FOR TOTAL QUALTY MANAGEMENT
e -desig scree W C W
ect o ot he ofice
emoe old forms esg ew orms
eamwok
Fgu
Solun efe agam
form and agreeng a plan wth uer a nd a tmecale to brng about the necear acton The reult wa mnmum drupton.
ANAYA MHODS
Method 47
Strati ficatio n
urpose
To assist i the deitio of a poblem by idetyig whee it does ad does ot occu When to use
I teams, whe tyig to idetify a poblem ad dee it pecisely as the st stage o poblemsolv ig o to use
Staticatio is a meth od o splittig dat a accodi g to whethe it does o does ot meet a set of cite ia The value of the techiue is to expose pattes i the data It is used bere data collectio begis to desig the way i which data will be collected, ad aer data collectio as a way of focus ig th e aalysis The pocess is vey si mple 1
Baistom a list of citeia o chaacteistics that could cause systema tic difeeces i the data These ae ot ecessaily thigs that d cause dieeces but thigs that culd cause them Desig the data collectio foms to iclude these items 3 Collect the data ad examie them fo ay pattes o teds Benefits
By focusig o data beoe collectio, staticatio esues that all ecess ay data ae col lected st ti me a d avoids wasted effot xpe
A team was exa mii g the poblem of abseteeism The lis t of chaacteis tics geeated at the stat was Name Age Sex Day Cosecutive days
MHOD OR OA QUAI Y MANAGM N
Month Bo name Wok aea Ovetime woked Total da abent The data colected wee analed accoding to thee categoe and patten looked fo Analt mght , fo eample , evea that the da afte pa da o cetain da of the month wee patcual pone to abentee im
Reeence B Bgmn nd B Kl ( 994 Q fom Csom Nd o Csom Ssfo N k: M-Hl
ANAYTIA MTHODS
Method 48
1 1
Syste m desi g
urpose
To apply specal scentc and engneerng know ledge to produce a basc unctonal prototype model, havng surveyed the relevant technology manuac turng envro nment an d customer need When to use
o desgn ualty nto the proce ss by carryng out resear ch and develop ment eperments n order to nd the best operatng condtons or the sats acton o the customer ow to use
Reject the noton that to be outsde specc aton s bad , but to be wthn speccaton, even by only a whsker s sats actory Replace t by the new dea o proc ess capablty n connecton w th the crtera C and Ck. Make people thnk n terms o amng or a targ et vaue an d work to reduc e the mean suare error that s the sum o the suare o the bas and the varance It s necessary to control both o them because the process reures zero bas and smallest possble varance
Benefits
t helps to take the prototype model and make t happen accordng to customer reurements xmple
The proc ess capablty nde
C s
C USS 2 6 6 When US and S are the upper and lower speccaton lmts respect vely As (standard devaton) decreases, C ncreases 2 s speccaton wdth
8
00 MTHODS OR TOTL QULTY MNMNT
Te second citeion is
pk minimum (CP CPU)
pl
when
wee J tget vlue J ctul vege, CPU uppe pocess cpbility nd CPL lowe pocess cpbility Nowdys, semiconduc to mn ufctues e stiving fo po cesses wi t 20 nd 1 5 which epesents bout 3 defecti ve pts pe million
p
pk
Reference PWM on Wy.
(990 Siil Md i Eii d Quliy AuNw Yok:
ANATIA METODS
Me thod 49 Tagu ch me thods urpos
A techniue or the optmzation o products or processes Taguchi nvolves a twostage eperimental design that gives the benets o robustness and ecency with the minimum number o eperiments. Whn to us
During desgn commissioning or postcommissoning when seeking the optimum operating charact erstics o a product or process
o to us
Eperimental design usually involves attempting to optmze a process whch can involve several actors (eg temperature time chemical compostion) at several levels (eg ve possible temperatures our possible tmes si possible chemical compositions) Factors can also be attrbutes or eample a switch can be on or o The actorial approach to eperimental design would thus nvolve 6 4 5 cells the replicatio ns needed or accuracy This is a minimum o 120 eperiments and could depending on the variability be 360 or more This can n practice be reduced by what are known as ractional actorial designs but the number is still large The Taguc hi approac h takes each o the act ors at two lev els (usuall y the etremes) and works out which has the greatest contribution to the end result These a ctors are then studied in more detail In our ea mple the eperime nt would probably involve studying three ac tors n eight eperi ments wth one repeat so that the intial design would involve 16 eperiments as compar ed with a possi ble 360. This de sign would allow all the actors and ther interactions to be estimated Designs o this type were src ina lly developed by lacket t and Burmann n the and tables are availab le givng m any dierent designs together wth analysis detals The desgns are in the orm o matrices called orthogonal arrays The stages to go through are actors and/or interactions to be evaluated 21 Selecton Selection o the number o levels o the actors 3 Selection o the approprate orthogonal array 4 Assgnment o the actors or nteractions to the columns
00 MTODS FOR TOTA QUATY MANAGMNT
20
able
Ohna aay esun fm pape manufaue expemens ato
al no. 2 3 6 7 8
A
B
C
2 2 2 2
2 2 2 2
2 2 2 2
D
E
G
2
2 2 2
2 2
2 2 2
2 2
2
2 2
2
2
Wate ( 2 3 9 8 22 20 26
5 Conduct the eperiment Analys e the result s 7 Carry out a conrmation eperimen t Bfs
Taguchi m etho ds give a fast and pragmatic approach to the optimi zation of products and processes They can also be used for tolerance design the setting of statistically based tolerances, allowing either improved perfor mance through tighter tolerances or cheap er designs in noncritical areas; and for fast type approval testing xam
A company involved in paper manufacture was seeking to optimize the process in terms of grade output The parameters thought to have inuence were Factor A Factor B Factor C Factor D
Machine speed Coating Wire position Clay content
55ft/s 02mm High High
0 ft/s 0mm ow ow
A decision was taken to carry out a sevenfactorineight eperiment design , using the remainin g four f actors to represen t interactions between the major factors The orthogonal array is given in Table The total waste for A was (2 5 3 9)/4 A2 was ( 22 20 + 2)/4 B was (2 + 5 + + 22)/4 = B2 was (3 + 9 + 20 + 2)/4
49/4 7/4 57/4 = /4
225 90 425 700
NLIL MHODS
11
C was (2 + 5 + 20 + 2)/4 C2 was ( 3 + 9 + + 22)/4
73/4 52/4
25 3.
wa s ( 2 + 3 + + 20)/4 2 was ( 5 + 9 + 22 + 2)/4
53/4 72/4
325 .
Machne speed was seen to be the major actor and the optimum operatng conditon was orecast to be
Al
B C2
Whch was late conrmed by eperment
Reference . Ross (9 Tagui Teiques for Quaity EgieerigNe Yok MGra-Hl.
METHODS F OR TOTAL QUALT Y MANAGE ME NT
Method 50
Tolerance design
Pus
To nd out b experiment where the variabilit in a process (product) occurs and where adjustments can be made Wh us
When the inuences of inner and outer sources of 'noise cannot be successful l reduced b use of arameter deign (Method 42 Here noise represents the effects of uncontrollable factors w us
It reuires the same steps as parameter design, but additional factors are considered that were previousl excluded because o c ost or the difcult of experim entatio n If this also fails, the tolerance of the products components are considered This means retaining the optimum nominal levels of actors, but reducing the tolerance of certain crucial fa ctors in an optimal and cost effective wa so that overall variabilit in the response is educed to acceptable evels Bfs
It assists in the stud of fac tors whi ch are expensive and difcult to change in experiments for the improvement of a process Exam
If we run a second outer arra centered on the optimum parameter condition s, this time the noise a ctors are noise in the pa rts themselves If factor is a resistor at a nominal level of 100 ohms, it becomes in the noise arra, a factor with lev els of, sa 100 ± 2 When the arra has been run , we n d how much variabilit is associated with each actor and, hence, where we have to make further improvements
A
Reference P.WM. Jon Wiey
(990) Sl Md n Engnng nd Ql AnNe Yok:
IDEA GENERATION Method 51
Brainstorming
Pus
To generate as many deas as ossble wthout assessng ther value. Wh us
n teams when tryng to dentfy ossble root causes or when seekng solutons to a roblem Branstormn g can also be used when decdng what roblem or mrovement actvty to work on and when lann ng the stes of a roject w us ,
Branstormng seems very smle . t works best when the team meetng s nformal. To hel ths there are eght basc rules:
1 Kee the meetng relaxed. 2 Select a leader to wrte the deas on a chart. 3 nvolve the rght eole n the team 4 Dene the roblem clearly You wll need to check that everyone esent has the same understandng of the roblem Ths can be dfcult to acheve n ractce A useful rst stage of any branstorm could nvolve a bref dscusson of the roblem before a denton s agreed. 5 Generate as many deas as ossble wthout dscusson or evaluaton Ths comes later There are two man ways of dong ths The rst s smly to nvte eole to contrbute and wrte the deas down as they are suggested. Ths s called the free wheel ng method The second s to go round the room askng each erson n turn for hs or her contrbu ton Ths s called the round robn method. 6 Encourage everyone to contrbute Ths s best done by begnnng the sesson wth a trval examle such as 'uses of a aer cu to get everyone started before movng on to the queston n hand. 7 Wrte down every dea There should be no censorsh and t here s no such thng as a bad dea Sometmes strange deas oen u a new area of thou ght.the branstorm a technque such as lst reducton Method Followng 59 should be used to reduce the branstormed lst to manageable roortons
1 4
1 00 METHODS FOR TOTAL QUALTY MANAGEMENT
Bfs
By encouragng everyone to contrbute, branstormng breaks down barre rs between dear tments and le vels of herarchy t theref ore allows everyone to contrbute equally to the team Branstormng encourages cooeratve and collaboratve behavour and s also useful n the develo ment of gro uwork sklls Remember that branstormng nvolves collectng eoles deas and onons and that t mght be necessary to collect data followng the branstorm to allow any decsons to be taken on the bass of fact xam
An organaton was seek ng suggestons to red uce absenteesm They ran a seres of branstormng sessons to generate deas for tacklng the roblem
Reference Cag ad M. Nedzieki (993) Cag.
otuou Improemet Too. Caf a a d
IDEA GENERAION
Me thod 52
Brai nwri tin g
Pus
To generate as many ideas as ossible. Wh us
n teams, when trying to identify ossible root causes or when seeking solutions to a rblem Wherea s brainstorming generat es as many ideas as ossible brainwriting results in fewer but more fully develoed ideas w us
The success of brainwriting deends uon having an agreed denition of the roblem to be address ed befo re beginnin g the session This should be checked with all team members before the start. There are three stes involved
1 Team members individually write down their own ideas on cards 2 All the cards are laced on the table or stuck on the wall Team members then take someone elses card from the table and add to it. A set time i s allowed to do this Members may cont ribute ideas t o as many cards as they wish 3 Reeat ste 2 Bfs
f there is conict within the grou or if the subject is likely to be controversial brainwriting may be more successful than brainstorming Brainwriting en courages gro uwork and consensual teamwork by asking team members to build uon the ideas of other s as wel l as generate ideas of their own n a situation where it is imortant to have more develoed ideas brainwriting wi ll rove more us eful than brainstorming. Exam
An organiation was considering the i ntroduction of exitime against some known oosition They decided to use brainwriting in teams to allow everyo ne to contribute and build a lan for its int roduction They foc used on the question How do we introduce exitime successfully? Everyone
MET HODS FOR TOTAL QUALT Y MANAG E ME NT
used cads to wite thei ow ideas ad the cotibuted by addig to the ideas of othes. By focusig o the ositive, ad geeatig a deth of ideas they wee able to build a la fo t he itoductio of exitim e. Team membes took esosibility fo the ideas that they had oigiated ad heled to big about a successful outcome
Reference Lnce Gbb ( 987) Too or problem oving London: PA Conulting Grou p inte rn report
IDA GN RA TION
Method 53
27
Brea k ng set
Pus
To overcome blocks in thin king by generating new ideas t is articularly useful in romting a grou to be more recetive to new suggestions Wh us
f a grou has run out of ideas or has become set on a articular line of thinking w us
The idea, soluti on or roblem is written on a i chart and is examined by the grou, asking several key questions under the following headings
1 Ada What else is like this? Are there any arallels in the ast? What could we coy? What other ideas are like this? 2 Mdify Can we ch ange th e meanin g, colour, sound, sme ll , feel , form, shae? 3 Magnify What can be added? More time, height, sie? Dulicate, multily, exaggerate? 4 Minimie What can be taken away? Smaller, condensed, simlied, streamlined, slit u, omitted? 5 Subiue Who rocesses, else, what aroaches? else, instead? Other sulie rs, com onents, materials, ower, 6 Rearrange Swa cause and effect? Other attern, layout, sequence, schedule? 7 Revere What are the oosites? Swa ositive and negative? Turn it round, uside down, change roles? 8 Cmbine Combine uroses, elements, ideas, blend, assortment, ensemble? Bfs
Differeideas nt ways of viewing same a good mean s of generating more Often when a the grou hasi dea triedaresomething and failed, it is reuctant to change track Breaking set rovides a way of building on an existing idea to genera te new id eas.
MEHODS FOR O AL QUAL Y MAN AGE M EN
Exam
An organiaion was looking for a novel design for a ead grid used in auomoive baery man ufacure Th e grid isel f had o conduc elecric iy , have a cerain hy sical srengh and be abe o have ase suck o i . There were oher characerisics The grid had been made of graviy cas ead in heno as
Ada Mdify Magnify Minimize Subiue Reaange Revee
Ne curains, he fron of an old wireless. Cyindric al , round hexag onal Coninuous no discree grids More holes less meal hinner Sines no a grid. Change he hrional and verical axes. Sack he grids aher han hanging as befor e.
The design changed as a resu o an exanded grid ne curains ha was coninuously cas raher han graviy cas.
IDEA GENERAION
Method 54
1 29
Bu zz g rou ps
Pus
A way of geng he mmedae reacon of a grou o a new dea or roblem. Wh us
Bu grou s are used o generae ene rgy when a grou s sagn ang They are very useful for changng he focus from revous work and brngng everyo ne back no he dscusson. w us
Bu grous wo rk bes for small numbers of eole A maxmum se of four s suggesed Ask he grou members o form several small grous and ask hem a smle queson or al locae a smle ask The queson could be Wha do you wan o do now? 'Where shall we ea ongh? or How shall we recogne successful rojec eams? The ask could be allocae he worksho no eams of wo f or he nex sesson They are asked o reor back on her dscussons Bfs
The bene comes n changng ack and generang lvely dscusson. Bu grous are very nosy and brng eole back no he worksho. They can also be useful f as worksho leader you need o refocus he drecon of he worksho. Exam
A comany mlemenng he TM rocess used he bu grous ech nque a he begnnng of each ranng sesson o record curren roblems and ge eole n he rgh frame of mnd for he ranng o come
Reference
. Be Mat ad P Rae 993 Workop t Work. Ne Yk MGa-Hl
M ETHODS FOR TOTAL QUALT Y MAN AGE ME NT
Method 55
Idea writing
Pus
To bring all artici ants into grou work Wh us
dea writing is used to generate energy when a grou is stagnating t is useful for changing the focus from revious work and b ringin g everyone back into the dis cuss ion w us
dea writin g is a very sime roces s involving ve stes 1 Break the worksho articants into subgrous of about four or ve eole and give every one an idea writing sheet.
W ame .................... y question or statement
irst response
esponse o o ther paicipants
Fgue
Typil ide wriing hee
......
DEA GENERATON
2 Eac atcat wtes s o e ame ad a statemet o questo o wc commet s vted. 3 He o se te wte s s o e ow esose to te statemet o questo o te same seet. 4 Te seets ae assed to ot e membe s of te subgou wo add te ow esoses te seet. All wtg sould be comleted a maxmum of 30 to mutes ad sould be doe sletly. 5 We comlete te ogal wte eads out te questo ad all te esoses to t usg t to geeate dscusso. Bfs
Te beet of dea wtg s as a dveso we te wokso leade judges tat te atcats eed some quet tme. It s atculaly useful tadtoal low sessos suc as afte luc as t volves lttle eegetc actvty but muc tougt. Exam
A tyc al dea wtg seet s gve Fgue
Refeence T. Bouer Martin and P ace 993) Wkp Wk. Nw Yok McGaw-Hi
M ETHODS FOR TOTA QUAIT Y MAN AGE M ENT
Met hod 56
Ima gi neering
Pus
To assst a co many to dentfy areas of oor tunty by concentratng on the deal outcome then workng back from t. Wh us
When clarfy ng vson and buldng what to do.
a lst of actons to assst n lannng
w us
There are ve stes n magneerng
1 Branstorm a lst of features that charactere the deal stuaton. Ths lst can be develoed n a grou atmos here nvolvng everyone
2 For each of the referred or actual characterstcs dented, state the actual current stuaton n relaton to t.
3 For each of the characterstcs dentfy the ga to be brdged to brng about the deal st uaton 4 Use cause and effect analyss to break down the ga nto small areas that can then be addressed 5 For each of the small areas dented agree an owner and tme-scale for cometon. Bfs
magneerng breaks down what can seem a dauntng task nto a lst of actons whch can be ndvdually acheved
rvc ow
Fgure
v rvc
mairi diaram to brid ap from actual to idal situatio
IDEA GEN ERA TION
1 33
Exam
The maage of a bak is egaged i a custome cae ogamme ad is wokig out what is the ideal situatio fo the baks customes The st stage is to baistom a list of ideal chaacteistics: fo examle wam geetig attetive atmoshee ivacy asfo desied etc. staff The ext stage is tosevice idetify easat the acual cuet situatio exame ufie dy sevice sow digy suo udigs oe la o y etc The thid stage is to idetify the causes ad the ga to be led Fo each of the costituet ats Figue ) a la is the daw u to bidge the ga.
METHO DS FOR TO TAL QUALT Y MANAGE M ENT
Method 57
Improve internal process (P) plan
Pus
To rovide te structure to develo work lan detais for a task using various factors, suc as measurables resonsible resources, times and revious task owners Wh t us
Wen you w ant to im rove internal rocess I IP by using information at a secic level of te develoment of te roject or task w t us
1 List and number eac task 2 Provide te measure of comletion for eac task. Te measure sould be clearly dened data, quantity and level of erformance
3 Al locate res onsibi lity for comletion of ea c task 4 Give details of te resources required for eac task.
S
Write down te time wen eac task will be comleted
6 Find te revious owner or eac tas Bfts
It rovides te structure to imrove internal rocess Exam
An examle of te information matrix required for an IP lan is given below in Figure
o
ask
eet the deadline
2
oforder each
0
eeting delivery date
llocation
veral responsibility
easurement
esources
ime
Previous owner
tatus
m m
ending the correct order
Packaged properly
o » m m Z m
�
z
ke
opal K
Zero defect
Packaging material
weeks
ike
Fure
Infomation matrix required or P
M ODS OR OA QUAI Y MAN AGM N
Method 58
La teral thi nk ing
Pus
A way of tranferrng from one frame of reference to another enablng you to break down barrer whch nhbt creatve thought W us
Lateral thnkng ued when a team tryng to dentfy roblem or oble oluton t can alo be ued a a dveron to relax a workho grou w us
There are everal welletablhed way of begnnng lateral thnkng Mg mphs
Th nvolve ung a metahor to brng a new look to a tuaton or roblem Suoe that you are lannng to ntroduce tattcal roce control and are exerencng oblem o accetance at mddle manage ment level. Thnk of omethng that ha arallel wth ntroducng a change ; for exam le redecoratng a room Now magne what you can do to make the redecoraton a ucceful a oble. Brantorm a lt of dea Now for each control dea dcu relate mght to thebentroducton of tattcal roce (SPChow An texamle 'make t more lght The arallel wth SPC mght be to look at educaton and tranng and conder atttude Rdm jp
Th nvolv e ntroducng a comletely new noton to al low more dea to be generated A grou ma y be conderng fac tor n the degn of chool room. At a certan tag e the dea of 'nfant mgh t be ntroduc ed to extend the co e of the dea generat ed Sppg
Th nvolve what de Bono 970) call the 'ntermedate moble The dea bet llutrated by an examle An organaton exerencng
DEA GE NERA TON
1 37
problems wth the outgong qualty of product The rst response s We cant affo rd to employ any more nspecto rs. The 'n termedate mpossble would be Make everyone an nspector Ths would result n changes n role and tranng to allow everyone to take responsblty for ther own work Bfs
Establshed patterns of thoug ht can stand n the way of nnovatve deas Lateral th nkng provdes a way of removng barre rs to new de as. Exam
Lateral th nkn g was used by an organat on lo okng for dffe rent ways of marketng a veday resdental course Random juxtaposton was used and the dea spouses was ntroduc ed As a result the organaton off ered a deal for spouses that ncluded a day at the races followng the course
Reference Edad de n (90 Lterl Ne Yrk Harer and R
METODS OR TOTAL QUALITY MANAGEMENT
Method 59
ist redu ctio n
Pus
To reduce a list of ideas to oe of maageable sie Wh t us
At the e d of a braistormig sessio or ay sessio ivol vig the geeratio of ideas w t us
There are two mai ways of reducig lists of ideas. These are 'hurdlig ad votig. Hudng
A way of testig the items o the list agaist a list of hurdles or criteria. Examples of this might be 'Achievable i weeks or 'Costs less tha £10000' Those items that fail to jump the hurdle are the discarded This idea ca be exted ed by compilig a list o f 'musts ad 'wats The list is divided ito the two categ ories. f there are o 'wa ts, this ca be achieved by weightig the 'musts by givig each a score o a scale . The higher the score the more importat the 'must Agai, oly those items that clear the hurdles remai o the list. Vng
t ca be used by itself or with hurdlig The team decides how may votes each perso ca be allocated Each team member the allocates their votes amog the ideas geerated The ideas receivig fewest votes are crossed off ad the process is repeated util a al list of the key ideas is reached. Bfts
Dealig with a very log list of ideas geerated durig a braistorm ca be dautig List reductio gives a structured approach to this task Exam
A braistor mig sessio examied ways of reducig sickess absece. At the ed of the sessio the group used hurdles to reduce the list geerated The
DEA GENERATON
hurdles used were, rst 'Wll ot lead drectly to dustral urest ad secodly, Ca be troduc ed at o addtoal cost. The redu ced lst forme d the bass of a acto pla
Reference T Bouer
Marti d P Rac 99) Wkp Wrk New Yor: McGraw-Hi
1 40
1 00 THODS FOR TOTA QUAITY AAGT
Method 60
Mi d ppi
Pus
A way o geneang and ecodng deas ndvduay ahe han n a gou Mn d mang mak es use o wod assocaons, encouagng you o oo w you own hough aens wheeve hey may ead aso ovdes a wen ecod o he deas geneaed Wh us
can be used as an aenave o s -makng as a way o geneang deas w us
Thee ae sx sme ses nvoved
1 Pace he oc, ssue o obem a he cene o a age cean shee o ae
2 Ao w you m nd o wande abou he oc 3 Sang wh you s hough daw a ne o m he cene an d abe wng aong he ne oow he hough by dawng banches ou om he ne, abeng hem as you go. 4 Whou ausng, when one dea uns ou , sa a ne w one wh a new ne sang om he cene Take cae no o evauae o cce wha you have wen down We down a deas as hey occu he wh deaseach dy u ake a deen cooued en and jon u deas 6 When assocaed ohe Bfs
Oen when you ae hnk ng ceavey abou a obem a o he hough ow ges os and s dcu o ecea e . The ouu om ed by mnd mang ovdes an exceen emnde o you an o hough xam
An exame o mnd mang s shown
n gu e 1
Reference za (9) e r Head Ld Aiel k
l) 't
�
� �
o m » G m Z m : » -
,1
� m
0
6
m
." VSE
Fgue 1
Example of mind map
z
42
00 MEHODS FOR OAL QUALIY MANAGEMEN
Method 61
Morphologcal forced connectons
Pus
To geeae ew deas o ways of aoachg oble ms I combes ls s of abues ad fo ces ew coecos bewee hem , so ggeg ew oos. Wh us
To d ew ways of aoachg a old oblem whe old assumos abou he oblem ae blockg eceo abou wha ca be doe w us
Thee ae fou smle ses
ttibutes
deising
aget
Costs
oation
iming
Cuent methods
ews
gs
200/ day
ome ounties
/2 days
boad
2 hs
tenatie methods
nempoyed xeuties Bs Chambes of Commee
hoo eaes
Betting shops et mai Psons
gur
orpologial ored onneion
pen eaning fte wok Between shifts
DEA GEN ERA TON
List in a ow along the to of a i chat the main attibutes o chaacteistics of the aoach o device cuently used 2 Below each attibute, list as many altenatives as ossible without any evalua tion o censo shi. 3 Geneate seveal diff eent fo ced connections aco ss the columns them u. by geneating new ideas Follow though 4 oin likely ideas found.
o by ageeing t o examine any
Bfs
When it is difcult to think of new ways to aoach an old toic foced connections geneate ossible new aoaches that allow you to beak fom outofdate assumtions Exam
The examle Figue shows f oced connecti ons used to g ene ate altenative ways of maketing quality management couses. The ideas joined togethe fom a new idea fo maketing the couses The og ania tion decided to advetise in the TEs aiming at the unemoyed and un the couses in schools at weekends
METHOD S FOR T OTAL QUALI TY MANAGE ME NT
Meth od 62
Mu lto t ng
Pus
To seec he mos ouar or mora ems from a s. Wh us
Muv og s ofe u sed a he ed of a brasormg sesso o reduce he s of ems o a more maageabe se
whe ryg
w us
Mu vog s coduced by usg a seres of voes, each cug he s haf, aowg h e fas reduc o of arge ss of deas.
1 Geerae a s of ems; for exame , deas by bras ormg. 2 Combe ay smar ems. 3 Number he ems
4 Each eam member s aocaed a umber of voes equa o oehrd of he oa umber of ems o he s. They aocae hese voes o he umber of h e em Oy oe voe er em er ea m mem ber s aowed 5 The eam me mbers he ca ou her voes as ea ch em umb er ur s dscussed 6 Emae hose wh oe voe or ess f he eam s arge, hs umber may be creased o wo or hree 7 Re umb er he s ad reea s es u o y a few ems rema f a hs sage here s o cear favoure use ared comarsos o reac h a a cocuso Bfs
Muvog ads he reduco of a arge s of ems. Exam
A he e d of a worksho whc h h ad deveoe d a s of cr ca busess rocesses, he grou used muvog o reduce he a caaogue of 7 rocesses o a maageabe s of seve.
Referene Arturo Onne (99) Quality taly
e Lne o o QCateamonte T PoK Pubaton on
IDEA GENERATION
Method 63
Nom ina g roup t ec hnique
Pus
A way of generating ide as from a grou and id entify ing the e ve of su ort within the grou for those ideas. Wh us
When a grou is trying to reach a concusion on a coective course of action or when one individua is dominating a grou and you want to invove others. t can aso be used at the end of a brainstorming s ession to bring together the ideas w us
This is a very sime technique that invoves six stes Agree a grou ead er to record the ideas on a i chart. 2 The eader asks each grou member in turn to give an idea or suggestion and records this on the i chart There is no discussion or evauation of the ideas at this stage. As with brainstorming, the ideas shoud be there for a to see. 3 Reeat ste 2 unti a the idea s that the grou has have been exhausted. 4 Aow the grou members to seek carication of any suggestions that they do not understand. 5 Ask each g rou memb er to write down on a iece of aer their v e most imortant ideas in order giving the score of 5 to the most imortant, down to for the east imortant When this is comete the members are asked to record their scores on the i chart next to the idea being evauated 6 The grou eader adds u a the scores and reorts the resuts. f the grou began with a ot of ideas it may be necessary to go through the rocess twice to reach a concusion Bfs
To add structure and therefore contro to an ideagenerating session. The grou eaves the session committed to the outcome. t can aso be usefu for a work sho ead er need ing to exert contro and bring a grou ba ck to order.
M ETHODS OR TOTA QUAT Y MANAGE ME NT
The techque ca be vaed, fo examle, by gvg gou membes coloued makes ad usg these to do the al scog, o by smly gvg ve votes ad allocatg these be twee deas Exam
Nomal gou techque was used at the ed of a woksho whee the leade was hav g dfculty b gg the sesso to a close Th e techque was used to geeate ad agee a l st of ct cal bae s to the m leme tato of total qualty. The aoach was esecally useful as he atmos hee was emotoally chaged ad the leade eeded to calm the stuato.
Reference T Boue V Martn and P Rae
(99 Wokop Wok. Ne York: MGra -Hl
IDEA GENERATION
Method 64
1 47
Opporu nity a n alysis
Purpose
Offers the op portunity to evaluate quickly a long list of options against desired goals and available resources When to use
This method can be use d by an individual or group when presented with many opportunities where it is difcult to decide what to do rst How to use
1
2 3
Write down all your goals in the situation under review Rank the importanc e (to satsfy the customer ) of each goal, and rate your ability to complete them. Do "you have the require d resources?
Benefits
It provides a rapd eval uatio n of a long l st of optons in r elati on to desired goals and the a valable resources.
mpoace High Medium ow
oals 1
Provide timely invo ce
X
Match delvey date as equested
X
3 Update al the iformato egulaly goods to the order
X
X X
6 Pack products popely
X
Opportunity analysis
X
X
5 Sed complete order
Figure 1
bility to complete High Medium Low
X X
48
00 METHODS FOR TOTA QUAITY MAAGEMET
Exam
Fgure rovdes an oortunty anayss for a comany that wants to satsfy ts customer requr emen ts. Here tems 2 and 4 of 'Hgh mortance and Hgh abty to comete shoud be done rst.
IDEA GEN ERA TION
Method 65
Ric h pictures
Pus
To aow a gou to catue a deas deveoed wtout judgement o anayss n a ctoa fom tat aows te stengt of te deas to be ecoded Wh t us
Wen a gou s exam nn g a obe m a c ctu e can be deveoed to sow wat aeas te gou as consdeed so tat a tougts about te obem ae taken nto account w t us
A c ctue deveos a ne of tougt as t occus wtout eacng concusons Bfts
Dung teamwok te stengt of an dea, o te way tat t s exessed often temts te gou nto com ng to concusons too eay n te oces s. Rc ctues catue tose deas and aow te team to move on. Te ctues fom a moe evocatve ecod of te gouwok and make t fa ease to emembe te dscusson tan s te case wt wtten ecods Exam
In Fgu e a team as been examnng quaty of se vce ssues n a ote can Te ctue sows te data as tey wee coecte d wtout anayss o judgem ent Foown g te team meetng, te deas wee taken off te c ctu e and tuned nto a an of acton .
J F I.'n"
coJn
.
t o
o o o
m
I o
"
�
C
:
Q
�
r
p C �
C
r
� � Z � G m
m
Z
Figure
1
Rich pictures generated during group work on hotel service issues
DEA GENERATON
Me thod 66
Sn owballi ng
Pus
Sometmes called yamdng snowballng s a technque fo gatheng nfomaton o deas Wh us
n a woksho settng, whee atcants do not know each othe, snowballng s an excellent way of beakng the ce and helng membes get to know each othe w us
Ths s a vey easy technque fo dea geneaton Thee ae fou smle stages:
Patcants a e gven a task whch s to be caed out nd vdu all y The task should nvol ve wtng down deas o nfomat on on ae 2 Patcants ae then aed togethe and asked to contnue the task, takng t futhe 3 The as ae then made nto gous of fou and asked to contnue the task 4 The combnng s contnued untl all the woksho atcants ae nvolved n one gou Bfs
Snowballng allows eveyone wth deas to contbute to the wok of the gou wthout the embaassment of havng to seak ndvdually t also allows ownesh of deas to develo n the gou An addtonal use of snowballn g s when a gou s becomng owdy an d the gou leade s seekng to e-exet contol The technque beaks the gou away fom what t was dong and gves the atcants a secc task quetenng them down Exam
A woksho gou of eght dectos s consd en g what key efomanc e ndcatos to set fo the busness:
MODS FOR OA QUAIY MAAGM
Sage 1 ndvdually we down e dea abou e ndcao Sage 2 n a combne e l an d educ e e numbe of ndcao o 5 Sage 3 n gou of fou combne e l and agan educe e numbe of ndc ao o 5 Sage 4 A wole gouofdebae wo l combne em and educe e anal numbe ndc e ao
Refeence Ber V Mar ad P. Race 993 Worksos at Work. Ne Yrk McG ra-Hll
IEA GEN ERA TION
Me thod 67
Sug ges tion schemes
Pus
To generate deas for mrovement. Wh us
At a later stage of a total qualty rocess when there s a qualty mrovement lan n lace an d a mechansm for selectng deas for act on. w us
Suggeston schemes can seem decetvely smle and many fal because deas are aske d for wthout any mec hans m beng set u f or decsontakng or communcaton of those decsons. The folowng are gudelnes for runnng successful suggeston s chemes Set u a steerng grou to oversee the rocess 2 Delegate as far down as ractcal the decson on whether or not to mlement the suggeston 3 Gve awards to suervsors and managers whose emloyees generate most deas 4 In the early stages go for quantty of deas rather than qualty. 5 To generate deas n manageable numbers try to set themes for deas and change the themes regularly 6 Ensure that al deas are acknowledged quckly and that the erson generatng t he dea s told whether t s to be mlemented an d f not why not 7 Look for reasons to say 'ye s rather than no . 8 In the early stages be reared for deas to be about envronmental asects rather than oeratonal ones Bfs
Suggeston schemes rovde a way of movng to contnuous mrovement va smal ncremental changes Exam
A head ofce functon of a major servce organaton ntroduced a suggeston scheme as art of ts total qualty rocess. In the rst year 8 suggestons emlo yee were generated an d 90and er cent acted uon. The organatoner began a 'Suggeston of the month 'Suggeston of the year award and the scheme ourshed
DATA COLLECTON, ANALYSS AND DSPLAY Method 68
Ba r ch art s
Pus
To display discrete data collected by checksheets so that patterns can be discovered. W t us
n the early stages of problem-solving when a team is trying to nd out what is happening w t us
The number of times an event occurs is shown on the vertical axis The value at which it occurs is shown on the horizontal axis. Figure 1 shows a bar chart of the reasons for intest failure of a printed circuit board Bar charts are sometimes used in conjunction with Pareto analyiMethod 0) Bfts
Bar charts give a simple pictorial representation of data Exam
The following data were used to prod uce the bar chart in Figure 1
ailure ce
Exlanation
A B C
Missing component Wrong component Wrongly inserted Cold solder joi nt Short pins Test machine fault
E F
No failing ,8 980 6 40 409 1
The bar chart shows clearly which reasons for fail ure ne ed to be addres sed to have any impact on the problem
DATA OLLETON ANALYSIS AND DSLAY
200 00 000 00 00
,
00 00
z
00 00 00 200 00 0
Fgure
B
a cod
Ba chat pntd ct boad
Reference H Kum 985) Sl Md Quly mpmn
MEODS FOR O QIY MNGEMEN
Meth od 69
Bas ic stati stics
Pus
The mean median mode ange and sandad deviaion ae ways of summaiing and descibing age voumes of daa The s hee ae measue s of ocaion ; he as wo ae measues of sead. Wh us
When o oking fo aen s in daa o when ying o comae age voumes of daa hese aamees give saisicay based measues ha aid decision-aking w us h an
The mean is he sime aihmeic aveage of a he daa oins To cacuae he me an, add ogehe a he oins and divide by he numbe of oins in he sum. The mean is usuay wien onounced X ba). The fomua is
X
+ + + .. . . n
whee hee ae n oins caed o As an e xam e he foowing i s gives he voage dei ved fo m a cic ui having vaiabiiy due o oo oeance cono . .3 .5 .7 8 9 2.0
.2 .3 . .6 .7 .8 .9 2.0
.2 . .5 .6 .7 .9 .9 2.0
3 . 5 .6 8 .9 2.0 2.
2. 2.3
2.
2.2
2.2
Thee ae 37 oins in oa.
ATA OTION ANAYSIS AN ISAY The sum X s 1. 1 + 12 + . . . + 23 The mean voltage 63.3/37 .71
57
63.3.
Te medan
The medan s the 50 er cent ont, the ont above whch and below whch hal f of the onts le There s no formula for calculatng the medan As an examle, the above data are lad out n the form of a tally chart as follows
Cun
Cumulaive cun
1 12 1.3 14 1.5 16 17
Value
1 2 3 4 3 3 3
1 3 6 10 13 16 19
18 19 20 21 22 2.3
3 5 4 3 2 1
22 27 31 34
36 37
The medan wll be the 19th o nt , some where between 1 . 7 and 1 8 t could be descrbed as 17+ volts. Te mde
Ths s smly the most frequently occurrng ont For examle, n the data above, t s 1 9 volts . When the data ar e symmetrcal abo ut the mean , the mean, medan and mode have the same value Te nge
The range s a measure of th e overa ll sread of a set of valu es t s dened as the arthmetc df ference between the largest and smallest value n the above examle, the larg est va lue s 2 3 and the smallest value s 1 1 so the range s 23 1 1 1.2 volts. Te standad deatn
The standard devaton s otherwse known as the root mean square devaton of all values from the mean. t s calculated by takng the dfference of each value away from the mea n, squarng t , and addng t to
METHODS OR TOTAL QUALT Y MAAGE ME T
al l the other square d deviations . The total is then divided by the number of values involved minus The square root is then taken This is written
n
1
In the examle the calculation is:
+
36
0.3 volts
There are simle methods of testing whether means ranges and standard deviations of different samles dif fer stat isticall y Bs
Statistical techniques use the basic data and rovide facts uon which decisions can be made Exam
Examles of the mean, median, mode, range and standard deviation are given above.
Reerence G.K Kni (99)
Si
London age
DATA OLLETON ANASIS AND DSLA
Method 0
Box and whisker plots
Pus
To rovde a s mle way of drawng the basc shae of the dstrbut set of data.
on of a
Wh us
n qualty montorng resentaton when you want to descrbe the d strbu ton of large data and th er dserson n order to ca rry out exlorator y data analyss w us
The central box contans the central h alf of the data wth the end of the box markng the quartles and the central l ne the med an . The whskers g o out to the extreme values Wth larger data sets there s a conventon that the whsker lnes enclo se say 97 5 er cent of t he data wth more extreme values marked ndvdually Wth small data sets the whskers end at the largest and smallest values Bfs
Samle method to descrbe the dstrbuton of data and ther varabl ty. n artcular they hel to exlore large data n the work of rocess control.
c I >
N I
0
Figue 1 A box and hisker plot
L c I >
N I
a
20
0
0
M ETHODS FOR TOTAL QUALIT Y MAN AGEM E NT
Exam
A box and wske lo fo daa conanng 50 obsevaons s gven n Fgue 78 8 9
06 35 9 8
33
Hee,
32 2 6 5 7 Q
Q Q
7 3 2
252 8 3
quale Medan 3 quale
29 27 5 3 25 25% 50% 75%
3 33 22 7 5
293 7
87 20 23
30
6
Mn value Max value
=
28 2 9 5 28
3 30
Reference B Begman and B. Klefjo (99) Qty: rom stomer Needs to stomer Stscto. Ne Yok McGa-Hil Tkey (99 Exortory Dt A y Ne Yok Addon-Weley
DAA OTION ANAYSIS A ND DIS AY
Me thod 7 1
161
cha
Pus
To dentfy when the numbe of defects n a samle of constant se s chang ng ove tme Wh us
When montong a ocess to detect changes o when a change has been made to ocess nuts to nd out whethe the numbe of defects o oblems also changes. chats ae used whe n the samle se s constant, o does not vay by moe than 5 e cent of the aveage samle se.
w us
Thee ae sx smle stes nvolved:
Collect data showng the numbe of oblems o defects ove tme Daw u a table sho wng the num be of de fects fo each lot n umbe The numbe o f defects s called The total numbe of lots s called m Plot the dat a fom the table onto the contol chat The successve lot numbes ae shown on the hoontal axs the numbe of defects o oblems s shown on the ve tcal axs 3 Calculate the cente lne Ths s calcated as the sum of all the s dvded by the sum of all the s and s wtten as
"n
4 Calculate the contol lmts whch ae They a e calculated a s
± 3 sd about the cental lne
+ 3 3
Ue contol lmt UC L Lowe contol lmt LCL
f the lowe contol lmt s less than eo t s taken to be eo 5 Daw the cental lne and the contol lmts on the contol chat. 6 nteet the esults Bfs
t can be dfcult to seaate out andom vaaton often called common cause o nonassgnable vaaton fom eal vaaton caused by changes to
M ETHODS FOR TOTAL QUALT Y MANAGE ME NT
the rocess charts gve a way to do ths for the umber of defects or roblems wth a samle se that s costat Exam
A orgaato s exerecg roblems wth the umber of errors or m scodgs of voces ad a chart was used to motor the error rat e A costat umber of 80 voces wa s samed each week wt h the foowg result s see Fgure 1
Lot no 1 2 3 5
6 87 9 10 1 12 13 1 5
16 17 18 19 20
No. of error 5
3 0 7 0 92 5
6 0 8 5
0 6 2 7 0 7
For the data above the cacuatos were as foows: 80 m 20 80 CL n
m 20 K = 3 3 \= 6 UCL + K 1 0 LCL K = 0
Refeene M Amen H E Bu an R T. Am n 99 SP Simped or Sr Lnn: Cap man an Hal
o
Dprtm
Plan!
no
lC_
Avege.
0 0
c
u
0
T1 SmeSize:
q o
'
�
1
(
r m
b
(
�
L 3
(
�. : 0
: z � z � r (
e
(
mpe (nJ
5
Numbr (p )
7 0
� z o o ( "
�
rotIn p )
-
Oef
Any Change n pe maeial uipent, me s o evrmeho b oted d . Thee noes wl elp o tk o rrece or pross mproement acon when sga� d b e cntol cha .
Figure 1
Bar chart of prnted circut board
�
e w
00
1 64
ETHODS FOR TOTAL QUALITY ANAGEENT
Meth od 72
Ch eckshee ts
Purpose
To collect data when the number of times a defect or value occurs is important. When to use
Either during problem deni tion when you are collectin g data to nd out what is happenin g, or when you have implemented a solution and you are collecting data to monitor the new situation How to use
There are ve simple steps to draw a checksheet diagram: 1 2 3
4
5
Agree the data to be collected This step is vital: you cannot analyse data that have not been collected. Desi gn the check she et Test the checksheet using someone who has not been involved in the design Get him or her to use the checksheet wthout assstance If necessary, modify the checksheet. Design a maste r checksheet. If more than one person is to be involv ed in data collection , you will need to bring t ogether all the data co llected The way to do this is to use a master checksheet Collect the data. ay
otal
bsences
Monday
,; u j
uesday
J ,
'I
22 6
Wednesday
I
3
hursday
\, I
4 18
riday Figure
1
H � > 1/\
Checksheet on absenteeism
DAA OLLETION ANALYSIS AND DISLAY
Bfs
By estabishing the facts about the incidence of faiure, a team can an to identi fy the causes of faiure and ook for ways of removing them Actions are taken on the basis of evidence not feeing. Checksheets area an exceent way of ing eothat e i ncan quaibetyeasiy imrov e ment. They give sime method ofinvov data coection understood and aied in a wide range of areas Exam
During an attemt to reduce unauthoried absence from work a team coected data on the day of the week that absences occurred The checksheet shown in Figure 1 gives the resuts of t he study over a six-month eriod The team used the data in consutation with deartmenta managers to crack down on absenteeism .
Rn H Kume 98) Ss hods fo Qy Imomn.
METHO S FOR T OTAL QUALT Y MANAGE ME NT
Method 3
Concen trat ion di ag ram s
Pus
To collect data when the lo caton of a def ect or roblem s mortant Wh us
Ether durng roblem denton when you are coectng data to nd out what s ha enng, or when you have mleme nted a soluton and you are coectng data to montor the new stuaton w us
There are ve smle stes to draw a concentraton dagram
2 3 4 5
Agree the data to be coected. Ths ste s vtal you cannot analyse data that have not been collected. Desgn the concentraton dagram. Test the dagram usng someone who ha s not been nvol ved n the desgn Get hm or her to use the dagram wthout assstance f necessary, modfy the dagram Desgn a master concentraton dagram f more than one erson s to be nvolved n data collecton, you wll need to brng together all the data collected The way t o do ths s to use a master dagra m Collect the data
Bfs
By establshng the facts about the locaton of falure a team can lan to dentfy the causes of falure and look for ways of removng them Actons are taken on the bass of evdence, not feelng Concentraton dagrams are an excellent way of nvolvng eole n admnstraton areas n qualty mrovement. They rovde a smle method of data collecton that can be easly understood and aled n ofce areas xam
A large organaton was concerned about the number o f road accdents haenng on the ste, coverng aroxmately 0 mles by 3 mles
DATA OLLETIO N ANALYSS AN D D SLA Y
In an attempt to nd out why the accdents wee happenng, the oad map of the ste was used as a concentaton dagam and each tme an accdent occued the poston of the accdent was ecoded on the map The concentaton dagam showed cea pattens n the ocaton of accdents. The ghts oganaton omaton p ace 'seepng nd tafc to sow used downthe thenftafc at thetopaces whee mostpocemen accdent as had been happenng. The stuaton was montoed and a fa n the numbe of accdents was noted.
MHODS OR OA QUA IY MAN AGM N
Method 74
Cusu m cha
Pus
To denfy wen e mean
vaue s cangn g over me
Wh us
Wen monorng a roce ss o deec canges, or wen a cange as been mad e o ro cess nus o nd ou weer e mean canges Cumuave sum or cusum cars can be used o mon or bo varabes and arbues daa , and are arcuary u sefu for deecng ong-e rm rends n daa
w us
Tere are sx sme ses nvoved Coec e daa Cusums are arcuary usefu for ong day or weeky averages 2 Decde wc vaue o subac from eac vaue Te vaues used mos ofen are e overa mean and e arge vaue Cusums are very sensve o e vaue seeced Too age a vaue and e cusum w dsaear off e boom of e car; oo sma a vaue and dsa ears off e o f n doub, e overa mean soud be used, n s case e cusum w end on e x axs 3 ubrac s vaue from eac daa eemen Te resung number s
5 6
caed e resdua Accumuae e resdua vaues Po e accumuaed resduas on e axs and e me on e axs nerre e resus A cange n mean vaue s sown by a can ge n soe f e cusum s movng downwards, e curren mean s ess an average; f s movng uwards, more an average
Bfs
Cusum c ars are very sensve o can ge and so can gve very fa s warnng of rocess canges Tey are aso very accurae so can ofen non exacy wen a cange occur red Te downsde of ese wo carac ers cs s a ey can somemes deec unrea canges Cusums can aso be used as ar of exermena desgn wen r yng o reae nu and ouu varabes Concurren canges n bo cars
DAA O IO N A NAY S S AN D DIS LAY
abl 1 Cusum data and vaues
Wk n.
rdvy v (
m v
�(X- T)
57
-5
-5
2 4 5 6 0 2 4 5 6
6
2 0 4 2
2 4 4 4 0 2
17
5 6 6
45 4 42 56 2
6 54 52 6 71
5 53
6 6 55 0
6
20 2 22
-4
4
95
65
-4
5 4 5 4 6
suggest that thee s a lnkage and that the nut vaable s affectng the outut Ths then needs to be checked statstcally Exam
An nsuance was lannng to movng cay out some tanng n thecom clams handlngcomany dea tment amed at oductvty The any decded to use a cusum to monto ndvdual oductvty to assess how effectve the tanng was The taget value chosen fom hstoc data fo each wokes oductvty was 62 e cent The data collected and cusum va lues ae gven n Table The cusum chat Fgue clealy shows a change followng week 3 when the tanng couse took lace Fom week to week 3 the oductvty value was aoxmately 62 e cent, the hstoc value Fom week the sloe of the cus um changed coesond ng to a new value of 7 e cent The tanng couse had been effectve
Rfnc
M. Own a SP nd Bsinss Imomnt ndn: S ns.
1 70
1 00 MET HODS FOR TOTAL QUAIT Y MANAGEMENT C C ,
C
C
c: , ,
O O
�
c
C
c
<
'
C
DATA OETION ANAYSS AND DSAY
Method 7 5
11
Dot pots
Pus
A
smle grahc devce whch resents observatons as dots on a horontal scale. Wh us
When there are less than 30 observatons and you want to use a smle dagram on a ece of aer wthout any fuss. w us
Draw a horontal scale and mark the observaton above t wth dots Bfs
ery easy to use and useful whe n you have fewer than 30 observatons It s also useful when you are comarng two sames.
•
0 gr
•
•
• 0
• •
•
•
•
•
2
0
• •
••
0
Dot plot
• •
0
•
•
ge 2 Dot plot fo two amples
•
•
2
• • • 2
•
• 76
•
•
• 0
MHDS R A QAY MAAGM
xam
gue how a do t pot o a poce y el d o te n un n a chemcal pla nt 66 77 68 69 7 70 72 73 62 75
(ample)
gue 2 ho w the compot on o two ampe o a poce yel d om a chemcal pant by dot pot Sample 72 7 73 76 7 77 6 9 80 6 8 79 Sample 2 72 7 79 73 7 67 78 77 80 65 Hee we ee that the aeage o the two ampl e ae about the ame but ample 1 ha le aaton than ampe 2
Rc PW h 990) Ss ehods n Enneern nd Q Assne rk Wly
DATA OETON ANAYSS AND DSAY
Mehod 76
1 73
Fl owch ar s
Pus
To generate a icture of how work gets done by iking together a the stes taken in a rocess. Wh us
When a team i s working on roces s im rovemen t it is rst ne cessary for a members of the team to have a common understanding of the roc ess. Fowcharts are aso a necessary stage in the introduction of ISO 9000 w us
Having the correct team is essentia when drawing a owchart It necessary to invove a those who are concrned with the rocess There is a sime rocedure to foow when drawing a owchart: 1
2 3
5
Brainstorm a the individua activities that make u the rocess. List the activities in the order in which they are done. Using waaer or some other arge sheets of aer, draw out the activities in schematic form. Common owcharting symbos are shown in Figure 1 . Ask each member of the grou in turn whether any activities have been missed out and whether he or she agrees with the process as drawn. Make changes as necessary. Test the owchart by taking an exame and 'waking it through the owchart
Bfs
Often rocesses in organiations are not designed, but have evoved over time Fowcharting aows rocesses to be chaenged, and gas duica tions and dead ends iden tied . It therefo re eads to roc ess simication xam
Figure gives a owchart for drawing a owchart
74
100 METHODS FOR TOTAL QUALITY MANAGEMENT
gree changes
Figure 1
Flowchart for drawing a owchart
Reference Gar Bon
( 1 994) Process Management to Quality Improvement. Nw York: W
DATA OLLETION A NAY SIS AN D DIS LAY
Geometric moving average
Method 77 Pus
To idenify rends in small changes in he rocess mean The geomeric moving average is someimes called he exonenially weighed moving average EWMA .
Wh t us
When monioring a rocess o deec chang es in he rocess mea n o r when a change has been mad e o rocess inus o nd ou wheher he rocess mean changes Geomeric moving average chars can be used o monior boh variables and aribues daa and are aricularly useful fo r deecing longerm rends in daa. One advanage is ha he roce dure rovides a forecas of where he rocess mean will be a he nex ime eriod w t us
There a re four simle ses involve d
1
Collec he daa The ye of daa could be sales gures, facory volumes error rae or exchange rae. 2 Calcu lae he subgrou averages by adding ogeher he daa in each subgou and dividing by he subgrou size. 3 Calcul ae he overa ll mean of he daa by adding ogeher he subgrou averages and dividing by he number of subgrous. This saisic is designaed 4 For each successive subgrou, calculae he saisic W as fo llows:
where r is a consan beween 0 and The choice is gove rned by he radeoff beween he need o deec an i moran change wihou false alarms Conrol limis can also be calculaed Bfts
Geomeric moving average chars are relaively insensiive o shorerm changes. This means ha heir main use is when you are rying o mask shor- erm variaion in he rocess o highli gh lo ngererm variaions The
MHODS FO R OA QUAI Y MA AM Tabl Deene beteen saes tagets and aeements Sgr
Sgr rg
w
1
0.4
002
2 3 4 6 7 8 9 0 1 12 13 14 1 16 17 18 9 20
1.1 -004 .04 -0.04 -068 0.8 081 062 038 04 08 0.33 0.32 0.44 0.22 0.88 0.6 0.9 1 09
0.344 0.28 0.0682 0.062 -029 0.3744 -0.0783 -0.0963 0.0228 01 179 01 134 0.1826 02170 02728 0.296 04147 030 0.8 0.6481
technique can aso be used to edict whee the next data ont w be usi ng ast data. xam
Tabe 1 shows the dffeence between saes tagets and achevements fo a saes tea m The data a e eco ded daiy , based uon a same se of ve The subgou aveage is based uon these ve eadngs The vaue of chosen was 025
Rn Thms . Ry 1989) Sttst ethods fo Qt Imroement. Yk Wly I
DATA COEC TO , AAY S S A D DS P AY
Metod
78
. 177
istors
Pus
To dsplay contnuous data colected by check can be dscoe ed
sheets so that any pattens
Wh us
At the ealy stages o poblemsong when a team s tyng to nd out what s happ enng us
Thee ae ou smple steps nol ed
1
Colect the data usng a checksheet 2 Use the etcal as to dspay the numbe o tmes each alue occus 3 Use the hozontal as to dsplay the alues ntepe t the hstogam Deent pattens o hstogam suggest that the pobem beng studed has patcula chaactestcs attens eea when two o moe thngs ae beng med o eampe deent ways o pocessng cams They aso show when data a e ben g censo ed o eampe when someone s alng to ecod cetan data tems They can also ndcate when thee s tme dependence n the data o eample when someth ng can take a ey long tme but when t s mpossble to take a shot tme Bfs
Assumptons o nomaty made about data need to be checked be oe data can be analysed usng statstcs that depend upon nomalty Hstogams ae a smpe sua way o ewng data that hghghts nonnomal stuatons When these ae dented the data can necessay, be analysed uthe The pctue seen can ge useul adce to teams tyng to estabsh ac ts about what s h appenng xams Cobke sogr
The hstogam n Fgue 1 shows two sets o haemogobn measuements taken by two nuses usng sghty deent methods The eadngs they
METHODS FOR TOTAL QUALTY MANAGEMENT
obtaied aear sligtly dislaced o te istogram imyig tat te differ eces i te metod of takig t e readigs are resutig i differet aemoglobi readigs t was ecessary to stadardie te metod to obtai reeatable readigs S imi lar examles would occur i mi xig iuts from two suliers usig a scale tat i s too e or i readig e rrors due for exame, to arallax r oblems Clflke hstogm
Te istogram i Fgure 2 sows data were a goo go gauge revets te resece of data beyod certai bouds t would terefore be ysically imossible to ave a readig below 23 mm ad tis sows as a cliff-like face o te left of te istogram Tis occurs were tere is cesorsi of data were it is ysically imossible to roduce certai values because of a costrait Ske hstogms
Te istogram i Figure 3 sows data were a mould is exibitig wear over te erio d of te sift te readig is slowly movig to te rigt Oter examles of tis woul d be age to failure of comoets or te amout o f time to rocess a isur ace caim.
Referene H. Km 98 S Med r Quy Impreme. .
2
0 � �
0
' <" :
(
r r
0
z
(
=
2 0
12
17
1
19
20
2
l agl (dgr) Figr
22
2
2
2
2
2 2
0
� z � r e U � z 0 0 U
Cmb-ke stgam
METHODS OR TOTA QUATY MANAGEMENT
�
I
I
I
C
I
DATA COLLECTO,
AALY SS A D DS PLA Y
2 0
6
2 2 / / / 20/2 222 2/2 2/2 2/2 0/ i ag (dgr) ig 3 kew htgm
MEHODS FOR OAL QUALY MAAGEME
Method 79 Hos hin kan ri ( ua it y poicy depoyment) Pus
To delght the customer through the ma nufactur ng and servcn g rocess by mlementng the qualty goals of the organaton Wh us
When objectves are dented at each level of an organaton by to down and bottomu consultaton and the overall goals of the organaton have been set as secc targets w us
Dene sh ortte rm and long-term goals of the organ aton . 2 dentfy the measurable goals . 3 Decde the crtcal rocesses nvolved n achevng these Ask the te ams to agree on erformance nd cators at arorate stages of the rocess. 5 Cha lle nge every level of the rocess n order to force the organaton to change the qualty culture 6 Organatonal goals are to be used as measurable goals n order to make the emloyee understa nd the mortance of the qual ty mrove ment rocess. Bfs
t shows the emloyee what the overall goals of the organaton ae and where he or she ts n so that everybody ull s n the same drecton towar ds clearly dened goals. xam
Examl es of goals set at dfferent levels of an organ aton mak ng them evdent n all organaton rocesses, are as follows:
DAA COLLECON, ANALYSS AND DSPLAY
Level
Goal
ororate Deartment Mantenance Manufacturng
Deght the customer Reduce cost of oor qua ty Reduce machne faure by 20 er cent Less than 3 er cent defect
Devery
Less than 5 er cent late deveres
Referene Hin Kni Picy Deyment fr Succefu TQMmidg Y. Ak ( sshstts: dtivity ss.
1 84
1 00 M ETODS O TO TA QA TY MAAEME T
Mtho
80
Is/is not ati
Ps
To denty pattens n obseved chaactestcs by a stuctued om o statcaton Wh s
In teams when tyn g to dent y a poblem pec sely by oganzng ava abe knowledge and deas about the poblem Ths s/s not mat asks a sees o uestons that am to pnpont the pobem so gudng data colecton
sn f
(whrwh to what s tt or to whom s t occur
W
(th ocato whr t otc
Wn
(th ay hour moth or thvt rato to othr vt
W kn w u th typ or catory r z
W
(roup or vua prt or ar th vt Fgure
Exe /s rx
(whr wh s n tc. s n occur wh t cou
(what mht Tf pla th pattr
DAA COECON, ANAYSS AND DSPAY
u n
Figre
I
I n
T
fgr ad g th pa appar ymmtrca
jot
n
pparty radom; watg o pam of had
Cotuou
ky to b ratd to arthrt or rhumatm; uuua for phyca probm to b ymmtrca Brought o by vt
kn I w u
hootg pa; go o for varab tm
u pa or ach
o pattr
Brought o by pcfc dvdua
ot pury phyca coud b tr ratd
I/i t tri ed i edic dii
w us
Th rs t st is t inti fy th rbl m situatin r ia t b analys Thn ask a sri s f qusti ns as giv n in igur Bfs
This is/is nt matri allws th rganiatin f knwlg an infrmatin in a structur frmat This alws ata cllctin t b gui by rir knwlg Exam
A ctr is iagnsing a atint wh has rsnt with mystrius ains in hans an lgs Th ctr uss th is/is nt matri as a gui t iagnsis igur 2 Th ctr cnclus that th ains ar strss rlat an cnrms th iagnsis by taking th atints bl rssur
Rfnc rs H r d Bmi B rg ( 91 Nw Ronl Mng rit w Jrsy: J. bisig.
1 00 MET HOD S OR T OTAL QAL TY MAAGE ME T
1 86
Method 81
Mtix dt ysis
Pus
T rv a ctur f numrical ata frm a matri iagram n an fcnt way W us
It can b us t btain a ctur f fr aml iffr nt ruc ts an markt chara ctrstics It s an i mrtant mans f ana ysing mu ltvarat ata
w us
Its us i s quivalnt t rincial cm nnt analysis which is a mth f muti-variat statistical analysis It rqurs sm knwlg f statstical mths byn th sc f this bk
tomach pt
Bad
Brad
Brad 2
Brad Brad Brad
Fgure
Mar daa aay
ctv
DAA COLLECON, ANALYSS AND DSPLAY
Bfts
Hlful in analysing multivariat data It shws all th ky data clarly and rvids a schma f such things as diffrnt rducts and markt charactristics. Exam
Figur 1 rvids th rsults f an invstigatin int th ffctivnss f diffrnt brands f a sirin Brands 2 and 3 ar ffctiv and d nt caus stmach ust whras Brand 1 is ffctiv but causs stmach ust. Brands 4 5 and 6 a nt ffctiv.
Referene B. Bgma ad B. sjo ( Nw Yok MGawHi
Qy rom omr Nd o omr Son.
1 88
1 00 METH ODS FOR TO TA L QUALITY MANAGEMENT
Metho d 82
Matri x dia gram
Purpose
To provide information about the rel method elements of the subject.
ationship and i mportan ce of task and
When to use
rgest amount of data for the To organize and illustrate graphically the la ortance of logical connection between various elements to show the imp different relations by using graphic symbols How to use
The most commonly us ed matrix diagram is shown in quality function deployment (see Method 26) Benefits
Simple method to show relationship between task and subect
Operator
Supervisor
Product engneering
Yes Rejects
Machine produces rejects
Stop machine
Quality contro
Stop production
No
Analyse the reject Yes Repar possible
Repair the problem Sta
Ready for producton
Figure
1
Matrix diagram
Maintenance
Repair . machIne
J
Machne repared
DATA COLLECTO, AALYSS AD DSPLAY
189
Exam
igur mns trats a matr i iagram in th situat in whr a machin is rucing r jcts
90
00 MEH ODS OR OAL QUAL Y MANAG E ME N
Method 83
Movin g average
Pus
o ieni y ens in aa w hen shoem vaiaion o cyclical atens ae conusing he longeem icue. Wh t us
When monioing a ocess o eec changes o when a change has been mae o ocess inus o n ou whehe he ocess mean changes. Moving aveage cha s can be use o monio boh vaiables an aibues aa an ae aiculaly useul o eecing longem ens in aa w t us
hee ae seven simle ses involve
2 4 5
Collec he aa. he ye o aa coul be sales gues acoy volumes, eo ae o echange ae Decie which eio o ake he aveage ove. he aa have a 1 -week cycle a 13week eio woul be ao iae; i aa ae collece seve n imes a ay, a moving aveage eio o seven woul be coec. Accumul ae he s seven gues o a eio o seven an ivie by seve n. his is he s gu e o be lote. Remove he s gue om he calculaion, a in he ne an again ivie by seven. Reea se 4 unil he aa ae all use
6 Plo he aa he ime sequence is usually lote on he moving aveage is loe on he y ais. 7 nee he esuls
x ais he
Bfts
Moving aveage chas ae elaively insensiive o shotem changes his means ha hei main us is when you ae ying o mask sho- em vaiaion in he ocess o highligh longeem vaiaions Exam
A secialiy eamen was eamining acos aecting lan yiel ab le . chemical A movingeney ave age yiel was loe ove a 52- week eio his involve 1 1 2 baches he eio o he moving aveage was ou baches see Figue .
DAA COLLECON, ANALYSS AND DSPLAY
0 Q
Q
0
0
;
0
20 2 0 2 20 Batch igr
vin averae chart
abl Pant ied data in a chemica rener Batch
Bach no
Yild (%
Sum
Mov avag
9 9 95
5. 59 55
5 9 44
4 5
9 99 9 94
48 4 89 4
4 5 48
45 455 85 5
9
.8
8.
545
A lt f the raw ata ehibit a large amunt f lca l variatin, making the icture very har t ee. The mvng averag e am thi wn revealing the lngerterm icture. There ha been a te change in the ye at abut week 32 Thi crrene t a hyca change in the way that the temeratue wa meaure befre the rce wa te. The reult wa that the rce terminate at a lwer temerature imrving the yel
Referene R Caucut t 995 Aceng Qly Impoemen London: Chapman and Hal
1 92
00 MEHO DS FOR OAL QUA Y MAAGEME
Me th od 84
Mu tiv ari cha s
Ps
T w e pern n prce ve r e r n ng erm ng grpc cnr cr Wh s
I cn be e rng prbem ny ep e e ce f prbem n nern e prce wen eer be r nbe w s
2 3 4 5
Seec e p rce n cr cer c be ny e n mpe nerny Devep me f recrn g e mpe bervn C ec e mpe n recr e ve f e crcerc P ec pn n e cr n jn e we pn e ge pn w rg ne If e ne re f e m e eng n f e y re n e me reve pn en e prce cn be cnere be erwe ere n gnbe ce
Bfs
I ep nern e vrn n prce ver e r n ng erm Exam
ve pr were ken frm rnng pern every 25 mne Te meer f ec pr w mere n pe n e cr A rg verc ne w rwn rg ec e f ve pn (gre 1) In e mvr c r e pern ver e r me per be b e pern ver e nger per k pc In ny e eng f e ne ecrbe e pern ver r per f me n e cn f e ne nce e pern ver ng per f m e
DAA COLLECON , ANALYSS AN D DS PLAY
193
.
2
m Figu 1
Mli-ri chr
Maro PerezWison (1992) Miri hrt nd Anyi: Preexerimention Techniqe. Sosdae, Arizona: Advaned Sysems Cnsutn.
194
1 00 ME HODS FO R OAL QUALY MAAGE ME
Method 85
N cha
us
T ientify wen te nmber f efective tem in mpe f cntnt ize i cnging ver tme Wh us
Wen mnitring prce t etect cnge r wen cnge been me t prce inpt t n t weter te nmber f efective item cnge NP cr t re e wen te mpe ize i cn tnt
w us
Tere re even impe tep nvve: 1 C ect te t Drw p tb e w ing te nmber f efective tem fr ec t nmber Te nmber f efective item i ce n. Te tt nmber f t i ce m. 2 Pt te t frm te tbe nt te n cntr crt ee igre 1 n pge 96) e cc eve nmbe re wn n te riznt i t e nmber f efective nit i wn n te vertic i 3 Ccte te centr ine np !nm. 4 Ccte p npn. 5 Ccte te cntr im it wic re ± 3 bt te centr ine n
re ccte : pper cntr imit Lwer cntr imit
CL np + 3 x x LCL np
If te wer cn tr imit i e tn zer it tken t be zer 6 Drw te centr ine n te cntr imit n te cntr crt 7 Interpret te ret Bfs
It cn be ifct t eprte t rnm vritin ften ce cmmn ce r nn-ignbe vrtn frm re vr itin c e by cnge t te prce NP crt re wy f ing ti fr te nmber f efectve item wit mpe ize tt i cntnt
DAA COLLEC ON, ANALYSS AN D DS PL AY
195
Exam
A cmpny mnrng e nmber f defe cve bndng f rep r A bndn g defec ve wen e cnen verp e er cver A mpe f 100 repr ken n dy b nd e gre re fw:
Lot no 1 2 3 4 5 6 7 8 9
Defective laing (n) 5 6 5 4 2 2
2
10 11 12
13 14 15
16 17
18 19 20
T
34 3 4 3 2 3 2 4 3 2 60 Smpe ze 00 n 60 np 60/20 3 3100 003 3 520 0985 3 x 512 CL 3 512 812 LCL 3 512 0
=
Te cnr cr w e prce nder c cnr gre 1 n pge 196
Referene Ma Own ( 989b SP d mpem London Pubatons
a veage
Cl .
lC ·
aeae e
Fuency
s I o e " :
f<
' u
o o
Q
oP P
Dparen
r p c » r
s » z » G s Z
Sampe
DateTle
s or vroet shouldb od Ay Chage i peope, matenals Thes s l lp o k ecte opross pvt atIO wh saUed b th tol cha
Figr 1
NPhart
DAA COLLEC ON , ANALYSS AN D DS PLA Y
Method 86
Paynter ch a rts
Pus
T ispay infrmatin ver time in a way tat a ws canges in pattern s f faire t be iscvere Paynter carts wi sw w en ne faire me taes ver f rm ante r in terms f imprtance r wen t e vera faire rate is canging ver time W us
Wen mnitring perfrmance an setting prites fr imprvement Paynter carts can sw canges in p atterns f faire ve r ti me Tey can as be se t recr faires errrs r te ccrrence f events tat a team wises t sty an remey w us
Te nmber f times an event ccrs is swn n te vertica ais Te ifferent time peris are swn n t e riznta ais Te ccrrences are given in ran rer wit te ig est rs t Te previs ranin g is given fr eac ccrrence Te tta fr eac year is given in te etreme rigt an c mn Bfs
Paynter carts are abe t ispay a arge amnt f ata at ne time as ppse t aving a ifferent cart fr eac faire me r ccrrence By ispaying a te ata tgeter a team is better abe t see te tta pictre Exam
igre 1 sws a Paynter cart f te reasns fr scrap in an atmtive sppier Te riznta ais sws canges mnt by mnt in te patter n Te r aning wn te eft- an sie f te cart sws ca nges in imprtance f te fats It can be seen tat in te rst mnt 'N treas became mre prevaent Te Paynter cart as sws tat Ba spray is an increasingy imprtant faire me
Previos rakig 2 2
c' 0
0
escrptio thers Cocked gaskets o threads Bad spray eted bodies eakig seams ie bodies set gaskets Poor pi ccetic assy. ole i body eakig weds -sq. theads ight theads plit b 'plate otal scap otal pod
igr Pater chart
Ja eb ar. p ay Je Jly 00 00 2 28 2 20
2
2 0
g ept. ct. ov. ec.
ota
-
-
-
-
00 00
-
-
-
-
-
2 23 2 20 2 2
-
-
-
-
99
0
o o s I o .
r
p c » r
s » z » G s Z
DAA COLLECON ANALYSS AND DSPLAY
Method 87
199
chart
Puose
T ientify wen te percentage f efective size is canging ver time
items in a sampe f variabe
When to use
Wen mnitr ing a prce ss t etect canges, r wen a cange as been mae t prcess inpts t n t weter te percentage f efective items as canges. P carts are se wen te sa mpe size varies by mre tan 25 per cent f te mean sampe size ow to use
Tere are eigt simpe steps invve:
1 Cect te ata. Draw p a tabe swing te nmber f efective item s fr eac t n mber (see Tabe 1 Te nmber f efective items is cae n Te sampe size fr e ac t is ca e n Te tta nmber f ts is cae
m
2 r eac t c aca te te percentage efective % by iviing n by n 3 t te ata frm te tabe nt te cntr car t. Te sccessive t nmbers are swn n te riznta axis, te percentage f efective nits % is swn n te vertica axis. 4 Cacate te Tis is cacate as te sm f a te
ns ivie by
te sm f a te s an is written:
n/n 5 Cacate te centra ine as x 100% 6 Cacate t e cntr imi ts wic are ± 3 s abt te centra ine an are ifferent fr eac ifferent vae f n. Tey are cacate as: pper cntr imit (C L) 3x Lwer cntr imit (LCL) 3
x
If te wer cntr imit is ess tan zer it is taken t be zer. 7 Draw te centra ine an te cntr imits n te cntr cart. 8 Interpret te rests
200
00 M EHODS OR O A L QUAL Y MANAGE ME N
Table 1 Perceage o deecie ioice Suboup no.
Suboup s
n
%
C
C
19
8 2
37
62
40 25
40 79
0 0 0
8
68
0
1 2 3
20 8
4
13
upp conto imit C ow conto mt.
Bfs
It cn be fct t eprte t rnm vrtn (ften ce cmmn ce r nn-gnbe v rtn) frm re vrtn ce by cnge t te prce P crt re wy f ng t fr te nmber f efect ve tem wt mpe ze tt vre by mre tn 25 per cent xam
An eectrc trbtr ccte P crt fr te percentge f efectve nvce e by t nvcng eprtment Te gr e re gv en n Tbe n te cnt r cr t n gre
In 60 15 In 8 18/60 0.3 Refeene Ma Own 989b SPC and Contnuous movement. ondon IFS ubcatons.
!
p
a!n
n
U
Avag"
l
ag ap' S:
' <" :
D
o
P
»
r r
�z
» z » r
» z o o
Sampe
"
( np C)
7
{
opoo ( )
Dale/Time
A e e.e et, e evet hd ted Thee tew he t tke etVe veet we ed b e h
igur
Pchart
N o
202
00 ME HODS FOR OA L QUAL Y MANAGE ME N
Method 88
Pie ch art
Pus
A way f pcray repre enng aa pe car are an eff ecve mea n f wng e reave ze f e nva par e a Wh t us
Wen y wan epc e reave ze f nva par e a w t us
Tere are fr mpe ep n e cacan f a pe car 1 Cec e aa an preen n abar frm 2 Ta e aa em an caca e e percenage f eac em e we e percenage wen ae p m eqa 100 per cen 3 Cnver eac percenage a n e reevan prn f e crce. Snce a crce repreen 360 egree eac prn f aa w ccpy a ce f e we r examp e a gre f 84 2 per cen f e we cacae repreen a prn f e crce a fw:
)
ut % tal halth
igr
Pe chat of health athot ependte
DAA COLLECON, ANALYSS AND DSPLAY
360
x 842/00
203
303
4 Draw the pe chart usng a protractor to show the sze of the slces wth n the pe . Remember that the pe chat shows reatve values. f you want to compare several pe charts, or f the sze of the populaton changes, you must ensure that the relatve value of all d ata s consstent for each. Bfts
Pe charts gve an easly understood pctue that ads decsontakng Exam
The fol lowng expendture prole o f a health author ty s represented a s a pe chart n Fgure Acute servc es Mental health servces Servces for the elderly Communty servc es
£38.2 M £11.9M £11.9M £ 0M
4% .2% .2% 0.2%
199° 62 62° 3
Total
£69.0M
100%
360
204
00 MEHOD S FOR OA L QUAL Y MANAGEM EN
Method 89
Proc es s an alysis
Puos
Enables a grop t o look or oppor tntes to prove proc esses It can also be sed to denty standards and easres or crtcal parts o pr ocesses Wh to us
A proble-solvng tea old se pro cess analyss to dscover gaps dead ends or dplcatons n bsness processes and to so ere prove ents can be ade. ow to us
Ts sple tecnqe nvolves eanng all te steps n a process and evalatng eac o t e crtcally.
1
Dra an otlne ocart o te process Ts sold be done by al e bers o te gro p o st be represent atve o te departents to c te poce pae. 2 Eac tea eber ten dras a detaled ocart or s or er on part o te process Ts sol d ten be agreed by oter ebers o s or er depar tent . 3 Analyse te ocarts to look or dead ends dplcatons and parts o te process tat are tat ssng otenprocess neglected detal sc as standard nstrctons, leads Itto secess varablty 4 Plan to cange te process accordng to te ndngs n 3 5 se te ocart to denty at to easre and ere to easre t Tese e asres can be ed back an d sed to prove te proce ss Bts
Oten people n te sae departent beleve tat tey are carryng ot tasks n te sae ay bt en te process s eaned n detal t can be seen tat tere are crtcal derences An addtonal benetjobs s tat any processes are not desgned bt develop t peoples Freqently no one callenges eter te process tsel s stll necessary or eter te steps gve te best ay to acev e te desred res lt
DAA COLLECON, ANALYSS AND DSPLAY
205
Exam
A company noticed that there was an apparenty random probem wi th the dispatch of fauty products. his coud not easiy be traced to shifts, products or tim es he pro bem-so ving group dre w an out ine owchart as shown in Figure 1 (on page 206 he two then drew their detaied owcharts for their ownparts part ofofthe thedepartment process. he detaied owchart for the second part is shown in Figure 2 on page 20 When the detaied information was coected it became apparent that specications were missing for key quaity characteristics. his ed to different operators interpeting the requirements in different ways and therefore the apparenty random pattern. he group was abe to put in pace the missing information and the probem disappeared
Refeence G Bon 994 Pocess Managemen ua Impovemen Nw Yok W
206
1 00 MEHO DS FOR OA L QUAY MANAG EM EN
Cold ed spet wae
o
Pat of poess de old ed oto
ejet afeted wae fom leh
eeze wae paed
ae oetve atio at hot ed
o
Pat of poess de hot ed oto
ejet wae at hot ed
Fg Otlie owchart to reet critical defect (CD) ito aced ware Sorce: Kaji ad Ashe 3
DAA COLLECON, ANALYSS AND DSPLAY
Cld end inspect pack are
2
reeze 0 min. packed are
nrm ht end t reject munt rm sectin until clear
eject mul d rm lehr until gap appears
btain engineering help t reciy #55 (i applicabl e)
esrt / check rzen pal lets
g incid nt details and actins taken
reeze urther 0 min packed are
Fgr 2 Detailed owhart (reventon of CDs nto a) of area nder old end control Sore Kaji ad Asr, 3
208
00 MEH ODS FOR O AL QUAL Y MANAGEM EN
Metho d 90
Proce ss cap ab il it y
Pus
To demonstrat e whether a proces s is capab le of meeting a specication and to calculate an index to show this apabilit Wh t us
During design commissioning or postcommissioning to demonstrate the capabilit of a product or process w t us
The calculation of process capabilit depends on whether the process is measured in terms of variables data or attributes data For vribl d, capabiit is expressed in terms of an in dex caled is dependent upon the actual centre of the process performance being the same as the centre of the specication When this is so the index is calculated as: _ Total width Total process spread where the total process spread is estimated using the formula Total process spread 6 d where is the mean of the ranges (see Method 6) and constant depending upon the same sie as shown below
2 3 4 5
d is a statistical
.28 .63 2.05 2326
arious values of together with the parts per milion produced above specication are as follows
(ppm) of items
DATA COLLECTON, ANALYSS AND DSPLAY
.00 .33 1 2.00
ppm 350 32 0.3 0.001
or attbut data, th apabilit of th pro ss is oftn alld th rst run apabilit (RC) and is alulatd as RC ( ) X 100% whr is th prntag of dfti itms (s Mthod 8). Bfts
B stabl ishin g th fats about pross prf orman it bom s asir fo r thnial , markting and m anufaturi ng dpartmnts t o talk to ah ot hr and for ompanis to talk to thir ustomrs and supplirs Exm
During th manufatu r of glass bottls, th intrnal diamtr of th bottl nk is a ritial paramtr Th spi ation of th nk i s agrd wi th ustom rs as 1 5 m ± 0 01 m Th total spi ation sp rad is thr for 002cm. Bas d upon a st of sampls of si z 5 th man o f th rangs was 0005 m Total pross sprad was alulatd as X /d2 0.01 9. Th apabil it was thrf or 0.02/ 09 1 035 Using th gurs gin abo , i an b sn that approimatl 350 ppm ar bng produd out of spiation.
Rfnc Ml Own 989) SP nd un mvmn. London S ubictions
20
00 MEHODS FOR O AL QUAL Y MANAGE ME N
Me thod 91
Sa mpli ng
Puos
A mthod by which a small nmbr of itms (th sampl) is drawn from a largr nmbr of itms (th poplation) in ordr to draw a conclsion abot th poplation basd pon information from th sampl. Wh to us
Sampl ing is sd whn th ovra ll siz of th poplation is sch that to gain fll information wold b impossibl, timconsming or vry costly. ndr som circmstancs, sampling can b mor accrat than infor mation obtaind from th total poplation. ow to us
Thr ar v common mthods of sampling in common sag. Random ampling
This is whn th samp l is dsignd so that vry mmbr of th poplat ion has an qal probabilit y of bing chosn. It is sal to do this sing ran dom nmbrs to tll yo which itms to slct. Random sampling is sd wh n thr ar no known pattrns or trnds in th data. Th disadvantags ar that it can b costly and that, if th poplation is ill dnd, it is impossibl to nmbr th itms to allow th sampl to b dsignd tratied ampling
This is whn th sampl is dsignd to rct prior knowldg abot th poplation. An ampl of this is social trnds whr grops A, B, C D and E ar chosn. A random samp is thn takn within ach grop or stratm Stratid samping givs mor prcis information than random sampling. Cluter ampling
This is whn th poplation is split into grops or clstrs and thn a random choic of clstrs is mad. Random sampling thn taks plac within th clstr.
DAA COLLEC ON , ANALY SS AN D DSPLA Y
21 1
An eampe o this is the use o opinion p os to predic t the outcome o an eection by picing poing stations and then carrying out a random sampe o tho se eaving the poing station. Sytetic li
his is sim py when the popuat ion is beie ved to be rando my spread a nd a samp e o 1 in n is tae n. An eampe woud be tai ng a sampe o students by taing every th student rom the ist o names Qot li
his is a orm o stratied samping where those taing the sampe are given a quota to u . hey interview everyone ttin g the quota unti the quota is compete his is common in maret research where peope have to interview so many mothers with chidren etc. Whicheve r method o samping is used the proces s to be oowe d is the same 1
Dene the popuation rom which the sampe is to be taen his is the most dicut step Most probems resuting rom samping are caus ed by ac o carity at this stage. 2 Dene the samp ing technique his may be aready nown rom inormation on the popuation. 3 Dene the sampe size. 4 ae the sampe . Bfts
Samping is a aster process than using the whoe popuation due to the smaer amounts obedata being coected It is aaso cheaper because ess y. time is used. It can more accurate and have nown degree o accurac Exam
As part o the diagnostic phase o a tota quaity proces s an organization used strat ied sampi ng to determine how many empoyees to intervie w to nd representative views o quaity probems in the organization.
Reference A Pccl Gud o Ssc M.R Bgd RJ Mklk d BA Olso (199 Pocss Imomn Amstd V ostd Rhold
212
1 00 M ETODS OR T OT AL QUALTY MAAGE ME T
Method 92
Sca tte r di a ram s
s
To allow th rlatonshp btwn as and t to b stablshd W s
Sattr dagrams ar sd whn a grop s trng to tst whthr a rlato nshp sts btwn two tms otn a as and t w s
Thr ar or sm pl stags to dra w a sattr dagram 1 Collt data abot th ass and ts 2 Draw th as on th horontal as 3 Draw th t on th rtal as 4 Draw th sattr dagram
0 •
2
20
• •
0 •
•
•
0 Fgue
•
Scer digr
•
2
lege
DAA COLLEC O AALY S S A D D SP LAY
23
Bf
Scae dagam help o ng he ac o ea when dcung p olem The help o educ e he am oun o gu eelng n volved wh he polem olvng wod o cauon ecaue appea o e a elaonhp doe no mean ha oneJu hng cauehee anohe; he elaonhp mgh e ouou o h ough a hd unnown vaale I a elaonhp app ea o have een ound poo mu e ough Exm
eam examn ng he elaonhp eween he ml eage o alemen and he volume o une old Mleage he caue ploed along he hoonal ax and ale pe annum he eec ae ploed along he vecal ax The cae dagam how a clea elaonhp eween mleage and ale Fgue 1
Rfrnc P Lyonnet 992) Toos of Tota ua Lonon Capman an Hal
214
1 00 MEHODS FOR O AL QUAL Y MANA GE ME N
Method 93
Spider web diag rams
Pups
To sow perorma nce against a target wen see ral criteria are being set . Wh t us
Wen team members need motiation to aciee resuts tat tey migt preiou sly ae belieed were impossib le , one way to do tis is to sow ow tei r perormance compar es wit te b est acieable Spid er web or aracnoid) diagrams gie a ery isible way o sowing progress and perorm ance against seeral target s at te sa me tim e.
(
Procedre developmet ( 00%)
ditors traied (00%)
People traied (00%)
dit completed ( 00% o tme)
tstadig corrective actio reqests
ocompliaces fod
Best i class tates of cret peformace
Fgue
ider we digrm
DAA COLLECON, ANALYSS AND DSPLAY
215
ow to us
For ech of the preters or trget s invo lved identify precise denition of the trget nd how it cn be esure d Mesur e the perfo rnce ginst the trget nd disply it on the digr. Ech tie the perfornce is esured new digr i s constructed Bfts
The spider web digr shows t glnce how progress is being de towrds trgets. When ben chrkin g, the spider web digr cn be used to show current perfornce, the iedite i the verge in clss or the overll best in clss perfornce Examp
A copny decided to use the spider web digr to show the develop ent of its qulit y syste ginst the trge t set (Figure 1 ) The pre ters nd esures chosen were • • • • • •
procedure developent % developed on tie internl uditors trined % ginst trget internl udits copeted % ginst trget non-coplinces found totl nuber corrective ction requests outstnding nuber people trined in the syst e % gi nst trget
26
00 METHO DS FOR TO TA QUAT Y MAAGEM ET
Method 94
Statistical process contro (SC)
P
To ident iy wen procee are canging over time Wh t
e n monito ring a proce to detect cange or wen a cange a been made to proce input to nd out weter te proce output cange Statitical proce control ( SPC) can be u ed or manuac turing and ervic e procee uing te owcart o te proce a a bai w t
Tere are eigt imple tep involved 1 Ide ntiy te proce to be tud ied. 2 Draw an outline owcart o te proce. 3 Draw a detailed owcart o te proce 4 Uing t e owcart d ecide wic da ta to collec t Uing tally cart or concentration diagram collect te data 6 Analye te data uing iogram or oter tecnique to enure tat te data are uitable or control carting. 7 en neceary, ue problem-olving tecnique uc a Pareto anal yi and caue and eec t ana lyi , to remove pecial caue variation rom te proce 8 Draw te data on a control cart Tere are two baic type o data
Vaiable data wen te caracteritic being meaured i continuouly variable over a range o value Ti migt be temperature weigt output time to proce a orm 2 Attibute data wen te caracteritic being meaured a one o tw o value Ti migt be on/o preent/abent go/no go. Attribute data come in two orm • defective unit or exampl e orm wit error deliverie wit ault
•
etc Defect: or example, error in orm were tere can be more tan one error per orm ault in ipment were ipment can ave everal error.
DATA COECT O , AAYSS A D DS P AY
217
Bf
I t can be dfcult to separate out random araton (often called common cause or nonassgnable araton) from real araton cause d b changes to the process Con trol charts allow decson s to be made about processe s on the bass of fact rather than gut feelng Exm
gure on page 2 ) shows how to select th e correct tpe of contr ol chart accordng to tpe of data and s ample s e.
Refeence S Oaklad ad R. llwl (10) Hma
Sascal Pocess Conol Ld Brwrh/
o o Variables
es
m
s I o
0
m
ttribtes
Xad MR hat
r C l r
s l z l G s Z
Fgr Chsig the right ctrl chrt
DAA COECON, ANAYSS AND DSPAY
Method 95
29
Stem and leaf diagram
Puos
o present raw data and to show ther dstrbuton vsually Wh o us
As a tool n exploratory data analyss such as hstograms. It s useul or llustratng arge amounts o data ow o us
Round numercal data to th e number o tens by trm mng o the last dgts. In gure 1 t he ascend ng order o the data s 3, 5, 6, 7, 7, 8, 8, 9 33, 35 he last dg t o each observaton s call ed a lea and the oth er dgts or m a stem . or an observaton o 35, the lea s 5 and the stem s 3 Bfs
1
A stem and ea dagram sdew ays provdes a hstogram o t he data wth numer cal values 2 he me thod can be mod e d to sut derent data 3 uartles e . , etc. can be read drectly. 4 he dagram gves a useul venumber summary e lower quartle medan , upper qu artle the smallest and the target value.
ens 0 0 2 2 tem
Fgr
ts 0 22 0 2 0 2 eaf
Stem ad eaf diagram
otal 0 2 " 0
220
1 00 MEHODS OR O AL QUALY MANAGE ME N
Eam
A sem and e a dag ram s gen n Fgure 1 The sem n h s ase s he ens and o he rgh are he un s Th e op a lue s hus 3 and he boom alue s 35 The dagram sdewas also prodes a hsogram wh he numer al alues an d > and
Refeence B. Begman and B. Klfsj 994) Nw Yk McGaw-il.
ualt from Customer Needs to Customer Satsfato.
DATA COLLECTON, ANAL
Me thod 96
YSS AN D D SPL AY
221
Tall y ch ars
Purpose To collec t data when the value of a defec t or probl em is important When to use Either during problem denition when you are collecting data to nd out what is happenin g, or when you have imp lemented a solution and you a re collecting data to monitor the new situation How to use There ar e ve simple steps to draw a tally chart 1 2
Agree the data to be collected. This step is vital: you cannot analyse data that have not been collected Design the tally chart Type of call
Tally
Stationery 've done t wrong
I
9
\ I
9
.
..
How do I . ? Soft crash
)
Other soft
J
5
Can I have . . ?
w
5
Code numbers
.J 1\
11 6
13
Hardware
I
3
Other int
P
6
Other ext.
I \ I
4
Total Figur 1
Frequency
Tally chart analysis of telephone calls
71
222
00 MEHODS OR O A L QUA LY MANAGEM EN
3 Tst t art using somon wo as not bn nvovd in t dsign Gt m or r to us t art witout ass istan If nssa ry, mod ify t art 4 Dsign a mastr ta y art. If mor tan on prson i to b invo vd in data otion you wi n d to bring tog tr a t data otd
T wayt to data do tis is to us a mastr art Cot
Bt
By stabising t fats about t vau of faurs a tam an pan to idntify t auss of faiur and ook for ways of rmoving tm Ations ar takn on t bass of vidn not fng Tay arts ar an nt way of invoving pop n a aras in quaty i mprovmnt Ty provid a simp mtod of data otion tat an b asi y undrsto od and appd n of and wo rk aras Oftn tr ar simp ways to draw tay arts witout asking pop to rord t rsuts Marbs or gof bas an b usd to r ord vnts vry sim py Exam
1
T tay art sown in Figur rprsnts t rsuts of an anayss of as to a ompanys omputr p dsk T data wr usd to assst n t dsign of indution trainng
DAA OEON ANASS AND DSA
Me thod 97
223
Tree dia gr ams
Pups
o identi y the task s and methods needed to solve a proble m and reach a goal Wh us
When you want to break down vaguely or mulated customer wishes about a product into customer requirements on a manageable level or to investigate all possible aspects o customer wishes that present a problem ree diagrams can be used to develop short-term goals beore nalizing long-term goals. w us
Build a 'tree sy stematicall y rom a statement o a goal thro ugh headings to 'branches o plans and action. rees are best developed as a team General issues are subdivided into specic issues A particular us e o this structure is that o a 'ault tree where the tree structure is used in a cause and eect ashion to analyse causes o aults.
Case
asy to an dle Poblem
asy to adst Case B
asy access Case C
Figr 1
Tree digrm
asy to o ld Case asy to gri p Case impe mecanism Case B imple eqipment Case B mal andle Case C dstabe grip Case C
224
1 00 M EHOD S FOR OA L QUALY MAAGEM E
Bft
To solve a coplcae probe an o acheve he esre goa n a ssea c wa. Eam
Wh he a of a re e agra a copan s seekng o e sgn equpen ha s eas o hanle gure ) .
Reeece B Brmn n B. Kjo (14) uat: fom Custome Needs to Custome Satsfacton. N York MGr-H.
DATA COLLEC TO , AA LYSS A D DS PLAY
Method 98
225
U ch a
s
To ideni when he numer o deec in a ample o variale ize i chang ing over ime The U char coul d e ued o monio r error in ping error on drawing mark in maerial Wh s
When monioring a proce o dee c change or when a change ha ee n made o proce inpu o nd ou wheher he numer o deec alo change U char ar e ued when he amp le ize varie more han 25 per cen o he mean ample ize w s
There are even imple ep involved 1 Collec he daa Draw up a ale howing he numer o deec or each lo num er The num er o deec i calle d . The ample ize o r each o i called The oal numer o o i called m U dividing 2 For each lo calculae he numer o deec per uni 3 Plo he daa rom he ale ono he U conrol char The ucc eive lo numer are hown on he horizonal axi he numer o d eec per uni U i hown on he axii calculaed a he um o all he 4 Calculae he cenre lnverical e U Thi divided he um o all he and i wrien a
U
/
5 Calculae he conrol limi The conrol limi are ± 3 d aou he cenral line and are dieren or each dieren value o The are calculaed a pper c onrol limi ( CL) Lower conrol limi (LCL)
U 3 U 3
I he lower conrol l imi i le han zero i i aken o e zero 6 Draw he cenral line and he conrol limi on he conrol char 7 Inerpre he reul
226
00 M EHODS FOR O A L Q UAL Y MANAGEM EN
bl 1 Rport sht rrors D 3/5 4/5
6/5 9/5 10/5 1 1/5 12/5 13/5 16/5 1/5 1 8/5 19/5 20/5 Tol
hi
No shs n
No. os
Eos/sh w
6 x 2 2 x 10 6 2
3 3 3
6 5
200 1.33 1 66
2 6 6 2 6 2 6 2 6 6 6 2 6 2 6
x x x x x x x
10 2 2 10 2 10 2 10 2 2 2 10 2 10 2
1 3 3 3 3 3 3 3 1 3 3 3 2 3
3 0 6 8 1 3 4 2 2 1 0 3 6 2
300 0 200 266 066 1 00 1.33 0.67 0.6 1 00 033 0 1.00 2.00 06
2 6 2 6
x x x
10 2 10 2
3 3 2 3
3 2 1 3
1 .00 06 050 1 00
22
�C
58
66
Bfts
It an be difult to separate out random variation often aled ommon ause or nonassgnable varation ) from real variation aused b hanges to the proess. U harts are a wa of dong ths for the numb er of defe ts with 5 a sample sze that varies b more than 2 per e nt. Examp
For data on report sheet er rors see Table 1 . T he resutn g U hart is given n Figure 1.
- �C 66 �n 58 114
CL U
3 32
upper ontrol lmt
lower ontrol imit
Yn
Reference TP. Ryn (1989) Statstica Metho fo uat movement Nw Yok: Wiy nsin.
P
t
Pia'l
Averag.
U.
np
lC.
TtSa Sz:
c
o
�
( r r
�
c' 0
l Z l r
Same I Nmb { c) oto (pu)
\ 3 b 3 J \ . J G 7
l Z o o
Dafme
y hag matas, m ms vrmt shu b t s s wl h yt akrctN9 ss mrvmt ac wsga y ctl chart
Figr
hrt
N N
228
MEHODS FOR OA QUAY MANAGEMEN
Method 99
X mov ing
range
(X-MR) cha
Pus
To denfy when a vale s hangng over me Wh t us
Whe n monorng a proess o dee hanges or when a hange has been made o proess nps o nd o wheher he mean vale hanges X movn g range s se d o monor varables daa when he vales an vary over a onnos rang e and when he sample se a eah samplng pon s Ths for exam ple s he a se wh monhly sales gr es when he re s only one gre prodon gres or when he os of akng he measremen s very hgh. w t us
There are sx smple seps nvolved 1
Col le he daa . Draw p a able showng h e readngs for eah of he samples a eah lo. The nmber of samples n eah lo s 1 The oal nmber of los s alle d 2 For eah lo al lae he movng range by sbrang he X val e from e prevos vale here s no movng range or he rs X vale. 3 Workng down he able add ogeher all he X vales and dvde he resl by he nmber of los Ths vale s he grand mean and s wren X (alled X bar) Agan, workng down he able add ogeher all he movng range vales and dvde by - Ths vale s wren 4 Cal la e he onrol lm s for he X- har sng he form la ppe r on rol lm ( C) ower onrol lm (C)
X + 3 X /d X 3 X /d
d s a onsan ha whe n ml pl ed by 3 gves 3 sd for he X har Plo he daa from he able ono he X- onrol har The sessve lo nmbers are shown on he horonal axs; he X vales are shown on he veral axs. 6 nerpre he ress Bfts
an be dfl o separae o random varaon (ofen alled ommon ase or non-assgnable varaon) rom real varaon ased by hanges o
DATA COLLECON, ANALYSS AND DSPLAY
229
abl Time o acieve le tan 5mba peue Sht ette
Cs
e <5
C
1
9.0
A A A
2 3 4 5 6 8 9 10 11 12 3 4 5 6
5.5 .0
B B B B C C C C D D D D D D
A A A A A B B
8 9 20 21 22 23 24 25 26
7. 5
0 80 65 60 65 90 80 9.5
90 6.5 55 60 .5
7.5
M e
3.5 1.5 05 0.5 1 0 15 0.5 05 2.5 0 0.5 05 2.5 0 05 5 0 2.5 0.5
9.0 95 90 80 105 85 1.0
2.0 25
7.5
3. 5
05
O 1.5
he poess -MR has ae a wa o dong his o vai ae s da a wih a sampe sie ha is 1 . Exam
The ime o ahieve a pessue is shown o 26 suessve poduon uns (Tae 1. Sne hee s on one suh me on eah poduion un he MR ha s he appopae wa o ono his poess (Figue 1 on page 230.
26 I 20 5; X I/m 79 IR 35 = IRm 1 = .38 Fo n 2 d 1128 heeoe 3 /d = 367 UCL X + 3 /d 153 LCL X 3 /d = 1 m =
Reerence R Clt 1995
Ahe Qat Imremet Ld Chp d
Plant
ChaaceiSTI
i
� :
' <"
cfica,
amlg - Fequecy
=
5
:
o o
N W o
I o
1\ '
R
Deparme
r p C l r
=
�
3
p
me Dae Lases is es
A Ay Change i n people. mateas. equent methods or environme shoud be noed These noes wl crecte hep ou too processmproemen aco whe sgaled the trol chart
Fgr
XMRchart
l Z l G Z
DAA COCON, ANAYSS AND DSPAY
Method 100
23
XR cha
Pus
To identify when the mean alue or range in a sample of constant size is changing oer time Wh t us
When monitoring a process to detect changes, or when a change has been made to process input s to nd out whether the me an or range changes X-R charts are used to monito r ariables data , when the al ues can ar y oer a continuous range, and when the sample size at each sampling point is greater than
l
Eamples of data drawn on an XR chart include t ime to process insurance claims, temperature of a chemical reaction and width of metal slabs w t us
There are see n si mple steps ino led: 1
Collect the data. Draw up a table showing the readings for each of the samples at each sampling point The number of samples at each sampling point is called n The total number of sampling points is called m For each sampl e, cal culate the mean X by addin g together the X alues and diiding by the sample size n For each sample, calculate the range by subtractin g the small est alue in each sample fro m the largest. This alue i s called R. 3 Working down the table, add together all the X alues and diide the result by the number_of sampling points m This alue is the grand mean and is written (called X bar bar) Again, working down the table, add together all the R alues and diide by m This alue is written R and called R bar 4 Calcula te the control limit s for the X chart using the formula Upper c ontrol limi t (UC ) ower control limit (C)
= =
X A X R X - A X R
A is a constant, that limits when multi s dformula for the X chart R, gies Calculate the control for thepliedRbychart using 3the Upper cont rol li mi t (UC)
=
D X R
232
1 00 METH ODS FOR TOTAL QUALITY MANAG EM EN T 205
20.0
igure 1
-----------------------------
Line graph of
4 .5
1 .0 0.5
Fgure
2
Lne graph of R
Lower control limit ( LCL)
6
7
=
0 X R
R chart. 04 and 0 are statistical constants that give ± 3 s.d. for the R. These limits are not symmetrical about the central line Plot the data from the table onto the X-R control chart. The successive sampling points are shown on the horizontal axis, the X and R values are shown on the horizontal axis. Conventon all y the R chart is plotted below the X chart. Interpret the results.
DAA COECON ANAYSS AND DSPAY
233
ab Dt usd to moor thkss of mhd prts
L n
3
4
5
Sum
Man
Rang R .8 8 6 1.3 5
�
97 . 97 18.6 .
+
8 8. 18.6 19.8 18.6 19.5
976 97. 975 9.5 975
195 19. 195 95 189 95
8 8
+
8.8 83
957 9
9 896
.1 .9
Tal
567
57
Man
89
19
89 99 19 186 8 181
3 5 6
1. 98 19.5 5 8
+ + + +
9 3
181 .9
. + 19
+ +
98 19. 19.5 99 8.8 19.
+ . 8
+ +
97.5
2.7
Bf
t can be dfcult to sepaate out ando vaaton (often called coon cause o nonas sgnab le vaat on) fo eal vaat on caused by changes to the pocess X-R chats ae a way of dong ths fo the vaables data wth a saple se that s geate than Exam
gues and 2 show the u se of an XR chat to onto the thckness of achned pats usng the data gven n Table The nube of saplng ponts m = 30 the se of each saple n = 5
30n = 5 R = m 57 30 = 9 X mX = 567 30 = 8 9 L D4R = 2. X 9 = 4 LL D3R = 0 x 1. 9 = 0 A2R_ 058 x .9 = . Lx = +AR = 89 11 = 20 LLx A2R = 189 . = 7.8 m =
R
Reerence JS Oakland and RF Fwl 99) Hnmann
Ss Pss Cnndn Buerwrth/
REFERENCES
Akao, Y ( 1 991 Hoshin Kani Polic Deploment fo Successful TM Cambidge Maachue: Poduciviy Pe Amden DM , Bue E and Amden RT ( 1 99 SPC Simpled fo Sevices London: Chapman and a Andeon D R Sw eeney .1 and Wiliam TA (1994 An ntoduction Management Sciences. New Yok We Bake Thoma D (1994 ualit b Epeimental Design. New Yok: Macel Dekke ( 1 992 A Pactical Guide to Statistical Beauegad MR, Mikuak R. and Olon BA Pocess mpovement Amedam: Van Noand Reinhod. Be D McBide P and Wlon G (1994 Managing ualt London: Buewoh/ einemann Begman B. and Kefo B ( 1 994 ualit fom Custome Needs to Custome Satisfaction New Yok McGaw-l de Bono, Edwad (1970 Lateal Thining. Nw Yok ape and Row Bon, G (1994 Pocess Management to ualit mpovement New Yok Wley. Boe , 1. ( 1 99 1 ualit unction Deploment a Pactitiones Appoach Mlwaukee ASQC Quaiy Pe Boun T, Main V and Race P (1993 oshops that o New Yok: McGaw ill. Buzan T (194 Use ou Head London Ae Book Camp , RC . ( 1 989 enchmaing the Seach fo ndust est Pactices that lead to Supeio Pefomance. Miwaukee: ASQC Pe Cau\cu R (1995 Achieving ualit mpovement London Chapman and all (1993 Continuous mpovement Tools Cafonia Richad Chang R and Ndzwieck, M Chang. Cook, Saah (1992 Custome Cae New Yok Kogan Page Coby PB (1979 ualit is ee Nw Yok McG aw-i Coby PB (1984 ualt without Teas New Yok McGawl Dahgaad GK andin Kienen K and(1990 A compaave udyManagement, of quaiy Total ualt conol mehod and pincipe apan Koea Dnmak' 1.1 Kanji 1 1532 Dae B G and Plunk e, 1.1. (1991 ualit Costing London Chapman and all Day RG (1993 ualit unction Deploment Lining a Compan with its Custome Milwauke: ASQC Quaiy Pe Deming W ( 1 98 Out of the Cisis Cambdge, Maachue MT Pe Fodyce K and Wel R. (1978 Managing with People New Yok Addion-Weley Gbb Lance (198 Too fo poblem ovng London PA Conuing Goup nenal epo uchn D. (1985 ualit Cicles Handoo. London Piman ma, M. (198 Kaizen the Ke to Japans Competitive Success New Yok Ra ndom ou e hkawa, Kaou (1985 Guide to ualit Contol Tokyo Aian Poducivy Pe ohn PWM (1990 Statistical Methods in Engineeing and ualit Assuance. New Yok Wiey Kane, Vco E (1989 Defect Pevention New Yok: Mace Dekke Kani GK (1993 Statistical Tests London Sage (1993 Total ualit Management Pocess a Sstematic Kan GK and Ahe M Appoach Oxfod Caax
REFERENCES
235
Kaf, B. and Ostbm, S 994) Bchmarkig a Sigos o Excc i Quaiy ad Produciiy New k: Wley Kene CH and Tee B.B 98) Th Nw Raioa Maagr. Pinetn New esey: .M. Publishn Kme, H 98) Saisica Mhods for Quaiy Imrom.Tky AOTS Lthetis, N and Wynn, H 99) Quaiy hrough Dsig. Oxd Oxfd Sene Pubiatins Lynnet P 99) Toos of Toa Quai.Lndn: Chaman and Hall MAndew and OSan, S . 993) FMEAs: a Maagr's Hadbook. Lndn Stanley Thes Nakaima Seiih 988) Iroducio o TPM Cambide Massahusetts: Pdutity Pess Oakand, S and Flwel RF 990) Saisica Procss Coro. Lndn: Bttewth Henemann Onnes A 99) Th Laguag of Toa Quaiy. Casteamnte, Itay: T Pk Pubiat in n Qualty Owen, M 989) SPC ad Busiss Imrom. Lndn: IFS Pbatins Pake, W. 99) Achiig Cos Eci Quaiy. Lndn: we PeezWsn, Mai 99) Muiari Char ad Aaysis a Prxrimaio Tchiqu Sttsdae, Aizna: Adaned Systems Cnsutant. Rss P 988) Taguchi Tchiqus for Quaiy EgirigNew k: Maw -Hi Ryan, TPP 99) New k Wi ey I ntesene 989) Saisi a Mhod for Quaiy Imrom. Seney, WordcCass Prformac hrough Toa Quaiy. Lndn: Chaman and Hal Tukey , 99) Exoraory Daa Aaysis. New k Addisn-Wesley
DEX
accptabl qality lvl (AQL) aity diagram 23 all work is procss 3 aalytical mtods 10 appraisal 85 arrow diagram 25 Asr bar carts 154 basic statistics 56 bc markig 27 box ad wiskr plot brai writig 125
20
159
123 braistormig brakig st 127 BS 60001 22 bzz grops 129
C cart 61 cas ad ct aalysis 79 cck sts 164 clilik istogram 178 clstr samplig 20 comblik isto gram 177 comparativ bc makig 27 comptitiv bcig 27 coctratio diagrams 166 cocpts or improvmt 2 coormac 27 cotigcy plaig 32 cotis improvmt 2 cotios improvmt cycl 5 cosss racig 30 cor cocpts or improvmt 2 critria tstig 35 critical pat aalysis 81 cstomrs' cotigcy tabl 37 cstomr satisactio 2 csm cart 168
dot plots 171 dobl samplig
2
rror proog (Pokayok) 43 voltio opratio (EVP) 89 ailr 85 ailr mod ad ct aalysis (FMEA) 91 alt tr aalysis 96 owcart 177 orc aalysis 44 orc d aalysis 98 Gat cart s 46 gomtric movig avrag
175
istograms 177 Hosi Kari 182 ida gratio 1 1 ida writig 130 imagirig 32 improv itral procss pla 134 ir distrbacs 112 itrmdiat impossibl 136 itral cstomrs ar ral 3 is/is ot matrix 184 IS (9000) 48 Jst i Tim 50 Kaiz 5 1 Kaji 1 Latral tikig 136 List rdctio 138
data collctio aalysis ad display 1 1
maagmt by act maagmt mtods maactrig variatio 112 matrix data aalysis 186
1 dligt cstomr migt w l 39 dpartmtal cost o qality 85 dpartmtal prpos aalysis 41 domaial mappig 87
matrix diagram 3 188 masrmt mid mappig 140 mit aalysis 101 mixig mtapors 136
NDEX
morphological forced connection moving averages 190 multi-vari charts 192 multi voting 1 multiple sampling 21 mystery shopping 52 nominal group technique nonconformance 107 normal inspection 20
12
15
objective ranking 5 opportunity analysis 17 outer disturbance 112 P chart 199 paired comparisons 103 parameter design 105 Pareto analysis 56 Paynter charts 197 PDCA 39 people based management 2 people make quality pie charts 202 Popkayoke 3 potential problem analysis 59 prevention 5, 85 problem prevention plan 61 process analysis 20 process capability 208 process cost of quality 107 process decision programme chart 63 programme evaluation and review technique 65 pyramid model 6 pyramiding 151 quality circles 67 quality function deployment (OFD ) 89 quaity management system (OMS) 8 quality policy deployment 182 quality policy manual 9 quality procedure 9 quality records 9 quota sampling 211
237
random j uxtap osition 136 random sampling 210 reduced inspection relation diagram 72 reliability 110 rich pictures 19 robust design (offline quality control) 112 sampling 210 scatter diagrams 212 sequential sampling 21 single sampling 21 skew histograms 117 snowballing 151 solution effect analysis 113 spider web diagram 21 statistical process control (SPC) stem and leaf diagram 219 stratication 115 stratied sampling 210
216
153 suggesstion schemes system design 117 systematic sampling 211
Taguchi methods 119 tally chart s 221 team wor k 7 tightened inspection 21 tolerance design 122 total production maintenance 75 tota l quality management principles Toyota Motor Company 50 tree diagrams 223
U chart
225
variable data
216
whyhow charting 77 work instructions 9 X moving range (X-MR) chart R chart 231 zero defects 78
228
