Warehouse Activity Profiling and Data Mining

First page Table of contents 1 Next page Last page

Animated publication

Warehouse Activity Profiling, Pattern Recognition, and Data Mining

1. The P’s and P’s of Profiling 2. Sales Order Profiling 3. Purchase Order Profiling 4. Inventory Profile 5. Calendar-Clock Profile 6. Summary

Warehouse Activity Profiling and Data Mining | Copyright: RightChain™ Incorporated | All Rights Reserved 1. The P’s and P’s of Profiling Ir ehcaevnet lbyeseunmt emaac rhiiznegd atchtei vtiet ay cphri onfgi l ai ns gt haen d“ Pd’ sa taan md Pi n’ si nogf fPorro mf i loi nr ge ”t. h a n t w e n t y y e a r s . I 1.1 The Power of Profiling Wt h ihnekni nI gwi sr ittoe rae na de ww ha ratti colteh oe rr pb eooopk l, eo hnaevoef wt hrei t tf ei rns ta bt hoiuntgtshIe dpoa rt ot i csut il ma rutloapt ei cm. Iyf Io wa mn ph ar evpe apr irnegp at or et de aoc hn at hc el a st os poircatsoe ms t ii mn aurl,aItdeomt hye tshaimn kei nt hgi nagn, dI rteov iaevwo iwd hraeti novt he enrt ipnegotphl ee wt hheedeal .i Al yc taicvtiitvyi tpyr ooffi ltihneg wwaorrekhs otuhsees. aAms eywo ua ys;t ai tr its tloi kleo or ekaadti nt gh et hper joof ui l rens aol fwcruist tt eonmbeyr Sy uo up paorsrei vyeodu awt e trhe es i dc ko catnodr ’ sw eonf fti ct eo , t hh ee daol rc et oa rd fyo rh aa dd i aa gpnroessi cs rai pn tdi opnr ews cariitpi nt igo nf o. rWyhoeun, ww iotrhko, uett ce. .v Ienn etfaf el kc itn, hg et do i aygonuo, sl ee td aylooun we iltoho hk ii sn ge yaets yc ol ous, eedxaanmdi na irnagn dy oo mu , pd roei ns cgr ibpltoi oo nd generator. Needless to say, you would not be going back to that doctor for treatment. Unfortunately, the prescriptions for many sick warehouses are written and ti mo opl lse, ma ennd t/eodr wl ai tchk o uo tf mt i mu ceh, emx aa nmyi nwa tai roenh oo ru st ee s tr ien- ge .n gFi onre el ar icnkg o af nkdn olwa yl eodugt e p, lraocj ek c ot sf ceox pml mo reanticoenwoift ht hoeurt eaanl youpnpdoer trustnaint ideisnfgo ro fi mt hperroovoetmceanuts.e o f t h e p r o b l e m s a nd wi t h o u t Warehouse activity profiling is the systematic analysis of item and order p o ac b r t o j i e b v c l i t e t i y m v . e s A , b c p t a i i s n v i i p s t o y f i o p n r r t s o p f m r il o i a n j j e g o c r t h - o i t g p e h a p l m o ig r h t d u t e s n c i t i t h s ie i e o s r n f o o o m r t p a c k r a o i u n c s g e e . s s o W f im m e p a w r t o e i v l r l e i a s m l t a a e r n n t d t s w , i n a it f n h o d r s m p o r a m o t v i e o i d n o e f f s l t o a h w n e mr eav jioe rw ma oftui vl la tsieotnos f aenxda mp optl ee npt iraol f irl oe sa dabnldo ctkhse itro i snut ec rc pe sr seftua lt i po nr os f. i l iTnhge. eTxhaemnp wl e es ww ii ll ll sr ee qr vuei rteod t feoarc hr et-heen gpirni ne ec ri pi nl egs yoof uprr owf ai l ri ne hg oaunsde aosr adni sot ur itbl iuntei of no rc tehnet ef ru. l l Ws eet owf i pl l r foi nf i il sehs wpriothcesthseredqautiaregdaitnheprrionfgil,indga.ta compilation, data analysis, and data presentation

ob redg ei nr st,opful or cwh aa sset ho er duenrds ,eirtsetma nadcitni vgi toyf , ui nnvdeenr tl yoirnygl eavc et ilvs i, teyt ci m. , tphreocvreesa. t i v e j u i c e s s h o u l d Done properly, profiling quickly reveals warehouse design and planning oo pp pt i oorntsu nt hi tai et srtehaal ltymai rgehnt 'nt owt onratthu rcaol lnys bi deeirni nf rgo tnot obfeygoi nu . wPirt ohf. i l iMn ga nq yu i cwkal yr eehl iomu si nea rt ee s- ec hn ag ni nceeeirni nt hg epfrior sj et cpt lsa gc eo . aPwr royf i lbi ne cgapursoevwi dee swtohrek roi gnhat bc aosneclei npet tt oh abtengei nv ej ur srtei fayl li ny gh na de wa investments. 1.2 The Participation of Profiling Pp re oo fpi ll ienfgr ogme t smkaenyy paef foepc ltee di ngvrool uv pe ds .t oDpurroi nv igd tehdeapt ar ,otfoi l ivnegr ipf yr oacneds sr ,a itti oi sn anl ai zt eu rdaal t tao, aans dk two i thhe ltph ei ndt ee sr pi grne tprreoscue lst ss .. TFoi ntahl el ye, xptreonf ti l pi nego pp leer mh ai tvse abnede nmiontvi voal vt ee sd ,otbhj ee yc t hi vaev ed ehcei lspi oe nd mI awkoi nr kg eads owpipt ho s eodn et o cbl ii ae snet d wd ehcoissei o nt es amma dl ee awdiet rh l wi t tel e aofrf encot i ao nn aa ltyesl yi s oc ra ljluesdt i fCi caaptti aoinn. Cl oaorkoeuds el il ks .e , nNoo mmaat tt et er rwwh ahta tt htehceo dmapt aa nsya ci do,unl do amf f oa rt tde, rh ewwh aa ts gt ho ei n og rtdoehr a av ne dc apr rooufsi leel ss in the new design. You can imagine how successful that project was! 1.3 The Purpose of Profiling Ywoaurewhoilulseseaectiavitloytproofficlionmg.pWlexhysgtaottiostaiclal lthdeitsrtoriubbultei?ons in our journey through Imagine we are trying to determine the average number of items on an order. S2u. 1p, p5o0s eo rwd eerds iadr tehfeo ra noanley sl iins eb, a0s eadr eofno ra 2r ai tnedmosm, asnadm5p0l i na rgeoffo1r 030i toermd es .r sW. hI na tFiisg ut hr ee n o av p e e v p r e o a r s g e h e d a n p t u o p m e a n v b s e e ! r r a I o g f f e w s it , e e t m h a e r s e e p n n e t o r i t r o e c r a p d r l e e a r f n ? u n l - i - t n o I g t ' p a s l n a t d n w d o a e . n s d H i g d o n w e p s i r o g o f n t c e e b n s a s s d e w o d e i l s l o n t b h e a d f t i l s a h t w r a i p e b p d u e . t i n T o ? h n a - s - t a I i s s t why it is so important to go to the extra step to derive these profile distributions.

Figure 1. Items per order distribution completed for a small mail order company. The average is two lines per order but the average never happens. As is often the case, designing for “average conditions” is often a recipe for design failure.

1.4 The Principle and Principal of Profiling Vl ainl fdr ewd aosPoawr enteodwbays 2a0n%e coofnt ohme pi sot paunlda tgi oa nr daennde rt h. Ha te 2o0b%s e or vf et hd ei np eI taapl yo dt hs ai nt 8h0i s%g aorf dt ehne yf ei ewl d ferdo m8 0 %t h eo f ttrhi vei aple ams .aJnoys.e pAhc tJiuv ri tayn pc ar ol lfei ldi nt gh eapnrdi ndc iapt lae mt hien si ne gp airsa tei sosne onft itahl ley v itthael r8e0p/e2a0t epdr i na nc idp l set, roart eAgBi cC aa pn pa ll iycsai ts i. oEnx aomf pPlaerse ot of ’Ps a lraewt o, ’ ss oLma we t iwmoersk i rnegf ei nr r we da rteoh oa us s ti nh eg include… • The majority of the picking activity stems from a minority of the items. • The majority of the inventory cube stems from a minority of the items. • The majority of the shipping volume stems from a minority of the customers.

Figure 2. Pareto 360

1.5 The Pictures of Profiling Wt h oh ue ns aynodus si me eu lat apni ce touurse tohfo au gmh ot st.h eWr ec oadr de l ai ni mg ihnegr fnoer wt hbeo rsna mb aebeyf,f eycotu i ne xwp ea rr ieehnocues ae w p ac r e t e i s v a e i r t n e y t t p r t r y h o i e n f i g l i i n n t f g o o a r c m s a w p a t t e u i o p r n e a i t n t h o t e a d a r e e c c t p i i r v s e i i o t s y n e n o m t f a a t t h i k v e e e r w p s i a c a r t n e u d h re o d u w e s a v e r e i e l n o h p o p u i q c s t e u o i a r c c i k a t , i l v w f i o t i y r s m . e I n c s o o p n r w o se f e i n l i c s n a u g n s , decisions as a team. If a picture is worth a thousand words, then a video is worth a thousand pmi ec tdui ar ews . e I hn at evrea tcot i vv ied edoa tfao rvai scut iavl ii tzya tpi or onf, i lsi ni mg .uIlna tt ei or na c, tai vned daant ai mvai st ui oanl i zaartei otnh ei s cs li omspe sl yt dr eypnraems ei cn tcaht ai or nt i nogf c tohnet r op lhl eyds i cbayl t wh ea ruesheor .u Ss ei mwu liat ht i otnh ei s at yc tpi ivciat yl l y par ot wf i l oe -sd igme ne ne rs ai ot innagl msimovuilnagtiwona.rehouse objects according to the profiles. Animation is three-dimensional

Figure 3. Animation model of an Avon distribution center.

1.6 The Pitfalls of Profiling On enred wI af ar lnl iinngt obtehf iosr ter awpe ableogt i)n- tyoo pu rcoaf ni l ed rt ho we wn ai nr eyhoouur soew( an sparno fei lnegs i. nSeoe mr ae npde ol opgl ies tciaclsl tt hh iast py aorua fl oy sr gi se ot ft oa ns ao ll yv sei st h! eI fpyr ooub laerme . nYo ot uc ahraevf ue l t, oy obue cc aa nr egf ue tl tsoo dc raauwg htth ue pl i inne parnodf i sl ianyg, that is enough. One means of moderation is to ask the following four questions to help identify the sufficient profiling requirements. 1. What decision am I trying to make? 2. What questions need to be answered to make that decision? 3. What information is required to answer those questions? 4. What data is required to product that information?

2. Sales Order Profiling Mc uastteormi ael r asnedr viinc feo. r Wm ahtai ot nd os hc uo sutl odmf leorws rteharl ol yu gwha nat wf r ao rme ht oh ue swe atroe hf aocui sl iet a? t eT heexyc ewl l ae nn tt twhaeri rehoorudseer sopf ielrl eadt i.onTshiesnt,het hper of ifrilset otfhcinugs t wome emr oursdt eurnsd. er s ta nd t o p la n a nd d es i gn 2.1 Customer Profiling Spoomr t ei o cnu sotfo mt heer sa cptliavci tey s ui nc ht hhei g hw adreemh oa nu ds es, oann da wh aavr ee hsouucshe , hr iegphr ecsuesntto ms uecrh sae rl avri gc ee rweaqrueihr eomu seen tfso r t haa tp airt t imc ual ay r mc ua skteo ms eenr s oe r t bo ues isnt aebs sl i suhn iat -s eap awr aa rt ee haoruesae wwiitthhiinn tt hh ee wJ CaPr ee hn onue syet. hFaot rt he xe ay mh pa vl ee, aa mJ Ca Pj oernanpepyawr eal rme haonuusf ae cwt ui rt ehri nd toheesi rs owma ruechhobuus sei. n eAs smwa ji ot hr dMi as tyr iCboumt opraonfypwa cakraeghionugs de owe si t hs oi nmt hu ec hi r bwuas ri ne he sosu ws ei tfho rMMa ya yC oCmo mp apna yn,yt hs ha to pt hpei yn gh abvaeg sa. Te xhtirredm- pea. rItny pwuabrl iechwo uasrienhgo ut as ek se,sa itshl ee s wwai rt he hi no ut hsee wwairt ehhi no uas ewaarreedheoduiscea tne od tti oo ns pteoc i af inc csuinsgtolemceursst.omInecro. ntract warehouses the entire warehouse is devoted to the needs of a The example below is from one of our food and beverage clients. Note that jouf stthfoosuer ccuussttoommeerrss t mh eai kr eo wu pn 5z 0o %n e so fi nt ht ehier wD aCr ae chtoi vuistey .t Ionhtehl ips rceads ue cwe et raasvsei gl nt iemdeeaanc dh improve customer service to those four.

Warehouse Activity Profiling and Data Mining | Copyright: RightChain™ Incorporated | All Rights Reserved As another example, many warehouses serve multiple business units under tr he es osuarmc eesr ov oe rf .s uTsh ibsuosfitneenstsi mu ensi ti s“ caomn tarjoolr. ”p oFi no tr oefxcaomn pt el en,t ai o lna r- gt he et eelfef iccoi emnmc yuonfi csahtai or ends cdoi fmf epr ea nn ty sbt ur us igngel es sd tuhnr oi tus g hf rtohmi s ttrhaed esoaf fmree c ewnat rl ye.h oI tuhs ei s. t o rRi ceasl el yr vsee r vi ne vdefnotuorr yo r wf i vaes cI no mt hmi s i nc agsl ee d, t. hUe na irqguuemf eo nr wt faorrd“ pb iucski innegs sa ruenaist ”wc eo rnet reos lt awbol ins hoeudt . f oTrh ee amc ha j obru rs ei na es sosn uwnai ts. tahl leo wl a c kt hoef amd ea qn ua ga teer sw aorf e ht hoeu s ewma raenhaoguesme etnot soyfsf et er mesaacnhd bour sgianneiszsa t iuonniat l sau pt apiol or tr et do warehousing program. At the opposite end of the spectrum is another tdei lveecrosme mb uusni ni ceastsi ounns i tcsl ihe on ut stehda ti nh tahs e psearmf eec tde ids t trai bi luotrieodn wc eanrteehr o. uTs ihneg c po rmopg ar anmy si s f soor pwraorfeichioeunstinagt bwusairneehsosufosirntghetihr aint dtuhsetryy.are considering entering the third-party In another example, in a large publisher’s distribution center, a central pool op fe rrieosdeircvael ss. t o Ec ka ci hs ubsuesdi nteos ss u up np iot r th tahs r eael l odci sattiendc/t rbeus es ri nv ee s si nuvneint st o-r yr e ti na i l t, ht rea dc ee ,nat rnadl warehouse and distinct forward picking zones to facilitate excellent customer service. Figure .4. A Small Minority of Customers Yield a Large Majority of Sales (RightChain™ Insights)

Tt hhee bmuasni naegsesr ou fn ei ta ct ho fwo rhwi cahr dh ips i cpki ci nkgi nzgo nz eo nhea si sa rdeopt ot erdt i nl i gn .e rTehpeo rstoi nl i gd rlei nl aet iroenpsohri tpi nt og r p r e e ic c la k e t i i i v n o i g n n l s g i h n i e r p e s i , s s a o t n u o d r c t s h e h s e a , D r e e ir f d f e i s c c t h i o e i n p r p t o i f n h D g a i n s r d e tr s l i i o b n u u g r t c i o o e f n s . . cTe hn itsr ai sl t ihnev be ne st ot royf , b od tehd wi c ao trel dd s -f osrhwa ra er dd

Random Backstock

Retail

Periodicals

Trade

Bookclub

Order Fulfillment

Putaway

P acking

Packing Packing

Shipping and Receiving

Warehouse Activity Profiling and Data Mining | Copyright: RightChain™ Incorporated | All Rights Reserved The warehouse within a warehouse design philosophy works because small wt haarne lhaor ug se ews ,airne hg oe nu es er as .l , Thhaev ewha irgehheorups er owd iut chti inv iat ywaanr edhcouusst eo md eesri gs ne rpvhi ci leo ps oe pr fhoyr ma l laonwc es ui t se mt o adci tvi ivdi tey apnrdo fci ol ensq aureer dt he es i wg naerde ht oo ui sdee nmt ii fsys ioopnp. oMr taunnyi t ioef s t ht oe scuubs dt oi vmi de er ot hr de eernat inr de ww aa rr ee hh oo uu ss ee s , oopr ewr aa trieohno uisnetso w si tehl fi n- c tohnet awi na er de h owu as er e. Th ohui ss ed e spi rgoncaeps spirnoga c hc ei lsl ss, i mviil ratrutaol tshpaetciufysefldexiniblme amnaunfaucfatucrtuinrginwg hceelrles imnsainduefaaclaturgrienfgacatcotrivyi. ty profiles are designed to Figure.5. Warehouse within a Warehouse concept developed for Lifeway’s national logistics center in Nashville, Tennessee.

2.2. Sales Order Profile The sales order profile includes the following: • order mix distributions, • lines per order distributions, • cube per order distributions, and • lines and cube per order distributions.

The best way to explain each of these distributions and their interpretations is to review a series of examples. 2.2.1 Order Mix Profiles To ph ee rr ae tai nr eg as tvraart ieegtyy. o fTohrrdeeer omf itxhde i smt roi sbtu thi eolnpsf ut hl aatr ae rtehheefl apmf uill yf o mr pi xl odt tiisnt gr i bwuatri eohno, ut hs ee handling unit distribution, and the order increment distribution. Family Mix Profile Ibny mt haen cyussittouma tei or nosr dt he er mo vi ex r -a lt lhoe peexrtaetni nt gt os twr ahtiecghyoor fd ae rws ar reeqhuoi rues ei t es hmosu fl rdobme md iuc tl tai tpel de fi taemmi lsi eosn otfhiet me m, tsh. eInf ti th ei s oarnd eerasr layr ei npdui craet, oi r. e t. ht ea nt dz otnoi nhga vt he ej uws ta roenheo ou fs et hoen f at hma itl ibeas soi sf cporuoldducctrievaittye aanvdirctuusatlowmaerresheoruvsicees. within the warehouses; typically leading to high For example, the family mix distribution in Figure 6 is derived from a wa fhaoml ei lsya loef dmi setrrci hb au nt odri soef cf ianl lee dp af pl aet r ss t, oc cokp. y P/ rl ai ns et er rps amp ea rk, ea hn idg he nqvueal loi pt ye sb. r oCcaht eu gr eo sr yf rAo mi s twhiedsee, fal an tds9t oicnkcshoe fs f di neee pp .a pAe cr as .r tAo nc awr teoing hosf fal abtosutto 8c k0 ips oaubnodu st .3 C0 ai nt ecghoersyl oBnigs, c2u4t i sntcohceks, bi sa as bi coeui tg h2 t4- ai nn cdh- ae-shlaolnf -gb, y1- 0e l ei nvcehnecsowp ii edre a, annddl a1s0e ri npcrhi ne tsedr epeapp. e rA. cAa cr at or nt own eoifgchust asbt oocukt

2m0erpcohuannddsis. e.Category C is envelopes and labels - extremely small and lightweight

Figure 6. Family mix profile for a large paper distributor.

In this example we are trying to figure out if it makes sense to zone the wo radreerhsoaursee mb yi x tehdo, si .ee .t fhl raet es ti ot ecmk , cf aumt si ltioecsk-, falna td setno vc ke l, ocpuet ss tt eonc kd, taon adpepnevaerl ot opgeest. h Iefr t ohne couns tt oo mp eorf ot rhdaet ,r sa, nt dh epnuitn epnavl leel to pb eusi l doinn gt owpe owf ot uh ladt . s t aI fr tt hwaitt hi sf l taht es two caky, pwuet cz uo tn set ot hc ke wt haorseehzoounsees, wo re pmaasys ap apyaal l ebti gf rtorma v eolntei mz oenpee tnoa ltthyebneecxatu. s e w e w i l l h a v e t o t r a v el a c r o s s If the orders are pure, i.e. they tend to be complete-able out of just one item fparmo ci leys, stihnegnczeol lns i, negs pt he ec iwa lal yr eshi nocues ep ar ol odnugc tt sh et es ne dl i nt oe sbwe irlel ceesitvaebdl i si nhteof ftihc ei ewn tawr eahroeuhsoeu as es flat stock, cut stock, and envelope shipments. In Figure 6, 35% of the orders can be completed out of flat stock alone, 25% oT fhteh ge ooor dd enreswc as ni sbtehcaot m( 3p5l e%t e+d 2o 5u t%o f+c 1u 5t s%t o) c7k5a%l o no fe ,t ahne do 1r d5 e%r socuat no fbeencvoeml oppl eest eadl oonuet.

og of oa ds ipnrgol ed ui tcetmi v iftaym, ci ul ys tsoumg geersst ei nr gv itchea, ta znodnsi tnogr tahgee wd eanr es hi toyupseer bf oyr imt eamn cfea.m i l y w i l l y i e l d Handling Unit Mix Distributions Ttwheo fruelvl/eaplainrtgiahlapnadllleintgmuinxidt imstirxibduisttiorinbuantidonths.e full/broken case mix distribution are Full/Partial Pallet Mix Distribution Wa r ietahs tfhoer fpual ll l/ept apri tci ka il npgaal lnedt mc ai sxedpi si ct kr ii bnug t. i oI nn swo me ter wy taor edheot ue rs me si np ea lilfe wt aenndeceads es eppi ca kr ai nt eg ag reen eprearlf, oirt mi se da oguotoodf ti dh ee as at mo ee si tt ea mb l ilsohc astei po na r, aati es l ea,raena ds /foorr apraelal eot f at nh de wc aasree hpoi uc ks ei n. gI n- rdei sptlreinb iusthi oi nng hae l pc as s er e ipni fcokri cneg tlhi ne ep/ ao ri ne at af nr odmh eal pps atl ol e ti dreenstei rf yv e“/wpai crkeihnogu saer ewa .i t hTi nh i as warehouse” opportunities. In Figure 7, 50% of the orders are complete-able out of partial pallet qa nu da nt thi et i er es m, i . ae i. nj ui ns tg c2a0s %e poi cf kt hs ;e3o0r%d eor fs trheeq ou ri rdee br so at hr epfai lrltai ba ll ea fnrdo mf u lflupl lapl laeltl eqtuqaunat ni tti ietsi e. s ,

Figure 7. Full/partial pallet mix distribution.

Should we have a separate case picking and pallet picking area? If we did, wf uol lupl da l wl e et pp oa ryt ai obnisg opfetnhael toyr fdoerrm? -i -x Ne do ,owr deerres awl l hy i wc horne’ tq. uTihr ea tmoenrl gy i hn ag popf et hn es p2 0a r%t i aolf at nhde twi mi t he .i nF ot hr e8 0w%a roef ht oh ue soer. d eWr sh, zeonn ti nh ge boarsdeedr so nc op ma l lee ti/nctaos et hpei c wk i anrgechroeua st ee s ma awnaargeehmo ue snet so yr sat emmi x, ei tds ohroduel rd. cFl ao sr smi f yi xtehde omr di me rms et hdei awt eal ryeahsoau spea lml eat npai cg ke mo rednet rs, yas tceamr t os hn opui cl dk corredaet re, ap ipc ak l laerte pa o, or tr i mo ne, rag ce atshee pciacske ppoi rc tki oann, da pn da l leei tt hpeorr pt iaosnss tdhoe wf unl sl tpr ae lal emt pf roormt i opni ctkoi nt hg e. c a s e Full/Broken Case Mix Distribution Wa nidt hbtrhoi ks edni s tcrai sbeu tpi oi cnk iwn eg . t r yI nt os od me t ee rwmai nr eehi of uwsee ss,h fouul ll dacnrde abt reoskeepna rcaatseea pr ei caks i fnogr af ur lel pg eenr feorraml , ei tdi soau tg ooof dt hi de esaatmo ee sittaebml i sl ho csaetpi oa nr a, taei salree, aasnfdo/r of ur l al raenad obfr ot hk ee nwcaarseehpoiuc ks ei n. gI n- rdei sptlrei nb iusthi oi nng ha ebl pr os k reeni nc faosrec pe i ct kh ien gp ol i innet / aa nr eda hf reol pms at oc a si de ernetsi ef yr vwe /apr iechkoi nu gs ea rwe ai t. h Ti nh i as wd i as trreihbouut isoenoipnpFoirgtuurnei t8i ei sn. d Ai csa twe as st ht ha te ocna lsye awsimt ha tl lh pe oprat li loent /ocfatsheemo ri xd edri ss trrei bquuti iroenb, ot ht he ac afsuel l pai nc kdi nbgr owk ei lnl yciaesl de qt wu aonot irtdye. rHceonmc ep,l et ot i oc nr ezaot en esse pwairtaht ev earrye al ist tfloer mf ui lxl i na gn db ebtrwo ke ee nn them.

Figure.8. Full/broken case mix distribution.

Order Increment Distributions Ws i tiut ha ttihoen oar dpearl l ient c) rreemq eune st tdeids t or inb ua t icouns two me de er t oe rr dmeirn. e Ft hoer pe ox ar tmi opnl eo, f sauupnpiot sl oe atdh e( irne tahri es 15 00 0%coa fr ttohne spoanl l eat .p aI fl lteht earneda raec 8u 0s t coamr et or nosr do enras p5 a0l lceatr at onndsa. cI nu stthoamt ce ar soer dt heer ys o2 r0d, et hr eedy ordered 20% of the pallet.

Warehouse Activity Profiling and Data Mining | Copyright: RightChain™ Incorporated | All Rights Reserved load preconfigured. How can we build half and quarter pallet unit loads? In this particular case tohne tmh ea nb uo fradcetru ar inndg af al lcwi l iet yhiasvaet ttaocdh oe di st os etth tehwe aprael hl eot ui zseer. tTohpeur et ai spaapl l ae ltl ei nt i zpel ar cteh aa tb soiut st ft oh ue rwt ai mr eehs oauss eo fitse nn ot to abt tuai lcdh eqdu at or t me rapnaul fl ae ct st uarni nd gt, wt hi ceenaesx to fbt ee snt tsoc ebnuai lrdi ohiasl ft op ahlal ev tes t! hIef Figure 9. Pallet order increment distribution for a large office supplies distributor. What do you notice that is unusual about this distribution? (In almost all of tpheeaskes ?d iTs threi bpuetai okns sa rtehea rkoeuyndi nas i2g5h%t s aanr ed 5i n0%t heofpaepa ka lsl eatn. d v al l ey s .) W her e a r e t he Suppose there are 100 cartons on a pallet and a customer places an order for 1d 0i d0n ’ct ahrat ov ne st .o Wr eoaudl dt hyi sobuoroakt ht oe rf i gpui crke oa uftutl hl aptayl loeut wo ro u1l0d0p irnedf ei vr itdoupailc kc aar two hn os ?l e-p- aYl loeut ac ut sat otmi meer . aTs hwa te l il s. Tn ho et counsl tyo mg oeordwporualcdt ircaet hf oerr ryeocue, i vbeu at if tu lils pgaol loedt qpuraanc tt ii tcye tfhoart yt ho euyr can handle in one unit load as opposed to having to handle 100 loose cartons. Now, what operating decision should we make to take advantage of the l l d o o i a a st d d r s i p . b r u T e t h c i o o e n n n f , i i w g n u h F r e i e g n d u . a r I e c f u a 6 s ? c t o u R m st i e o g r m h t p e , l r w a c p e e l a s s c h a e o n s u o a ld r n d b o e u r r d i f l e o d r r s f a o o m q r u a e a h r q t a u e l a f r - r p p te a a r l l l l - e e t a t , , n w w d e e h h h a a a lf v v - e p e a t t l h h le a a t t t u u u n n n i i i t t t

spurpecpolinefrigbuurieldththeeuqnuitalrotaedrsaantdrhecaelfivpianlgle. t loads. And if not the supplier, then we can Can we encourage people to order in half, quarter, and/or layer quantity iancccruerma teenat sn?d Av bi ssi ob ll ue tteol yt!h eI nc mu sat no ymcears ae ns db yt hs ei mopr dl ye mr eank ti rnyg pt he er spoanl nl eet l/ lvai ya etrh qe ul oa ng itsi tt ii ec ss il no af odrsm. aWt i eo nc as ny sftue rmt h, ewre ecnacno eunr ac go eu rtahgee pt rhaecpt ircaec tbi yc eoof ff eorri nd ge rpi nr gi c ien dpi rs ec coounnftisg udreesdi gunne idt at hr eo usnadl ees f foi crigeannt i zhaatni odnl i nogni nt chree mc reonstss-.f u nI nc ttihoinsacl atseea tmh ewr ehwo al si t ea rrael lpyr erseesne tt a tt hi vee pf rroi cme breaks on the quarter and half-pallet quantities the next day. The two potential downsides of preconfiguring sub-pallet unit loads are (1) tl ohses coofms tpolreaxgi tei edse no fs imt yi.xFi nogr tFhI eF Op rraoct taitcieo nw, itthhe FwI Fa Or e rhootuast ei omn arne qa guei rme me netnstyssat enmd (s2h)o ut hl de bi ned ua sbtl rei etso FtIrFaOc kr edqaut ei r eamn de nltost arroet ant ai omn e dw iftahl isne l yF I Fa sO awn i ni md opwe ds i. m eI nbt etl oi e vweo irnl d m- c laansys wc oanrtei nh ou ue ds i ntgo phr ao cl dt i coeus .t AFsI FaOn eaxsa ma pbl ea,rIr ri eerc etnot l yp rwoodrukcet di vwi t yi t hi ma pc ar onvdeymc oe nmtps .a n y t h a t I can remember a design meeting on Valentine’s Day when the company was receiving product for the Halloween season. Indeed, there can be some large time windows within the FIFO requirements. There will be some loss in storage density since a pallet worth of cartons may nwoews hhoauvl ed 2b eo ar b4l eptaol lsettas csku2p ph ao lrvt ei ns gi ni t af of ur l ls oo pmeen ui nngi .t Fl ooardqsu. aFr ot err hpaal lf l ep tasl ,l we t eqmu aany tni tei ee sd, al o rs os wi n osf t oo pr aegnei ndgesntshi tayt as hr eo u1l5d%b et allel se sr tt hh aa nn t5h%e of opre nt hi negefnotri rsei nwg laerse. h oAus sae . r eAs nu ldt ,, tt hh ee pI f r tohf iel eysi ehlodu li ds tseul cl hu st ht haet pt ho et e lnotsi sa l ipnr os tdour catgi ve i tdye ynisei ltdy ai ss s oofcf isaette, dt hwei tphrtahcet inc ee ws hpor ua lcdt i cbee. itmh aptl de me cei snitoend .o bI jfenc toitv, etlhyei sptrhaec triec ae ssohnotuol dh anvoet tbhee ipmr opfl iel me ! e n t e d . T h e a b i l i t y t o m a k e

Figure 10. Layer picking profile for a large food and beverage company. Note that the large majority of all order line requests are for layer increments; strongly suggesting the evaluation and implementation of case layer picking technology; in this case leading to a 68% productivity improvement. (RightChain™ Insights) A common pallet order increment we encounter features customers ordering ionr df ue lrl llianyeesr aqr ue af no rt i tf ui el sl .l aAyne re xi na cmr pe ml e eins tisl l; ul es tardaitnegd t ion oFui rg us rt reo1n0g . rNe oc ot emtmh aetn 8d 0a t%i o no f t thhaet lt ahyr eoru g ph ipcukti ncga p ba ec i t yi m. Ap l we mi deen tveadr i ettoy odfr laamy eart i cpai cl lkyi n ignsc rs eyas tseemtshaer ep dr oe ds curci tbi ve di t yi n atnhde case picking chapter. With the case order increment distribution (Figure 11) we determine the pa roer t1i o0n0 opfi eac ef us l li nc aar ct oa nr t ot hna at ni sd rae qc uu setsot emde ro no rcduesrtso m5 0e,r t ohredceursst. o Fmoerr eoxradme rpel ed, hi fa tl fh et hr ee f ( c u F a l i r l g t c u o a n r r e . t o 2 W n .x . h ) , A a c t s u a d s t o r o e m y s o u e u l r t s , n w t o e e t n i c w d e o t t o u h l o d a r t l d i i k e s e r u t a n o r u o s s u e u n t a d p l r h a ic a b e l o f b u a r t e c t a a h k r i t s s o a n d t i a s a n t r h d i a b a l u f t q c i a u o r a n t n ? o t n i t I ( y n a c n t l h d o i s c s e r c e t a a o s t a e e aa lnmi no ns teor rpdaecrki nfgo rt haaht aqluf -acnatrittoynn)oawn dt oa ot radfeurl li nc afrutlol nc atrot oenn ci no uc rreamg ee nc ut ss.t o m e r s w h o a r e

Figure 11. Case order increment distribution for a large bio-tech.

The general principle is to prepackage in increments that people are likely to order in and to encourage customers to order in intelligent handling increments . A higher-level principle is that the supplier should do as much as ph ao vs se i tbhl ee st ou phpe ll ipe rpdr eopaasrme ut hceh pf or ro ud su cats fpoor sps ii cbkl ei n, tgh ae nn dwsehsi ph po ui nl dg . d Ao fat se rmwu ec hn ae sg op toi as st ei btl oe at ht at th emroemc eeinv ti nt gh adtowc ke thoagveet tphreo dl ouncgt erseta tdi my feo wr si nh di popwi nagv aani l da bpl ae cfkoirn gp, i cbkeicnagu/sseh ii tp pi si nagt pp rr ee pp aa rr aa tt ii oo nn . o f At hse sporoond uac st sthhoeu l odr bd ee ra t dar ompisn i mf o ur mt htaot mpereotdtuhcet , e vt he re s hhar inndkl ii nn gg t ai mn de window for product delivery. 2.2.2 Sales Order Volumetrics Oo rudre rR i gdhi st tCrhi ba ui nt i™o nWs ; a ruenhi ot su, s icnugb eI n, sai gnhdt s wseailgehs opredre ro rvdoel ur ;m ae nt rdi c sj oiinnct l uddi es t rl iinbeust i opne sr comprised of the above.

Lines per Order Distribution Ts uhcehl ii nt ei ss opneer oofr tdheer md ios st tr ii bmupt ioornt apnot rdt ri satyr si bt uh tei onnusms bi necreoef aucnhi qu un ei qSuKe USsK oUnoar nl i on redoenr . t Ah es ol arrdgeer mr eapj orreist ey notfs t ha ev wi s iotr tkol otahde i un nainqyu ew laorceahtoi ouns ef.oMr at hn aa tg iSnKgUl o. cLaotci ao tni ovni s vi tiss ii tsspaerreh at hp es tkheey md eaci ins imo nass tienrcyl urdeiqnugi roerdd ef or rp mi c ka ixni mg si zt ri na tgewg ya raenhdowu shee tphreordourc tni ov itt tyo. Hb ae tl pc hs pt oi cikn. f o r m The lines per order distribution in Figure 12 indicates that 50% of the orders isni xt ht oe nwi anree, haonuds e1 0a %r e ff oo rr ot enne ol irn emiot er em. , 1W5h%e rfeo ri st wt hoe, 1p 5e a%k ?f o rI t t hi sr eaer ot uo nf di v es ,i n1g0l %e l ifno er oi nr ddievri sd. uTa hl icsoins snuomt eurnsc oo mr tme cohnn, iecsi ap ne cs i aa rl ley pi nl atchi en gmoari ldoerr ds eorni nt dh eu swt rayr eo hr oi nu scea .s eWs we hn eorwe need to consider the operating strategies which take advantage of this order profile.

Warehouse Activity Profiling and Data Mining | Copyright: RightChain™ Incorporated | All Rights Reserved Figure 12. Lines per order distribution for a large mail order company. First, “singles” may be backorders. Backorders are an excellent opportunity fcoa rn c br oe s bs -adt oc hc ke idn gt o. gSeetchoenr df, o“rs i pn igcl ke isn” gmoany bs ien sg ml e a- llil n, ee mp ei cr kg ienngc yt oour rdse, ras n. dT hboys ep roi rndt ei nr gs sa idndgiltei o- lni n, et hoer od redr se ri nb al tocchaet si onna tsuerqaul leyn zc oe nwe et hcer ewa at er evheoruys ee fifni ct ioe nz ot npei sc kdi enfgi nteodu rbsy! t hI ne

ol epnpgot rht uonfi t yt hteo pc ri ce ka itne ga tdoyunra. m iTc hf iorrdw, asridn gpl ei c- kl i nl ien eo. r dI ne r tsh i ms ao yp e ar al st oi n gr espc reensaerni ot , aa nn w a b u a h t t i o c c h m h p a t i h t c e e k d r e e l d o i o a s n k a - d t a l s h e e e a t a s u d t p a i n a f l t u o o l n l t - g h c e a fa r d s t a t o n p y ’ i ’ s s c k w o - r p o s a r h t c h i k f t o l ’ i s f n o s e i r s n d . g elres l mi naeyoyr ideel dr sa. nTuhmo sbee rS Ko fUSsKcUa sn fboer

Figure 13. Lines per order grid analysis. (RightChain™ Insights)

Lines and Cube per Order Distribution Ti nhf eo rl mi naetsi oann dn eceudbeedpteor do er df i enre doirsdt reirb upti icokni nbgr isnt gr as t teoggye. t hI te ri si na oj onien tp rdoi sf itlrei bt hu et i oc nr i ttihc aa tl classifies all orders into lines per and cube per families. It illustrates the typical daily ap ni cdk ti nh ga taoc ct icvui pt yy. l eI ns st ht hi sa ne xaa cmu pb li ec f(oFoi gt uorf es p1 a4 c) et. hTe rheo saer eo r1d7e6r so radr ee rpsrwo bi tahb loyncealni nd ei diat et ems fcoarr tas , st iontgelse, oorp se hr ai pt opri ntgo cboanttcahi nteorgse. t hTehre rf oe ri spoi cnkei nogr di ne rt owci tohmmp aorrtemt he na nt a 1l i 0z el di n pe iict ke imn gs tchaantdiodcactuepfioers ams ionrgel et hoapner2a0t orc ut ob ipc i cf eketto, aa bpoaullteta. t hi r d of a p al l et . T ha t ord er i s a

Figure 14. Lines and cube per order distribution or a healthcare supplies distributor. The example in Figure 15 is a lines and cube per order distribution from a mt oayj o) ,r arnedt a lial rcgl iee n( tx. Nt ooyt e) . tTh ha ti st hi se toyrpdiecrasl l tye tnhde t coafsael l wi nht eoroenoensl my aa l lf e(wx t ooryd)e, rmteydpieusma r( ex rt he ea l el yx at mh epr lee; , as nmda tl lhoerr de earrsea rr eeapl li yc koendl yv iaa fsepwe cvi iaal bpl iec kc o- pnac cekp tcsa rf ot sr ht ho ol dsien go rt de ne rt ot ytpwees n. It ny ot or df ei vres po er dr ecrasr tp. eMr ecdai rutm. Lsairzgeeoor dr de er sr sa raer ep ipcikcekde dwoi tnhemaet dai utmi msei zi en cl aa rr tgse hoorlddei nr gc at wr tos holding a single order per tour.

Figure 15. Lines and cube per order distribution for a large retailer.

Figure 16. Lines and cube per order distribution for a lighting manufacturer and distributor.

Figure 17. Summary order volumetrics for a hardware supplier.

Figure 18. Cube and weight per order joint distribution. (RightChain™ Insights)

3. Purchase Order Profiling Tl i nh ee sppuer rc hoar sdee ro rddi se trr pi bruotf ii ol en i, nacnl du dl iensetsh pe esra mo red edri sdt ri si bt rui tbi uo tni so n( o) radsetrhme icxu ds ti os tmr iebru ot irodnesr, pp ruor fcihl ea. s eT hoer doenrl yp rdoi ff fi leer ei sn cues ei sd t ht oa tmt ha ek ea ct thi ev i st ya mi sei nbbaot cuhnidn gi nasnt eda dp roofc oe sust ibnogu sntdr .a tTehg ey ds tercai tsei og iness aa rs e wf oars r et hc ee i pc tuss at on md ep ru toa rwd ae yr s pa rsoof ipl ep oesxecde tpot otrhdee rbpa it cckhsi n. gK eaenpdi np mr oi cneds st hi nagt yl i os tu or fpl ui nr ec hi taesme onrudme rb ei sr sy, oduers sc ur ipppt il oi enrs’ s, acnuds tqoumaenrt iot iredse. rT. hTehoensl tyr ud ci ftfuerr ee ni sc et hi se tshaamt et h, ae pfruormchyaosuerowrdaererhisouinseb.ound to your warehouse and the customer order is outbound 4. Item Activity Profiling Ti t he me i t( e1m) wahc taitv si ttyo rpargoef iml eoids eutsheed ipt ermi msahr oi luy l dt obsel oats st hi genwe da rt eoh, (o2u)s eh,otwo md eucci dh es pf oarc ee at hc he ii tt ee mm sshhoouul ldd bbeea ll looccaatteedd. i nTthhee si tteomr a gaec tmi voi tdye , parnodf i (l e3 )i nwchl uedr ee si nt ht hee fsotlol or awgien gm oa dc tei vt iht ey profiles: • popularity profile, • cube-movement/volume profile,

• popularity-volume profile, • order completion profile, • demand correlation profile, and • demand variability profile.

Warehouse Activity Profiling and Data Mining | Copyright: RightChain™ Incorporated | All Rights Reserved 4.1 Item Popularity Profile At hse mp iecnktiinogn ae dc t ievai rt yl i .e rT, hae mp oi npourliat ryi toyf dt hi set ri itbe umt si oinn (as owma er teihmoeuss ec agl el ende ar ant Ae Ba Cmcaujrovrei t oy r oaf Pareto Distribution) indicates the x% of picks associated with y% of the SKUs (ranked

t p b h y g t ti theemdsi s(tFr ai bmuitl iyo An )mma ay ys umgagkees tuipt e5m0 %p o op futl ha rei tpyi cf ka mi n igl i aecs t. i vF iot ry ,etxhaemnpelxet, 1t h5e%t oopf 5t h%e oi tfetmh es (i tFeammsi l(yFBa m) mi l ya yC t)a ckoevuesr ttoh e8 0r %e moafi nt hi ne gp pi cikc ikni ng ga catci vt iivt iyt,ya. n Td ht ehseer feammaiilni ei ns gm8a0y%i no ft ut hr ne ss ut ogrgaegset tmh roedee a, lFt ae mr nial yt i vBe isnt oar as ge me mi - ao ud teosm- Fa taemd i lmy oAdi enr aa nt eal yu tpormo ad tuecdt ihvieg hpliyc kpirnogd mu cot di vee, ab nr eda Fk apmo iinl yt sCmi na yaa ml s oa ns uu ag lg pe si ct kt ihnegl omc ao tdi eo nt hoaf tt ho ef f ietresmhsi gwhi tsht ionr aa gset odreang sei tmy .o dTeh. eAf ai tme mi l ys liotecmatsedinitnhethe golden zone (close to a travel aisle and/or at or near waist level), B silver zone , and C items in the remaining spaces. o a p d t u t e l h s a e c r e 1 i n t 0 e d % m in s m g r o p e s o p t p r p u e o s la p e r n u i t t la y 9 r ) 0 . i t % e F m i g o s u f r r t e h e p e 1 r p 9 e i s c i e s k n i a n t g 7 cl 0 a a % s c s t i i o v c f i t p t y h o , e p a u p n l i d a c k r s i i o t n y g o d n a i c . s t t i K r v i e i b t y u y , b t i t r o h e n e a 5 k in p 0 d o % i i c n a m t t s i o n i s n

Figure 19. RightChain™ SKUs popularity profile for a large service parts company.

Figure 20. RightChain™ SKUs popularity Pareto for a large retailer. (RightChain™ Insights)

Warehouse Activity Profiling and Data Mining | Copyright: RightChain™ Incorporated | All Rights Reserved The overriding principle is to assign the most popular items to the most accessible warehouse locations. Unfortunately, many warehouse operators use the wn urmo nbge rmoef arseuqrueeos tf spfoopr ut lhaer iittye.mS. oImn et hues ee nsda, l ea sl l, os of mt hee sues ae r ue swa gr oe ,n agn! dT shoemn eu mu sbee rt hoef ri ne fqourems tast i of onr t aona si tsei gmn ii st etmh es ttor uset omr aegaes umr oe doefs po or peuvleanr i tt oy , l oh coawt ee vi teerm, ist wi si tnhoi nt es tnoorua gg he mw iot dh ei ns . t hTeh ea sps ri gonpeedr as tsos ri ga ng em me not doef iist ebma ss etdo os tno rt ah ge epmo poudleasr i at yn dd ias ltlroi cbauttiioonn oafnsdp at hc ee cbue bl oew- m. o vAe mj oei nntt dpi os tprui bl aurtiitoyn- c. uAbceu- mb eo-vme omveenmt e dn itsdt ri si bt rui tbi uo tni ofno lel xoawms .p l eFirsopmr e st he ne t ej odi ni nt pasospigunlamrietyn-tcsu. be-movement distribution we can make appropriate slotting 4.2 Cube-Movement Distribution Td ehcei smi oonsst ri es v tehael i ncgu bdei s- mt r iobvuetmi oennfto r( odre tveor lmu mi nei n) gdsi tsot rr iabguet imo no.d e Tahned cs up ba ec e- ma ol lvoecma t ei onnt distribution indicates the portion of items that fall into pre-specified cube-movement

rc au nb ge e- ms . oIvf et hmeepnrt e -dsips terciibf iuetdi orna nwg ei lsl ceosrsreens tpi ao lnl yd tsoo sl vt oe r at hg ee ms toodr ae gael t emr no adtei vae ss s, it ghne mn tehnet pmroonbtl he .mT. hFoosreeixt ae mm sp lme , ai yn bt heegfoi goudr cea, n1 d5 i%d aot fe tshfeo ri t setmo rsasghei pd rl easws et rhsa on r0b. 1i nc us hb ei cl vf ienegt . pAe rt tchueb ioct hf eeer te(nnde oa rf ltyh e2 0d i ps tarlilbe tust)i opnewr me foi nndt h1. 2 T%h oo sf et hiet ei mt e smms tahya bt em coavnedmi doartee st hfaonr 1b,l0o0c 0k stacking, double-deep rack, push-back rack, and/or pallet flow lanes. The principle is to assign items to storage modes based on their cube-movement.

25%

20%

15%

12%

15%

10%

% Items

<0.1 0.5 2

5 10 50 100 250 1,000 1,000+

Cube (FT 3 ) Picked/Shipped per Month

Figure 21. Cube-movement distribution.

4.3 Popularity-Cube-Movement Distribution Dt hoen ceupbreo- pmeor vl ye, ms leont tt i nd gi s tt ar ikbeust ii on nt o. aTchc eosuen td ibsot rt hi b tuht ei oint es mc apno bp eu l caor mi t yb idni es tdr iibnut ot i oanj oa inndt pdiicstkriinbguitsiopnr.esAennteedxabmelpolwe .popularity-cube-movement distribution for broken case

FLOW RACK

BIN SHELVING

CUBE MOVEMENT

CAROUSELS

STORAGE DRAWERS

POPULARITY

Figure 22. Popularity-Cube-Movement Distribution for Broken Case Picking

Warehouse Activity Profiling and Data Mining | Copyright: RightChain™ Incorporated | All Rights Reserved In the example, those items exceeding a certain cube-movement threshold are an seseidg nt eodbteo rceasrt toocnk ef dl o wf r erqaucekn. t Il yt e, ma ns dwni et he dh iag hl a cr guebre -smt oorvaegme el onctattui ornn oavsecrofmr epqaureendt ltyo, istteomr asg we mi t ho dme etdhiautmf a ca inl idt altoews rceusbt oe c- mk i on vg eamn edncto. nHdeenncsee st hl ea yr gne eset do rtaog eb el oac sa st ii og nn es da ltoon ga tahr ee pg ei cnke lriantei n-gc ma rat on ny pf l iocwk s rpaec rk .u nI itteomf ss pwaict eh tl ho awt tchuebye omc couvpeymaenndt da no dn oh ti gohc cpuoppyu ml a ur ict hy st hpiascceaas leolni gghtthdei rpei cctke dl i ncea.r oTuhseeyl s naereedr et oc obme mi ne nadheidg hbleyc apur os ed ut hc tei vpei cpkiicnkgi npgr omd ou dc tei .v iItny ihso uh si gi nhg as nodn wt hee cpai cnk al if nf oer da ntdh ed oc anroot unseeel sd ftoor biet erme sst otchkaet ddfor enqoute nntel eyd. (l Ca ragr eo ussteolrsadg oe nwoi tt hl e lnodwt hpeomp suel al vr ei tsy t oa nr de s tl oo wc k i cnugbaen- md oa vr ee me xepnet n cs ai vnen po et rbceu bj ui cs tf iof ioatbol yf shpoauc se e. )d Ii tne ma ns es txopreangsei v de rsatwo reargs e. mOo nd cee. Ht heen c set, ot rhaegye a rme ocdaen dai ds saitgens mf oern tbsi n hsahveel v bi nege na n md amdoed, utlhaer

pd ri setfrei rbeunt ci oe nrse. g i oTnh so sf oe r i et eamc hs sitno rtahgee bmoot tdoemb erci og hmt e- htahnedi r ppoorpt ui ol anr iot yf - tchueb ed mi s torvi be mu t ei onnt gHeennecrea, tteh tehye smh oous ltdp bi cek iansgs iagcnt ei vdi ttyo ppeor sui tni oi tnosf isnp tahc ee tghoel dy eonc czuopnye .i nTthhoesset oi treamg es mi notdhee. uo fp sppear crei gt hh et yh oa nc cdu apny di nl ot wh ee rs tloerf at ghea nmdo gd een. eHr ea nt ec ea t mh eoyd sehr oa tuel dn bu emabsesri gonf epdi ct ok sp po es irt iuonni st idni s tt rhieb ustiilovne rg eznoenrea. t e Ft ihnea fl el yw, et shtops ei c ki tsepmesr ui nn i tt hoef suppapc ee rt hl eeyf t ohc ac unpd y qaunaddtrhaenyt sohfo ut hl de be assigned positions in the bronze (least accessible) zone. This example is not meant to make an end-all recommendation for slotting br ar toek, etnh ec acsoespt i cokf i snpgascyes, t tehme s c. oTsht aot fd ce appe int da ls, ot nh empalna yn no ti nh ge rhf oa rc itzoor ns ,i nect cl u. d Ii nn gs ttehaedw, tahgi es euxs ae dm pi nl e tihs ep sr el ostetni nt egdptrooicl el ussst. r aOt en cheo wi n t hp el apc oe ,p tuhl ae rdi tiys-tcrui bbuet-imo no vpermo ve indteds i smt roi bs tu toi fo nt hies insights required for slotting the entire warehouse.

Figure 23. RightChain™ Warehousing Insights SKU demographics screen.

4.5 Item-Order Completion Distribution t t T h o h a e s t m i c t a e a n m ll f -i ol l rlda er gr ec og rmopulpest ioofnodr di set rr si b. uTthi oons e( sFmi g aulrl eg r2o4u) pi sd eonf ti ti fei me ss scma na lol fgt reonubpes aosfs ii gt enme ds order completion zones in which the productivity, processing rate, and processing quality are 2-5 times better than that found in the general warehouse.

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

Full Case Orders Broken Case Orders Overall

% Orders Complete

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

% Items

Figure 24. Item-order completion distribution.

The item-order completion distribution is constructed by ranking items from mp oopsut l at or il teeams ts ,ptohpeunl at hr .e tBhergeien nmi no gs t wp iot ph utlha er imt eoms ts , peot cp. utlhaer iitteemms, tahreenp tuht ea gt wa i on smt tohset oc or dmepr l es teet . tIon dt heitse er mx aimn ep lwe h1a0t%p oo fr tt ihoeni toefmt hs ec aonr dc oemr spal egt ei v5e0n%s uobf st he te oo fr dt he res i. tSeumpsp coas ne Io wf tahl ke i on rt do eyrosu. r wWahraeth wo uosuel da nydoiud ednot i wf y i 1t h0 %t h oo fs teh 1e 0i t%e m? s It hhaot pc ea nycooumwp loeut ledl y cfri lel a5t 0e %a warehouse within the warehouse or order completion zone for those 10%. The design principle is similar to that used in agile manufacturing, where we laonodk tfhoor ssempaalrl tgs rmo uapkse oufpp aa rst ms tahl la gt rhoauvpe tsei mc hinl aorl omg ya cchei lnl ewrhoeurtei ni ng st .h eT hmoasne umf aacct hu irni ne gs

efaffcitcoieryncays,aquwahliotlye,.and cycle time are dramatically improved over those found in the I recently worked with a large media (compact discs, cassettes, videos, etc.) dt hi set or irbduetrosr. aWn ed ahsesl ipgende dt ot hi doesne t 5i f%y 5t %o coa fr ti tosn4f ,l0o0w0 rSaKc kU sp owdhsi c( 3h fcloouwl dr ac cokmbpal ye st ep3e 5r %p oodf, o T 1 r h o d e p e e d r r s i a s f t t r r o o i r b m p u e t t h i r o e p n f o l c d o e w ) n a t r t e a t r c h k h e a a f s r t o n w n e t o a o n rl f y i t t h 6 s e t i n d im d is u e t s r s i t t b r h u y e ’ t s i o o p v n e r c o r e a d n l u l t c r e t a r i t v . e i O t o y p f e a t r h w a e a t d o r r d is s t f c r o o i r b u u t ld h ti e o p n i l c a c k s e - t n p t t a w e c r o k . years. Very frequently there is a driving force behind order completion. Examples cpor uo lddu ci nt cgl ruoduep c, uas tsoump pelrise ro, rad es ri zi ne ,g awci ot hl oi nr , aa bk ri ta, nedt,c .o Or uars RdiegphitcCt he da i ni n WFai gr ue hr eo u2s5i n ag Iant tsei gmhpt st ™t o pi da et tnetri nf y roer cdoegr nciot imo np l eatl igoonr iztohnmi ns g soepe pk o ar tnudn ist ui ersf awc ei t htihnotshee pwaat rt ee rhnosu sien. a n

Figure 25. Item-Order Completion Profile from a large food and beverage company.

Figure 26. Demand correlation profile for a large hardware distributor.

4.6 Demand Correlation Profile Jaucst ti vliitkye, ac emr tiani on r iittye mo fs t ihne ti theemws ai nr eah owuasree ht eonuds e tmo abkee ruepq ua ems taejdo rti ot yg eotfhtehre. pIinc kti nh ge eo xr da me rps .l eW, pea ai rrse ol fo iot ke mi n sg af or er gr ae nn ke readl bpaast et edr onns . t hI ne itrhfi rsecqauseenwc ye oa fr ea pe px ae ma riinnigntgodg ae tt ah ef rr oomn ac rme wa i nl oe cr kd esrwaepapt ea rr ,e Vl -cno emc pk asnwye. aTt ehre, tf ui rrsttl e3 ndeicgki t ss hrierpt , rpe lseeant te dt hpeasnt tysl ,eeot cf . )t h, teh iet emmi d(di . lee. dt hi ge i ltarset pdriegsi te nr et sp trhe es esni zt se tohf et hceo li ot er m( 1(=1w= shmi t ea ,l l2, =2 b= lma cekd, i 3u =mr,e3d=, l4a=r bg el u, e4,=5e=x gt rrae el anr, geet )c,. )a. n d

Sty

Item Number 189-2-4 493-2-1 007-3-3 119-2-1 999-1-8 207-4-2 662-1-9 339-7-4 112-3-8

Item Number 189-2-1 493-2-8 007-3-2 119-2-7 999-1-6 207-4-4 662-1-1 879-2-8 112-3-4

Pair Frequency

58 45 36 30 22 15 12

9 6

Warehouse Activity Profiling and Data Mining | Copyright: RightChain™ Incorporated | All Rights Reserved tahbroouutgl ho gtihs tei cps rios fsiul iensgi spor fotceens so f -f - bt oa sseu. rTf ahcee mt hy er i at rduot hf S! K (UUsn, of or rdteurnpaat et tl ey r on us ,rs ui nptpuliitei or sn, alongdistiinctseordpeepraetniodnens.t) decisions make it difficult to form reliable intuition about How do we take advantage of this demand-correlation information in slotting tt hh ee wf a ac tr oe rh ot huas te ?w iWl l ec raeraet el otohkei nl agr fgoers tt hf ae ml oi lwy eosft i ct oe mm sm. oI nn dt heins ocma si ne ai tt oi rs ot hf ec os ri zr ee l oa ft i tohne, imt eemd i. u mS os, , wl aer gzeosn, ea nt dh ee xwt raar el ahroguesse obf ya lilt es tmy l es si z. eWf iirtshti, nc re ea ac thi ns gi z ea az roenae, wf oer st thoer es mi t ea ml l ss, of the same style together, mixing colors within a style. This zoning strategy allows Figure 27. Demand correlation distribution (Style-Size-Color) for a large omni-channel retailer. What do you think people tend to order together from this mail order apparel ci na t tahl oe g coapt aelroagt .o) r ? W( hI at ht oduogehst ti th ew odui sl dt r ibbeu tsihoinr t si na nFdi gpu ar en t 2s . t2h6a ts luogogkeesdt ?g oIond tthoigs e ct ha seer ct huas tt ocmu setros mt eenrds ttoe nodr dteor igt eetmcsoomf ftohret as abml e ews ti tyhl e aa ncde rstiazien t os tgyelteh earn. dT theenedx pt ol a no ar dt ieorn iins u i m t n u w l l e t a s i s p s l a t e h s c e u o y r l p o w r r i s i l s l t e r o e t a t o u d t r d h n e v o a m n ri e a e r f t o k y r e t t o f i i n t t t g h i n e p i g r e . o w T p a h le r i d . s r w T o h a b a s e t . a i O s s f u t r h c p o e u r m i r s s e o e s t , t o t h i m m ey e p . o o r M r d ta o e n r r e t t h r im e e a p s s a o o m r n t e a t n o s t i g l z y o e ,

umsa tnoy c ri teeamt es poi cnk isnhgo tr ot -udriss tbaansceed po inc ks ii nz eg at no du rsst . y l eA. t Atsh ea rseasmu let , toi mr dee, r wp iec kwe irlsl cma na npai gc ke cpoonpguelastrioconlobryfosrperaeachdisntgyleouatt othrenesaizrews.aisGtolledveenl. zoning is used to store the most

Figure 28. Optimization based upon demand correlation.

Figure 29. Commodities per order distribution. (RightChain™ Insights)

4.7 Demand Variability Profile Td ah iel yddeemmaanndd- vf aorri ae abci lhi t iyt edmi s. t rUi bn uf ot ri ot unn(aFtiegl uy r, ea n3 i0t )e mi n’ ds idc aa itleys dtehme asnt adnids anrodt dp er ve di ai tcitoanb loef. Ds uucrhi nt gh aat reeaccehn tp pi crko fj ea cc te wh eel wd ae rdeatyr’ys i wn go tr ot hs iozfes tt ho ec kp. i cTkhfea mc eos tai vl oa nt ograwcaass et op mi c ka ikneg slui nr ee tphi cakt wf aec edsi ds inz oe dt nf oe er da nt oarveesrtaogcek da al yo’csadt ieomn adnudr ,i na gn dt hteh ed ac yl i .e nTthceo cuul dr rne on tt fdi ge us irgenohuat dwt hh ye twhhe yy . hIaf dt hteo pr iecskt of ac kc es ios ms i az endy fl oo rc at thi eo nasv edruargi en gd at hy e, ucnoluerssset ho ef tshaemdeaqy u. aInht iot py ei sy po iuc ks ee de edvaeyrsywshi negnl et hdeapyi, ctkh ef arceewi si lul nbde emr sainz eydd, at hy us swrheeqnu itrhi ne gpai crke fpal ceen iisshomv ee nr sti. z e d a n d m a n y The real objective was to make sure that there was no need to restock during af apc ei cmk iunsgt sbhei fsti z( 2e dpti oc ka cs choi fmt smpoedr adtaeyt ha en da v1e rr ae sgteodc ka yi n’ sg dsehmi f ta npde rpdl uasy )e.n Ho uegnhc et ,ot choevpeirc1k sdteavni da tai rodn sdoef vdi ae tmi oann doffo rd ae m1 %a ncdh af no cr e ao f 5r %e s t oc chkainncge. Oonf creetsht eo cpki icnkgf a ac ne sd w2e r set raensdi za er dd

Page 2 Page 3 Page 4 Page 5 Page 6 Page 7 Page 8 Page 9 Page 10 Page 11 Page 12 Page 13 Page 14 Page 15 Page 16 Page 17 Page 18 Page 19 Page 20 Page 21 Page 22 Page 23 Page 24 Page 25 Page 26 Page 27 Page 28 Page 29 Page 30 Page 31 Page 32 Page 33 Page 34 Page 35 Page 36 Page 37 Page 38 Page 39 Page 40 Page 41 Page 42 Page 43 Page 44 Page 45

Made with FlippingBook - Online magazine maker