PROMOTIONAL ARTICLE

IKS - Integrated KanbanSystem

Tony Campion, MIOM, Campion Associates Ltd,

Phil Parry, MIOM, Jullip International Ltd

and Martin Ellis, MIOM, Optimised Operations Ltd

"JUST TELL ME WHAT YOUWANT AND I'LL MAKE THEM!"

Heard it before? A simple andreasonable enough statement fromany shop floor supervisor, but onethat has on a daily basis, frustratedthe hell out of every operationsmanagement professional.

The answer is similarly simple:. take the demand for a productover a period, lets say a month. calculate how many of eachcomponent part we need forthat product (manufactured andpurchased). take away the quantity of eachpart we already have in our factoryand have already ordered. work out how long it takes us tomake or buy a quantity of each part. then work out when we need to

start making those additional parts sothat demand is always satisfied.

which will give a schedule, day-by-day, of what we need to make andbuy.

However, there are thecomplications, such as:. we don't only make one productwe make lOO'Sor even 1000'S

. products may be segregated intoproduct families but each family has adifferent process route and as suchhas different lead times

. scrap and rework at the differentprocess stages can lead to late orunder-supply. suppliers and sub-contractorsdon't deliver what they are supposed,when they are supposed - or don'tmake what they are supposed to dueto machinery breakdowns, peopleshortages, etc. a change to the whole dynamics ofthe process due to new products andprocesses and Lean programmes thatlead to reduced manufacturing leadtimes, reduce inventory levels,

24

reduced scrap and rework, etc.. Plus many other factors set tomake our planning lives difficult.

Over the years there have been anumber of systems and solutionsavailable to operations managementto assist in deriving an answer to thisconundrum and all its variability andcomplications.

PUSH MECHANISMS

Typical Material RequirementsPlanning (MRP) systems werecomputer packages derived frommanual cardex systems that carriedout the scheduling calculation todetermine what was needed when.

These systems had no resourcecapacity constraints associated withthem and as such manufacturingresources became significantly over-loaded, requiring teams of progresschasers to hunt out and push ordersthrough the processes.

The natural evolution was to

create Manufacturing ResourcePlanning (MRPII) systems whichdid try to take account of capacityconstraints. However, the improvedcapability of the systems did notalways lead to better schedulinginformation being delivered to linemanagers. These shortfalls were

Figure 1

Itt'::I1'Fa......Orde..Alr.dy .ReIeeHd ~ .,

Bill 01 ... ~~p'--..Materll,"

.[}

StockInSlo...

Schedule

CONTROL Number 4 2004 iomnet.org,uk

commonly associated with datacapture and feedback to the computersystem, a time consuming and oftencomplex task. The MRPII systemsoften required complex interfaceswith other business computer systemssuch as engineering, purchasingordering, accounts, etc, which alsolead to data management problems.

The data management andresponsiveness issues associated withMRPII systems led to the develop-ment of Enterprise Resource Planning(ERP) systems. ERP attempts to dothe same job as an MRPII system buttakes advantage of developments intechnology. However, this stillPUSHES orders for component partsinto the process chains with no easyway of being able to respond to day-to-day fluctuations in processperformance - which commonly

results in unpredictable inventoryaccumulations and deterioration indelivery performance.

PULL MECHANISMS

The focus on the Japanese approachto manufacturing through Kaizen andLean programmes has led to theadoption of Kanban systems whichPULL, as an alternative to thecomputer driven PUSH systems.Orders for component parts arereleased, based on usage downstreamthrough the process chains.

Typically, a buffer (supermarket)stock level is established for a

component part between each processstage. The buffer stock is divided intosmall chunks (at least 2 halves) eachchunk of stock is assigned a ticket orKanban. When the downstream

(consuming) process takes a chunk ofstock to be processed, the Kanban isreturned to the upstream (supplying)process as a signal to indicate that thebuffer stock needs to be replenished.

1

TriggerPoint

Figure 2

INFORMATION FLOW

SUPPLYINGPROCESS

MATERIAL FLOW

When a pre-defined number of signalshave been received, manufacture istriggered - this trigger level being setso that the buffer stock level will notreach zero.

The simplicity of the underlyingconcept - ie, only replace what hasbeen used, is embraced withenthusiasm at the shop floor level andprovides flexibility to cope with day-to-day fluctuations in processperformance.

The operations management rolein a Kanban system is to use theinformation on:. demand

. cycle time

. batch sizing

. manufacturing lead timeto determine the:

. size of the Kanban system (thesize of the buffer stock). number of Kanban signalsrequired. trigger point to commencereplenishment.

Demand will be satisfied within aknown stock level without the need to

constantly update masses ofcomputer records every time theprocess parameters change.

However, some problems that arefaced when running a manual Kanban

Figure 3

system include:. Adjustingthebuffer stock and No.of Kanbans as end

product demandchanges. Difficulty ingetting advancedwarning of supplyproblems -the firstindication often

being when thesupermarket buffer

is empty. The Kanban system is establishedwith higher buffer stocks to ensure nosupply problems -but then adjustingstocks down in a controlled manner isdifficult. How to ensure all Kanbans are still

live - eg no cards missing,especially when handling severalhundred part numbers, etc.

Most companies operating aKanban system still use MRPII/ERPas the main business system forstrategic planning, forecasting capacitydemands and capturing financialperformance, etc. Kanban is then usedas the call-out signal to the internaland external suppliers. Kanban sizingcalculations are usually done byspreadsheet separate from mainsystem. Therefore to update theKanban system data needs to beextracted from the main business

system and interpreted through thespreadsheet to make necessaryadjustments. Also updating the mainbusiness system with data to satisfyfinancial performance analysis (eg.Stock/WIP on hand) has to be carriedin addition to running the Kanbansystem. These tasks can be onerousand soften the advantages of runningthe Kanban system.

Demandchangesslowly over time

HighValueLow Volume

Typically1-10

PROMOTIONAL ARTICLE

eKanban

Recently, a new software systemhas been developed that enablescompanies to reap the benefits ofrunning a Kanban system for the day-to-day operation whilst maintainingand utilising data in the mainbusiness management system(MRPII/ERP) and having morerobust Kanban management.Effectively a hybrid Push/PullMechanism bridging the gap betweenthe two operations managementapproaches and offering the benefitsof both systems.

The system is called IKS(Integrated Kanban System) andhas been developed by Kanbanpractitioners through an associationof European businesses. IKS is aWindows style PC based systemrequiring minimal installation andtraining, and dramatically enhancesthe manual Kanban system by:. Determining the number ofKanbans in a loop. Printing and managing theKanban cards for each Kanban loop. Monitoring the location andmovement of Kanban cards

Recommending adjustments toKanban loops as end product demandchanges. Providing an early warning ofpotential supply problems. Enabling diagnosis of eachKanban loop performance. Displayinga computerisedKanban board giving visualmanagement. Integrating Kanban with the mainbusiness system so that data can beeasily exchanged.

The system is divided intomodules arranged as illustrated:

Figure 4

PARTSMediumValue

MediumVolumeTypically10-100

Low ValueHighVolume

Typically100->1000

nStochasticDemand

CONTROL Number 4 2004 iomnet.org.uk 25

Manual Kanban

IKSElectronic Kanban

Electronic Kanban?This should be decided on a case by case basis

PROMOTIONAL ARTICLE

The type of business that canbenefit from IKS is illustrated in the

matrix on the previous page: IKS is amajor contributor towards reducinginventory, shortening manufacturingand purchasing lead times and gaininggreater control over purchased partssupply with almost no stock-outs andlower stock levels. It is designed as an

About the authors

Tony Campion, MIOM hasbeen involved with a number

of MRPII / ERP system

implementations along with theimplementation of Kanbanthrough the supply chain.His early career began as anOperations Managementpractitioner with APV, BAE andTriumph and prior to becoming aManagement Consultant, Tonywas involved with Lean, starting

Industry Forum the.n'~'!i;"'{i" i~";:r.;~:,.;iJ...,.M;,..,,_" ~..

electronic tool to assist the operationsmanagement activities and enhanceKanban and business performance.At the direct operations level there isno difference in the running of theeKanban system compared to runninga traditional manual Kanban system.The differences and improvementsare experienced through reduced

Phil Parry, MIOMhas gainedwide aerospace experience fromhis involvement in managingmajor operations improvementprogrammes. Much of his workinvolvespromoting a morevisionary approach to thedevelopment of operationallogistics, continuous improve-ment and the introduction oflean manufacture focusing oncost, inventory and lead timereduction. Phi! is also,a member

inventory levels (higher stock turns),reduced manufacturing and purchaselead times, and improved deliveryperformance. Coupled with thesemore tangible measures all currentuses express an enthusiasm for thesystems simplicity of operation andthe confidence of assured supply itprovides.

Martin Ellis, MIOM hasindustrial experience from15 years at Rolls Royce aeroengines and Jaguar cars. He hasbeen a Management Consultantfor 10 years, with his initialconsulting experience comingfrom working for Ernst & Young,and Arthur D Little, in the USand UK before he founded

Optimised Operations Limitedin 2001. He specialises inhelT!>i '.'

26 CONTROL Number 4 2004 iomnet.org.uk