purpose in radically - lean frontiers · measures the just-in-time pace of market demand in...

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

31 JULY/AUGUST 2018 COST MANAGEMENT

In August 1988, entrepreneur Nor-man Bodek and his company, Pro-duc t iv it y Inc . , featured thesoon-to-be-famous lean consultingfirm Shingijutsu at Danaher Corpo-

ration in a fabulous workshop known as “5Days and 1 Night.” This was the prototypefor the now ubiquitous f ive-day kaizenworkshop. Beginning on Monday, consul-tants f rom Shingi jutsu and Product iv it yInc. led an improvement team of Danahermanagers, engineers, and guests from othercompanies in fol lowing Danaher’s opera-tors with stopwatches to gather data on theshop floor. The six lean metrics starred inExhibit 1 were employed to analyze the dataand redesign the f low of production. Onone very long Wednesday night, contrac-tors phys ica l ly reorganized, rew ired,replumbed, and reprogrammed the factory.On Thursday morning, Danaher employees

used the same metrics to test and validatefour functioning lean production cells. Inthe weeks that followed, Danaher supervi-sors and employees used the same metricsto manage daily production and highlightthe need for further improvement. Practi-t ioners of lean manufac tur ing and leantransactional processes — including healthcare — are intimately familiar with thesemetr ics , which have appeared in manykaizen workshops during the last 29 years.1

My goal is to make sense of these metricsf rom a cost management perspect ive. Inorder to do so, I must add one more metricto the l ist, namely, takt t ime. Takt t ime (t),or takt for short, is the rat io between t imeavailable to produce and expected demand:t = t ime avai lable to produce ÷ expecteddemand.

Every thing in lean production happensjust-in-t ime, as measured by takt; and, as

DECODING VALUE AT

THE SOURCE:

THE ABCs OF LEAN PRODUCTION

METRICSTOM JACKSON

T O M JAC K S O N i s manag ing d i re c t o r o f Mos s Adams LLP in S ea t t l e , Wa sh ing ton , and a membe r o f t h e RonaCon su l t ing Group, Mos s Adams’s l ean con su l t ing p ra c t i c e . Tom ha s b een t ea ch ing and con su l t ing in l ean s ince 1988and i s t h e au t h o r o f s e v e ra l b o ok s on s t ra t e g y d e p l o yment and l e an h e a l t h c a re . P l e a s e a dd re s s i n qu i r i e s t oTo m . Ja c k s o n @ Mo s s Ad a m s . c o m .

This ar t icle i l lustrates how lean production metr ics function

as par t of a sel f - regulating cost management system.

. . . . . . . . . . .

32 COST MANAGEMENT JULY/AUGUST 2018 LEAN METRICS

EXHIBIT 1 Lean Production Metrics

METRIC DEFINITION / EQUATION

PURPOSE IN RADICALLY DECENTRALIZED CONTROL SYSTEM

CT = cycle t ime of thecel l

Measured empir ical ly in seconds

Ensures adherence to target labor and materialcost in product ion cel ls ; ensures work bal -ance.

OCT = operator cyclet ime


Measu r e s a c t u a l wo r k c on t e n t o f t a s ksequences assigned to product ion operators.

D = demand Measured both histor ical lyand in real t ime

Ensures synchronizat ion of product ion wi thboth forecast and/or actual demand.

Floor space Measured empir ical ly in square f t .

Ensures adherence to targeted non-labor con-vers ion cost ; used to design, manage, andimprove layout of product ion areas.

K = # signal cards orkanban K = D • (L + b) ÷ Q

Where D = demand, L = lead t ime to replen-ishment, b = buffer stock, and Q = transferbatch size.

L = transfer batch leadt ime


Ensures that the r ight # of units are del iveredto t he r i gh t ce l l a t t he r i gh t t ime t o keepproduct ion moving at takt.

n = operators needed n = Σ ΟCT ÷ t Ensures that takt t ime is met at target cost.

Q = transfer batch sizeEstabl ished internal ly and

in negotiat ions with suppl iers

Str ic t ly contro ls the amount of inventory ineach transfer batch or container.

Scrap Measured empir ical ly in # units or lbs on gemba

Captures any material costs of poor qual i ty inproduct ion cel ls that are not ful ly quanti f iedby the loss of product ive t ime.

SWIP = standard workin process SWIP = [f(n)]

Ensures synchronizat ion with takt t ime in pro-duction cells, and adherence to target materialhandl ing cost.

t = takt t ime t = net t ime avai lable D

Measu res the j us t - i n - t ime pace o f marke tdemand in seconds; used to synchronize al laspects of product ion with demand.

Total lead t ime Measured empir ical ly inminutes or seconds

Measu res t he t ime pa r t s spend be ingt ransformed in ce l ls as we l l as t ime spentwait ing in buffers.

Total part / input path Measured empir ical ly in l inear f t .

Ensu res adhe rence t o t a rge t ma te r i a lmanagement cost between product ion cel ls;used to reduce L , replenishment lead t ime.

V = value-added rat io V = value-added t ime total lead t ime

Ensures adherence to target labor cost in pro-duct ion cel ls; used to opt imize n (operatorsneeded) and other variables.

Value-added t ime Measured empir ical lyin seconds

Measures how much t ime is spent “add ingvalue” by changing the form, fit, and/or functionof materials or information.

★

★ indicates or iginal Shingi jutsu/Product ivi ty Inc. metr ics

★

★

★

★

★

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Ben jamin Frank l in once s a id , “Time i smoney.”2 Practical ly speaking, takt t ime isthe average pace at which production mustmove to meet product ion targets w ithinthe t ime ava i l ab le — that i s , w i thoutovert ime. In each of Toyota’s factories, anelectronic production control board hangsabove the factory floor. On this board, takttime is prominently displayed together withproduction targets and running totals offinished production. Many lean metrics aremathemat ical funct ions of takt. It is thecontext in which every lean metric must beunderstood.

During the design phase of the productl ife cycle, Toyota works with its suppliersto achieve a target cost of the project’s com-ponents and materials. Target cost ing is asignificant departure from the tradit ionalpract ice of cost plus, in which producerssimply add an acceptable margin to whatevertheir costs happen to be. In target cost ing,the object ive is to price backward from amarke t pr i ce and then deduc t a prof i tacceptable to investors. The result is thetarget cost . In a process known as valueengineering, the manufacturer and its sup-pliers meticulously calibrate the form, fit ,function, and weight of ever y componentto achieve the target.3

The lean metrics in Exhibit 1 operation-alize the target cost as the product entersthe product ion phase of its l i fe cycle. Inproduction, the only problem that remainswith respect to cost is to hit the target. Al lthe information needed to control cost thatis not a l ready embedded in the product

design, bill of materials, and new productionequipment must be embedded in the pro-duc t ion l ayout and work f low of ever yoperator. This is the function of lean pro-duction metrics.

Of course, there is nothing new aboutt a rge t cos t ing . Many Amer ic an andEuropean companies star ted adopting itdecades ago. The only difference, accordingto Robin Cooper, be tween Japanesecompanies and the rest of the world is thatJapanese companies l ike Toyota actual lyhit the target. 4

Radical decentralizationIt i s genera l l y ag reed upon that thetradit ional standard cost ing model of theaccounting profession does not provide ther ight informat ion or prov ide it in a waythat is t imely enough to support advancedmanufacturing methods such as lean pro-duct ion. 5 How do lean organizat ions getthe right cost information when they needit? It’s s imple. Instead of rely ing on costvariance reports prepared a month or moreaf ter the fac t , operators control cost inexactly the way they control qualit y : Theyreliably detect and correct deviations fromstandard or target cost and qualit y condi-t ions as they occur on the shop floor.

Understanding how cost managementmoved from variance reporting to empoweredemployees requires a short history lesson.6

Economists v iew the histor y of businessorganization as a study in decentralization.In the mid-1800s, the American railroads

33LEAN METRICS JULY/AUGUST 2018 COST MANAGEMENT

EXHIBIT 2 The DNA of Lean Cost Management

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

and, later, big business adopted the com-mand-and-control systems of the Prussianmi l it ar y es t abl i shment . Beg inning w ithGeneral Motors in 1919, however, businesseshave progressively decentralized their deci-sion-making, becoming increasingly adaptiveand lean or, in the current lingo, agile. Decen-tralized decision-making in business orga-nizations is comparable to parallel processingin computers: They both decrease latencyor slowness in information processing.7 Inother words, parallelization gives us fastercomputers; decentralization gives us fasterorganizations. What makes Toyota’s systemunique is its radical decentral izat ion. Ineffect, a lean organization is comparable toa massively parallel supercomputer. Toyota’sdecentralized model of cost management isindependent of its centralized managementand f inancia l account ing systems. As H.Thomas Johnson advised, “To become lean,shed accounting.”8

To unders t and how r ad ica l ly decen-tralized cost management works requiresanother reference. In a famous ar t icle inthe Harvard Business Review, Steven Spearand H. Kent Bowen explained that the infor-mation about Toyota’s decentralized man-agement s ys tems i s l i ke a DNA codeconsisting of four discrete letters. In Exhibit2, I have reordered and restyled this codefor purposes of our discussion of cost man-agement. 9

By decoding the f i rs t three le t ters ofToyota’s DNA, we will see how the lean pro-duction metrics in Exhibit 1 are systema-tically embedded into frontline operationsto keep cost under control without man-agement intervention.• A. Flow explores how process syn-

chronizat ion and value are measuredin lean value streams. Flow providesreference condit ions for target cost,standards of synchronicity, and valueagainst which dev iat ions can be mea-sured “at the source” to trigger imme-diate correct ive act ion.

• B. Standardized work explores howthe metrics in Exhibit 1 are used toactualize the condit ions of f low withinproduction cel ls and the pull systemsthat connect them to form valuestreams.

• C. Autonomation explores how leanmetrics are used by operators to main-tain the state of f low in equil ibrium bydetect ing and correct ing defects thatdisrupt the f low. Most defects areaddressed within takt t ime.The ar t icle w i l l conclude w ith a br ief

discussion of Plan, Do, Check, Act (PDCA)and the scientific method.

A. Flow. The first element of Toyota’s codeis f low. Flow is a reference state in whichproduction pathways are free from wasteand variability, as measured by the lean pro-


EXHIBIT 3 Modular Cost Management: Cells and Pull Systems

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

duction metrics. Of course, those pathwaysmust be suff iciently f ree f rom waste andvariability to ensure that the target cost setin the design phase of the product life cycleis met in the production phase. Lean prac-titioners normally describe flow pathwaysas value streams. See Exhibit 3.

Value streams have two components, cellsand pull systems, which behave as discretemodules in a lean cost management system.A production cel l is a highly compact ar-rangement of people, production equipment,materials or other inputs, and informationdesigned to meet customer demand at targetcost and t arge t prof i t . 10 Al l produc t ionactivities within cells are synchronized withdemand by means of takt t ime. Productioncel ls are connected by pull systems, whichconsist of s tores of mater ia ls , known assupermarkets, and information or signals,known as kanban. Kanban are used to syn-chronize the movement of materials in stan-dardized containers or t ransfer batchesf rom supermarkets to product ion cel l s ,where materials are integrated into the pro-

duction flow. In section B on standard work,we wil l explore how these same two com-ponents are used to construct lean costmanagement systems.

Within production cel ls, operators andmachines perform work that changes thephys ic a l form, f i t , and/or f unc t ion ofmaterials and information into productsor serv ices that customers are wi l l ing topay for. 11 Such ac t iv it y i s refer red to asvalue-added. All other act iv ity is referredto as was te , which i s non–va lue added.Exhibit 4 sets forth a list of the seven classicnon–value added wastes, al l of which maybe measured or interpreted as wastes oftime. The ratio of value-added time to cyclet ime (CT) — the t ime a unit spends at restor in process — is known as the va lue-added rat io (a starred element in Exhibit1). Production cells are normally engineeredto meet target cost at a value-added rat io(V) between 50 percent and 90 percent .Although there are exceptions, manual workinvolv ing cr it ica l th ink ing , such as thepractice of medicine, will tend to have lower


EXHIBIT 4 The Seven Wastes

WASTE DEFINITION MEASUREDBY

TIME CONVERSION

1 Overproduct ion Too many product ion units(more than the demand)

Time conver-sion

Non–valueadded t ime =# units × CT

2 OverprocessingToo many process steps (ornon–value added steps) toget the job done

Empir ical t imestudy N/A

3 Waits/delaysPeople, information,machines wait ing for value-added act ion


4 Inventory Work in process wait ing forvalue-added act ion

Physicalcount or t imeconversion

Non–valueadded t ime =# units × CT

5 TransportThe physical distance thepart or person must travel toget the f inal unit completed


6 MotionBending, l i f t ing, etc., to pickup or drop off assembliesand inventory


7 Defects/rework

Poor product or servicequal i ty that generates test-ing + undoing mistakes +reworking to restore qual i ty+ scrap

Physicalcount, t imeconversion,t ime study

Non–valueadded t ime =# defects ×CT

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

value-added content; work that can easi lybe standardized or automated, such as man-ufacturing assembly, will tend to have highervalue-added content. All activities that occurbetween cel ls — the storage of inventoryand the transportation of materials — areconsidered to be non–value added.

To summarize, the state of flow providesa reference state for lean cost managementsystems. Cells and pull systems are the basicbuilding blocks of lean production systemsas well as lean cost management systems.To achieve a state of f low, production mustmove at takt t ime with at least 50 percentva lue-added (w ith le s s than 50 percentwaste) within production cells. Theoretically,the value-added ratio in a true state of flowis high enough to ensure that target costcondit ions s e t in the des ign phase a resatisfied. As we wil l see in section B, in theproduction engineering phase, iterat ionsof improvement increase the actual value-added ratio unti l target cost is met.12

B. Standard work. Flow is defined by taktt ime and the va lue -added r at io — themetr ics that underl ie the cel l s and pul lsystems in a value stream. It is implementedthrough the indus t r i a l eng ineer ing ofs t andard work . S t andard work c an be

def ined as the best way an organizat ionknows how to perform a sequence of oper-ations. Technically speaking, standard workincorporates the five elements in Exhibit5. In the production engineering phase ofthe produc t l i fe c yc le , l ean produc t ionmet r ic s a re u sed to cont rol l abor andmaterials management costs. Flow is estab-l ished through the design and test ing ofstandard work, which is specifical ly engi-neered to meet takt time with minimal wasteor losses to non–value added act iv it ies.

Cell design. As i l lustrated in Exhibit 6,standard work for value-adding act iv it iesperformed within production cells is tested,improved, and retested until the target costis met at the correct levels of operator cyclet ime (OCT), operators needed (n) , ands t andard work in process (SWIP) . Anyremaining problems with product qualit yare also addressed at this point.

The typical lean industrial engineeringprocedure conducts repeated t ime studiesto discover and el iminate hidden waste.This may result in the el imination of non–value added time and subsequent reductionsin OCT and n . Of ten , operators can becross-trained so that once-specialized taskscan be per formed by even fewer, more


EXHIBIT 5 Elements of Standard Work

ELEMENT REQUIREMENT PURPOSE

Standard taskEvery discrete operation mustbe performed in exact ly thesame way. Es tab l i shes cond i t ions fo r

measuring defects as devi-ations from standard task andsequence. Makes defects dis-coverable in space, not t ime.Standard sequence

Every sequence of discreteoperations must be performedin exact ly the same order.

Standard t imeEvery sequence of operationsmust be performed, on aver-age, in the same amount oft ime (OCT) .

Trea t s e l emen ts o f cos texac t l y l i ke t ask andsequence, the e lements ofqual i ty. Makes defects dis-coverable in t ime and space.Standard work in process

Every sequence of operationsmust be supported by a con-stant level of work-in-processinventory (SWIP) .

Standard documentat ionEvery sequence of operationsmus t be tes ted and docu-mented together with normaltask t imes.

Supports systematic trainingand provides a basel ine forimprovement.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

capable operators. Reductions in n al lowmore compact arrangement of machinesand inputs. Reduct ions in f loor space of50 percent or more are also common, per-mitting the construction of additional cellsin the same area and reducing costs of space,heat, and l ight.13

Within lean production cel ls, materialsare control led by SWIP. The control ofSWIP has three purposes. First, it helps tosynchronize production with takt t ime byregulating how much work-in-process eachoperator physical ly controls at each stepof the process. Second, the optimizat ionof SWIP opt imizes f loor space used forstoring work-in-process. Third, SWIP setsan upper l imit on the CT of the cel l . As aprac t ica l mat ter, SWIP i s a func t ion ofoperators needed (i.e., SWIP f(n(t))).

SWIP can be equal to n, where each oper-ator operates one machine or workstat ionand no add i t iona l work- in-proce s s i srequired to facilitate the movement of inputsfrom one machine to another or the loadingor unloading of machines. SWIP may begreater or less than n depending on con-textual factors such as the degree of automa-tion, difficulty in loading and unloading

mach ines , and the nature of te amworkwithin the cel l .

Pull system design. Once product ioncells are capable of meeting takt and targetcost , at tent ion turns to the pul l systemsthat connect them. Pull systems are designedto automatical ly ensure that materials orinputs are moved to production cel ls onlywhen operators are ready to produce at takt.Pull systems are governed by the kanbanequat ion: K = D × (L + b) ÷ Q (refer toExhibit 1).

The kanban equation governs the numberof signal cards or kanban , K, that circulatebetween two productions cells. As illustratedin Exhibit 7, kanban cards are used by pro-duction operators to communicate to mate-rials managers the need to move parts fromsupermarkets to production cells for value-added processing.14 The equation also stan-dard i ze s the amount of i nventor y insupermarke t s p lu s any cont a iner i zedmaterial in transit between supermarketsand cells. This number is simply the numberof containers in circulat ion, g iven by K,t imes the maximum permissible amountin each container, given by Q. To audit theamount of material on the f loor, operators


EXHIBIT 6 Cost Management in Production Cells

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . only have to count the number of kanbani n c i rcu l at ion and mu l t ip ly i t by thecontainer or transfer batch size.

Management’s attent ion is more l ikelyto fal l on L , the replenishment lead t ime. Lis a function of the total input path betweenthe cel ls, another lean production metricstarred in Exhibit 1.15 If the total input pathcan be reduced by, say, 50 percent, so canthe direct labor component of mater ia lsmanagement. Like the work of direct pro-duction, the work of materials managers isorganized into cel ls w ith their own taktt imes and defined by standard work andits metrics n and SWIP. L can often be mea-sured by how long i t t ake s mater i a l smanagers to walk the distance of the partpath. Thus, the total part/input path oftenbecomes a target for improvement.

To summarize the logic of the lean metricsof standard work, the condit ions for f loware established within cells during the pro-duction engineering phase of the productlife cycle by regulating V, OCT, n, and SWIP.

This ensures that a l l cel ls w i l l meet taktt ime at target cost. The conditions for f lowbetween cells (i.e., across the value stream)are established by optimizing the total inputpath. This regulates L, the number of unitsthat circulate between cel ls, and the costof materials management.

C. Autonomation. Once cel ls are capableof meet ing target cost , how is the targetcost equilibrium maintained? What are the“unambiguous yes-and-no ways” of sendingand receiv ing signals that aler t operatorswhen dev iat ions from the reference con-dit ions of f low occur in actual operations?Operators maintain the equil ibrium stateof f low by responding to signals generatedby dev iat ions f rom f low. In this context,defects are defined as dev iat ions from thereference state of f low, defined in terms oflean product ion metr ics . Standard workand work- in-proce s s a re spec i f i c a l l ydesigned to allow operators to perform thisdemanding intel lectual work at the sametime they make products or deliver services.


EXHIBIT 7 Cost Management in Pull Systems

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

The bifurcation of duties into intel lectualand phys ic a l component s i s known a sautonomation.

Autonomation frees human operators tocritically assess and adjust production activ-it ies at the same t ime they perform theirva lue-added work. Informed by s igna lsdefined by lean metrics, autonomation isthe motor of self-regulat ion. It operates atthree levels: defect detection, containment,and prevention. Any or al l of these threeelements may be automated mechanical lyor electronical ly. In many cases, however,simple systems such as visual order or work-place organizat ion , v i sua l cont rol , andsimple checklists can achieve the requiredresult w ith l it t le capita l investment (seeExhibit 8). Takt time and the lean productionmetrics of value-added, CT, n , SWIP, andtotal input path make it easy for operatorsthemselves, as well as supervisors and man-agers, to discover dev iat ions from flow assoon as they occur. As i l lustrated in Exhibit9, the effect of autonomation is to detect,contain, and correct defects before the costof poor qualit y control begins to rise.

The concept of autonomat ion can bei l lus t rated by the prac t ice of s top- the-l i ne , w i t h wh ich mos t re ader s w i l l b efamil iar. At Toyota, whenever an operatordiscovers a defect within a production cell,he or she pauses product ion momentari lyto address it before al lowing the workpieceto proceed to the next step in the process.Fellow workers and supervisors assist withdiff icult problems. Product ion is stoppedcompletely only when the problem is toocomplex to be solved within the windowof takt t ime.

Because the system is engineered to meettarget cost at takt t ime, pausing the l ine toaddress routine production inefficiencieswithin takt ensures that Toyota can meetits cost targets. Because the reference stateof flow and the standard work that articulatesflow incorporate the elements of standardtime and SWIP, the stop-the-line systemensures that cost, as wel l as qualit y, is keptunder control. Operators pause the line notonly when they discover physical defects,such as a dented fender or an incorrect doseof medicine, but also when they are unable


EXHIBIT 8 Methods of Autonomation

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

to complete their standard work within taktt ime or when ac tua l work- in-process i sinconsistent with SWIP.

The process of discovery and autocor-rection is summarized in the window analysisin Exhibit 10. Square 1 of the window, upperlef t , s ignif ies the state of equi l ibrium inwhich operators know and adhere to standardwork, as measured by takt time and the leanproduction metrics. Squares 2–8 representreciprocal situations in which there are devi-ations from the standard, because operatorseither do not fol low or do not know thestandard. In squares 2–8, the response isstraightforward: Take steps to adhere to thestandard and restore the conditions of flow.In most cases, simple deviations from known,effective standards can be discovered andautocorrected within takt.16

To summarize, the metrics of standardwork and pull systems indicate, in real time,

when production and materials managementprocesses are either in control or out ofcontrol , g iv ing r i se not to managementintervention but to self-regulat ion. Withlean production metrics directly embeddedinto critical steps in the work, autonomationensures that (a) dev iat ions from a knownand effective standard are discovered quickly,and (b) when they are discovered, they arenormally corrected within takt t ime.

ConclusionWhat happens when problems that causedev iat ions f rom flow are too complex tobe discovered or solved within takt? Forexample, what happens in manufacturingwhen the prob lem i s t he produc t orequipment design? What happens in healthcare when a pat ient develops an unknowncondit ion? These are situat ions that land


EXHIBIT 9 The Cost of Poor Quality Control

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

us in square 9 in Exhibit 10. The four thletter of lean code, scientific method, speaksto Toyota’s commitment to approach suchproblems systematical ly, not only with thestat ist ica l tools of modern scient if ic in-vestigation, but also with the direct obser-vat ion and creat ive thinking of f rontl ineproduction operators. But this takes us farbeyond the scope of this ar t icle, which hasconcentrated on why lean organizat ionsdon’t land in square 9 when they can avoidit. Operators have al l the information theyneed to intervene in real t ime to correctmost problems.

Toyota’s cost management system designsand maintains a f low of product ion thatproduces products and serv ices at a target

cost , w ithout management inter vent ion.An economist would say that lean systemsare designed to maintain an equi l ibrium— that is , a self-regulat ing state of f low.The DNA of a lean production system keepstrack of the plan — including targets andmetrics — for the f low of production and,like biological DNA, stores a complete copyof that plan in every production cell. Autono-mation, including the practice of stop-the-line — informed by lean production metricsdeeply embedded in the work — v ir tual lyel iminates exponent ia l ly r i s ing costs ofpoor qua l i t y by ensur ing that the mostobvious dev iat ions from the standard aredetec ted and autocorrec ted w ith in t aktt ime. Toyota’s operators also have al l the


EXHIBIT 10 Window Analysis of Lean Cost Management

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

information they need to keep problemsfrom spreading. Using the scientific method,they work to prevent problems f romrecurring in the future and even to anticipateproblems with advanced predictive models.Lean metrics are designed to support bothautonomation and science, both practicedby a community of engaged and empoweredproblem solvers. n

NOTES1The ar t ic le focuses pr imar i ly on the s ix starredmet r i cs . Other met r i cs in Exh ib i t 1 have beenincluded to support th is invest igat ion, includingtakt t ime, kanban , and scrap. In fast-paced envi-ronments, lean metr ics are often strat i f ied intotheir stat ist ical components to embed informationmore deeply into operat ions.

2Franklin, B., “Advice to a young tradesman,” FoundersOnline (July 21, 1748). Avai lable at: http:/ /founders.archives.gov/documents/Frankl in/01-03-02-0130.

3Cooper, R. and Slagmulder, R., Target Costing andValue Engineering. (Portland, OR: Productivity Press,1997). See also: Cooper, R. and Slagmulder, R.,Supply Chain Development for the Lean Enterprise:Interorganizat ional Cost Management. (Port land,OR: Productivity Press, 1999).

4Private conversation with Robin Cooper at EmoryUniversity in Atlanta, Georgia, 1999.

5See, for example, Maskel l , B. and Baggaley, B. ,Pract ical Lean Account ing: A Proven System forM e a s u r i n g a n d M a n a g i n g t h e Le a n E n t e r p r i s e .(Port land, OR: Productivity Press, 2004): 134–136,173–174.

6Accountants may be assured that there is nothingi n so - ca l l ed l ean accoun t i ng t ha t v i o l a tes t heprinciples of GAAP. See Cunningham, J. and Fiume,

O., Real Numbers: Management Accounting in aLean Organizat ion (Managing Times Press, 2003):29–36.

7Jackson, T.L., “The r ise of the C-form organization,”Rona Consult ing (2006) (unpubl ished white paper).

8Johnson, H.T., Lean accounting: To become lean,shed accounting, Cost Management 20, no.1 (2006):6–17.

9Spear, S. and Bowen, H.K., Decoding the DNA ofthe Toyota production system, Har vard BusinessReview, (Sept/Oct 1999): 97–106.

10 I n hea l th -ca re p rocesses , pa t ien ts themse lvesbecome inputs, a long with medic ines and othersupplies. The cost equations explored in this articleare not changed.

11 In health care, value-added work can be said tochange the information, functional i ty, f i tness, andfeel ing of the patient’s wel l -being.

12When ce l ls are connected by pu l l systems, thevalue-added rat io necessary to meet target costdrops be low 50 percent , because of non–va lueadded work-in-process stored in pull system super-markets.

13Stubborn production bottlenecks and quality problemsthat prevent achieving target cost become the focusof an intensive team-based improvement interventionknown as 3P (production preparat ion process).

14 Kanban are also used to init iate the production ofinputs by upstream processes for consumpt ion( i .e., further processing, integrat ion, or assembly)by downstream processes.

15The term L is often accompanied by additional termsthat add extra t ime and inventory and that al lowmaterials managers to respond to normal var iat ionin demand and to more disruptions in production.

16For addit ional information about window analysis,see Fukuda, R., “Using window analysis for accuratefact finding and effective countermeasures,” BuildingOrganizational Fitness: Management Methodologyfor Transformation and Strategic Advantage (Portland,OR: Productivity Press, 1998).


Lean Accounting & Management Summit Bridge the Gap Between

Accounting & Operations

Lean Accounting & Management Summit

September 12-13, 2019 // Austin, TX

The Lean Accounting & Management

Summit (AKA Leadership Week) has long

served the accounting community by exploring

accounting’s unique and critical role in the lean

enterprise. The goal of the Summit is to align

Accounting and Operations to better measure

and motivate lean thinking across the organiza-

tion. Further, the Summit seeks alignment across

all business functions by running simultaneously

with the Lean HR & People Development Summit

giving participantsmore, and diverse, learning

options.

Join your accounting and management peers

September 12-13, 2019 in Austin, TX. You

will literally meet the “who’s who” list of lean

accounting and management practitioners.

Many of them current and former CEOs, CFOs,

and COOs in some of the world’s most respected

lean organizations.

www.leanfrontiers.com/leadershipweek

w w w . l e a n f r o n t i e r s . c o m

Austin, T X • September 12-13, 2019

Lean Practices for the

Accounting Department

Nick Katko

Life Cycle of Lean Accounting

Jean Cunningham

The Lean Profit Model

Mike DeLuca

High-Performing Kaizen System

Hal Macomber

Does Lean Reduce Cost? YES!

Nick Katko

Improvement Kata

for Accountants

Gary Kapanowski

Communicating Financial

Information: Talking about

dollars in a way that makes sense

Tamara Koopman

BlockChain, IOT, 3D printing, &

Other Emerging Technologies in

the Lean Enterprise

Rebecca Morgan

Bridging the Gap Between

Accounting & Operations in a

Lean Enterprise – Simulation

Jerry Solomon

purpose in radically - lean frontiers · measures the just-in-time pace of market demand in...

Documents