14.15 Prof.saji Gopinath

7/25/2019 14.15 Prof.saji Gopinath

1/43

Conference 2012 PMI KeralaConference 2012 PMI Kerala June 09, 2012June 09, 2012

Critical Chain Approach toCritical Chain A

pproach to

Project Success

Project Success

Dr. Saji opinathDr. Saji opinath

IIM Ko!hi"o#eIIM Ko!hi"o#e

PMI- Project Management Conference 2012PMI- Project Management Conference 2012


2/43


Moti$ationMoti$ation

% &he'e of &his conference&he'e of &his conference

(Project Mana)e'ent* a +ife s"ill

% hat is co''on in all our projects-hat is co''on in all our projects-

( &i'e o$errun

( Cost o$erruns-

% h-h-

% More i'portantlMore i'portantl

( h such o$erruns happen rarel in our /personalprojects

(A ne a of loo"in) at projects-


3/43


PMI Stu#PMI St

u#


4/43


hat #o e ant fro' Project Mana)e'ent -hat #o e ant fro' Pro

ject Mana)e'ent -

% elia3le on ti'e in full to 3u#)et #eli$er perfor'anceelia3le on ti'e in full to 3u#)et #eli$er perfor'ance

More re$enue, 'ore Profit, happ custo'ers

% A sta3le planA sta3le plan

More Pro#ucti$e use of resources

% Si'ple, o3jecti$e 'easures of Project pro)ress 4 project health statusSi'ple, o3jecti$e 'easures of Project pro)ress 4 project health status

Shorter 'eetin)s, 3etter infor'e#sta"ehol#ers * less aste, 'ore pro#ucti$it

% Clear si)nals for hen correcti$e action is *Clear si)nals for hen correcti$e action is * an# is notan# is not* necessar* necessar

5etter #irecte# reco$er efforts * less aste,

'ore pro#ucti$it

% Direction for on)oin) i'pro$e'ent effortsDirection for on)oin) i'pro$e'ent efforts

&he future3rin)s 'ore re$enue, 'ore profit, happier

custo'ers than the present


5/43


An# hat e nor'all )et66An# hat e nor'all

)et66

% elia3le on ti'e in full to 3u#)et #eli$er perfor'ance -A continuous stru))le ith ti'e, cost an# scope -

% A sta3le plan

epeate# resche#ulin) -

% Si'ple, o3jecti$e 'easures of Project pro)ress -

Clarit at the start an# en#, thic" fo) in 3eteen -

% Measures of Project health status -Su3jecti$e assess'ents co'poun#e# 3 hu'an factors -

% Clear si)nals for hen correcti$e action is * an# is not *necessar -

Inter$enin) too 'uch too earl, an# too little too late -

% Direction for on)oin) i'pro$e'ent efforts -7e8ll i'pro$e our 'etho#s hen thin)s )et 3etter7


6/43


% Do our projects fail 3ecause of /3a# pannin)-Do our projects fail 3ecause of /3a# pannin)-

( is plannin)the 'ost i'portant PM S"ill-:

% Is the #elas happen #ue to /A Class ite's;Is the #elas happen #ue to /A Class ite's;

Acti$itiesActi$ities

% h #o our project finish hen it is on /Missionh #o our project finish hen it is on /Mission

Mo#e-Mo#e-

e$isitin) the 3eleifs


7/43Conference 2012 PMI KeralaConference 2012 PMI Kerala June 09, 2012June 09, 2012

% Classical Project Mana)e'entClassical Project Mana)e'ent( Perfor'ance = Increasin) Co'ple>it

( e*loo" at the fun#a'ental assu'ptions

% &echnical an# 5eha$ioral Di'ensions of Project&echnical an# 5eha$ioral Di'ensions of Project

Mana)e'entMana)e'ent

( h our projects are #elain)-

% e*thin"in) the a e 'ana)e our projectse*thin"in) the a e 'ana)e our projects



h #o our Projects #ela-h #o our Projects #ela-

% &he ti'e esti'ates are too ti)ht&he ti'e esti'ates are too ti)ht

% &he chan)es in en$iron'ent is too #rastic&he chan)es in en$iron'ent is too #rastic

% &he interfacin) a)encies are inefficient&he interfacin) a)encies are inefficient

% It is i'possi3le to pre#ict the ti'e of co'pletionIt is i'possi3le to pre#ict the ti'e of co'pletion

of a ?or" pac"a)e@ acti$it:of a ?or" pac"a)e@ acti$it:

% thersthers

-B # ti t

j t # ti -



Bo #o e esti'ate project #uration-Bo #o e esti'ateproject #uration-

% Past DataPast Data

% &op Don or 5otto' up approach&op Don or 5otto' up approach

% Cushions for uncertaintCushions for uncertaint

% Acti$it prece#ence netor" techniue:Acti$it prece#ence netor" techniue:

% Bo 'uch is the cushion ou pro$i#e--Bo 'uch is the cushion ou pro$i#e--



Project DurationProject Duration

% hat is the effect-hat is the effect-

TheoryTheory

ActualActual

The effect is that the typical cushions you provide onThe effect is that the typical cushions you provide on

each activity is around 0-100!each activity is around 0-100!

" #" #

$ 20$ 20

%%

1212



Con$entional Project Mana)e'entCon$entional Project Mana)e'ent

Task Time EstimatingTask Time Estimating

Take best guess at how long a task willTake best guess at how long a task willtaketake

Consider the efect o unknowns orConsider the efect o unknowns or

unplanned interruptionsunplanned interruptions

Add sucient saety to be able to deliverAdd sucient saety to be able to deliver

with 90 probabilitywith 90 probability



0!"

0!#

0!$

0!%

&!0

0

Time

2! 0!

&0!

T0 T&0



Pro3le's ith the con$entional project 'ana)e'ent techniuesPro3le's ith the con$entional project 'ana)e'ent techniues

% &hree 'echanis's that inflate ti'e esti'ates&hree 'echanis's that inflate ti'e esti'ates

( &he orst*case scenario

( A## safet ti'e to ensure project is on ti'e

( 5 inflatin) ori)inal ti'e to protect a)ainst a )lo3al cut

Median = 2 days Estimated = 5 days

Safety time =

3 days

Confidence

level 80-90%

0'$

0'"

0'2

0'1

1 2 " $ % #

Probability of

Comletin!

ro"ect #it$in

n&mber of

days

'&mber of days toComlete ro"ect

&o "eep project on sche#ule * $aria3ilit encoura)es

people to pa# in#i$i#ual acti$it ti'es ith safet ti'e.

50%



(uffers on )ach Activity(uffers on )ach Activity

*hy after all these cushions project get delayed+*hy after all these cushions project get delayed+



% &hree as to aste Safet ti'e&hree as to aste Safet ti'e( passin) on of pre$ious #elas

Depen#encies 3eteen acti$ities cause #elas to accu'ulateDepen#encies 3eteen acti$ities cause #elas to accu'ulate

( stu#ent sn#ro'e

ait until the last 'inute to start a tas"ait until the last 'inute to start a tas"

( 'ulti*tas"in)

Multitas"in) cause# 3 li'ite# resources resource contention :Multitas"in) cause# 3 li'ite# resources resource contention :

% How the Safety Gets WastedHow the Safety Gets Wasted

( A##in) safet ti'e to resol$e the resource contention pro3le' #oesnot help

( e a## safet e$erhere an# then e aste it6

Con$entionalCon$entional SAFETY TIMESAFETY TIMEinflates project co'pletion ti'einflates project co'pletion ti'e

5ut.. A##in) safet ti'e to protectin) project o3jecti$es

sel#o' or"s



,ue ,ate,ue ,ate

tart ,ate .inish ,ate ,ue ,atetart ,ate .inish ,ate ,ue ,ate

)ffect of .luctuation)ffect of .luctuation

5 h i l Di i f P j t i t5 h i l Di i f P

j t i t



5eha$ioral Di'ension of Project n$iron'ent5eha$ioral Di'ension of Project n$iron'ent

% A##iti$e uleA##iti$e ule

( Co''it'ents of #uration an# total cost of a project are3ase# upon a##in) up the #uration an# cost of in#i$i#ual

tas"s

% Par"inson@s +aPar"inson@s +a

( or" e>pan#s to fill its ti'e

% E*Minute uleE*Minute ule

( It@s not ualit if it is finishe# 3efore ti'e is up

% Boc"e Stic" Sn#ro'e stu#ent sn#ro'e:Boc"e Stic" Sn#ro'e stu#ent sn#ro'e:( aitin) to start a tas" #ue to 'ore i'portant or" at han#

&he Conflicte on ct


18/43


&he Conflicte on ct

ManageManage

ProjectsProjects

uccessfullyuccessfully

ManageManage

ProjectsProjects

uccessfullyuccessfully

AvoidAvoid

Par/isnonsPar/isnons

aa

AvoidAvoid

Par/isnonsPar/isnons

aa

MinimiseMinimise

ProjectProject

eadtimeeadtime

MinimiseMinimise

ProjectProject

eadtimeeadtime chedulechedule

*ithout*ithout

safetysafety

chedulechedule

*ithout*ithout

safetysafety

Meet ProjectMeet Project

PromisePromise

Meet ProjectMeet Project

PromisePromise

ProvideProvide

.or.or

MurphyMurphy

ProvideProvide

.or.or

MurphyMurphy

chedulechedule

*ith*ith

A.)T3A.)T3

chedulechedule

*ith*ith

A.)T3A.)T3

ConCon

flictflict


19/43


a ut-a ut-

% Plan A * in$est our ener) in re#ucin) the e>tent ofPlan A * in$est our ener) in re#ucin) the e>tent of

the $aria3ilitFthe $aria3ilitF* Alloin) lon)er in Project plannin) sta)e for preparin) esti'atesAlloin) lon)er in Project plannin) sta)e for preparin) esti'ates

* &rainin) staff in esti'atin)&rainin) staff in esti'atin)

* Gse of for'al esti'atin) 'etho#sGse of for'al esti'atin) 'etho#s

* Measurin) pro)ress an# fee#in) results 3ac" into esti'atin) practiceMeasurin) pro)ress an# fee#in) results 3ac" into esti'atin) practice

* More #etaile# specificationsMore #etaile# specifications* +ess fle>i3ilit o$er chan)es to specifications+ess fle>i3ilit o$er chan)es to specifications

* &rainin) the staff 3etter in their jo3 content&rainin) the staff 3etter in their jo3 content

* Gsin) in#i$i#ual perfor'ance 'easures to i#entif poor perfor'ersGsin) in#i$i#ual perfor'ance 'easures to i#entif poor perfor'ers

* Keeping projects short ( ! "onths#$ %rea&ing 'arger nderta&ingsKeeping projects short ( ! "onths#$ %rea&ing 'arger nderta&ings

into se)era' short *rojectsinto se)era' short *rojects

Doin) this can help, 3ut #oesn8tDoin) this can help, 3ut #oesn8t sol$esol$ethe pro3le'the pro3le'


20/43


Plan 5 * Copin) 3eha$ioursPlan 5 * Copin) 3eha$iours

% Project Mana)ersProject Mana)ers fi)htfi)ht to 3e assi)ne# the 'ost $ia3le Projectsto 3e assi)ne# the 'ost $ia3le Projects

% Project Mana)ersProject Mana)ers fi)htfi)htfor the 3est stafffor the 3est staff

% Project Mana)ersProject Mana)ers fi)htfi)ht to "eep the Project scope #onto "eep the Project scope #on

% Project Mana)ersProject Mana)ers e>ploite>ploitchan)es in scope to un#ul e>ten#chan)es in scope to un#ul e>ten#

ti'elines an# 3u#)etsti'elines an# 3u#)ets

% Project Mana)ersProject Mana)ers uituitlon) Projects ell 3efore the #eli$er #atelon) Projects ell 3efore the #eli$er #ate% Project Mana)ersProject Mana)ers #isre)ar# tar)ets#isre)ar# tar)etsthe "no to 3e i'possi3lethe "no to 3e i'possi3le

% Staff or"Staff or" #ou3le shifts#ou3le shiftsin the final ee"s ; 'onthsin the final ee"s ; 'onths

% Du'pin) theDu'pin) the 3la'e3la'eelsehereelsehere

Doin) these 'a help theDoin) these 'a help the in#i$i#ual,in#i$i#ual,3ut not the3ut not the or)anisationor)anisation


21/43


e cane can re#ucere#uce$aria3ilit, 3ut e cannot$aria3ilit, 3ut e cannot eli'inateeli'inateit,it,

3ecause it is3ecause it is inherentinherentto the nature of a Projectto the nature of a Project

Plan CFPlan CF Approach the pro3le' in a #ifferent aApproach the pro3le' in a #ifferent a

e 'uste 'ust 'ana)e'ana)ethethe $aria3ilit$aria3ilitthat re'ainsthat re'ains

h t i th t-h t i th

t-


22/43


hat is the a out-hat is the a out-

C iti l Ch i P j t M tC iti l Ch i P

j t M t


23/43


Critical Chain Project Mana)e'entCritical Chain Project Mana)e'ent

% A#aptation of Principles of &heor of constraintsA#aptation of Principles of &heor of constraints

%Applin) &C concepts to project 'ana)e'entApplin) &C concepts to project 'ana)e'ent( Critical Chain is a project 'ana)e'ent application of

the &heor of Constraints &C:

(Accor#in) to &C, the 'ain constraint in an project is

the ti'e ta"en for co'pletion of the +ritica' +hain

% Critical Chain Project Mana)e'entCritical Chain Project Mana)e'ent

( Con$entional approach focuses on successful on*ti'e

co'pletion of each in#i$i#ual acti$itin a project( &C approach focuses on successful on*ti'e

co'pletion of the entire project

C iti l Ch i P j t M tC iti l Ch i P

j t M t


24/43



TheThe TOC philosophyTOC philosophyapplied to pro'ectapplied to pro'ect

management attempts to remove themanagement attempts to remove the

undesirable efects (late) over*budget) andundesirable efects (late) over*budget) and

under*perormance pro'ects+ by attackingunder*perormance pro'ects+ by attacking

individual measurements and uncertainty!individual measurements and uncertainty!

,hat is T-C .hilosophy/,hat is T-C .hilosophy/

&h f C t i t&h

f C t i t


25/43


&heor of Constraints&heor of Constraints

A system improvement philosophy (asA system improvement philosophy (asopposed to a process improvementopposed to a process improvement

philosophy+philosophy+ -rganiations live or die as systems) not-rganiations live or die as systems) not

asas processesprocesses

1uccess or ailure is a unction o how well1uccess or ailure is a unction o how welldiferent component processesdiferent component processesinteractinteractwith one anotherwith one another

&heor of Constraints&he

or of Constraints


26/43



1ystems are analogous to chains) or1ystems are analogous to chains) or

networks o chainsnetworks o chains 2ike a chain) a system2ike a chain) a system @@s perormance iss perormance is

limited by the perormance o itslimited by the perormance o its

weakest linkweakest link

The weakest link is the systemThe weakest link is the system@@ss

constraintconstraint

&h f C t i t&he

or of Constraints


27/43



Another basic principle o T-CAnother basic principle o T-C

A large number o undesirable efects will becaused by a relatively small number o coredrivers

Eliminating a very ew core problems canresult in a huge improvement


28/43


Bo e han#le $aria3ilit in Critical ChainBo e han#le $aria3ilit in Critical Chain

% e #o not 3uil# in an contin)enc at the &as" le$ele #o not 3uil# in an contin)enc at the &as" le$el

% e 'o$e all the contin)enc to the Project le$el * call this thee 'o$e all the contin)enc to the Project le$el * call this the

Co'pletion 5ufferCo'pletion 5uffer

In#i$i#ual&as"scan no 3e late ithout affectin) the

co'pletion #ate of the Project

&he Project #ue #ate is protecte# as lon) as

the accu'ulate# lateness alon) an onechain is less than the co'pletion 3uffer

3ow does all this relate to pro'ects/3ow does all this relate to pro'ects/


29/43


hat #ifference #oes this 'a"e to our pro3a3ilit of 3ein) late -hat #ifference #oes this 'a"e to our pro3a3ilit of 3ein) late -

Gn#er 8nor'al8 practice, ifGn#er 8nor'al8 practice, if any tas&any tas& isis later thanlater than its contin)encits contin)enc

alloancealloance, e ha$e a pro3le', e ha$e a pro3le'

Gn#er Critical Chain, e onl ha$e a pro3le' if theGn#er Critical Chain, e onl ha$e a pro3le' if the tota'tota'latenesslatenesse>cee#s the total contin)ence>cee#s the total contin)enc

&his secon# con#ition is&his secon# con#ition is 'uch less li"el'uch less li"elthan the firstthan the first H +a of a$era)esH +a of a$era)es

; Central li'it theore'; Central li'it theore'an#an# increasin)lincreasin)lso as the nu'3er of tas"sso as the nu'3er of tas"s

increasesincreases

Critical Chain Project Mana)e'entCritical Chai

n Project Mana)e'ent


30/43



4ncertainty always present4ncertainty always present ((it doesnit doesn@@t go awayt go away

Take the saety out o each o the critical path tasksTake the saety out o each o the critical path tasks

and lump them into a saety net at the end o theand lump them into a saety net at the end o the

pro'ectpro'ect

5dentiy constraints along the path and set up5dentiy constraints along the path and set up

bufers in ront o tasks that can sufer rom thebufers in ront o tasks that can sufer rom the

constraint (constraints 6 time and resources+constraint (constraints 6 time and resources+

Allow tasks to start when predecessors areAllow tasks to start when predecessors are

completed and resources are availablecompleted and resources are available


31/43


4evisiting activities4evisiting activities


32/43


Poer of AggregationPoer of Aggregation


33/43


e'o$e safet ti'e fro' in#i$i#ual acti$itiese'o$e safet ti'e fro' in#i$i#ual acti$ities

Jo3

Jo3 2

Jo3 E

Jo3 1

Con$entional Project Sche#ule &as" 3ufferssafet ti'e: are hi##en

ithin in#i$i#ual acti$ities

Critical Chain Sche#ule

5uffers are poole#,

an# 'a#e e>plicit

*roject ,ffer,

% Safet 3uffersSafet 3uffers

% Poole# 3uffersPoole# 3uffers

Project 5uffer is safety time added to the end of the critical chain

to protect the completion date of the project'


34/43


% ee#in) 3uffer on the non*critical pathee#in) 3uffer on the non*critical path

ee#in) 5ufferProject 5uffer

If Slac" re'ains,

then sche#ule as

late as possi3le

Critical Chain

(eedin!

at$

.eeding 5uffers are designed to protect the critical chain from

delays on non-critical paths


35/43


% esource 3uffersesource 3uffers

( a a"eup call to alert resources to 3e rea# to or" on

critical tas"s

( Sche#ule# i#le ti'e can pro$i#e 3etter info a3out resource@sa$aila3ilit capacit:

esource

5uffers

Critical Chain

Project

5uffer

ee#in)

5uffer

A'ert W&r A

A'ert W&r ,

A'ert W&r +




36/43



Critical ChainCritical Chain* set o tasks which determines overall* set o tasks which determines overallpro'ect duration) taking into accountpro'ect duration) taking into account both precedence andboth precedence and

resource dependencies7resource dependencies7improvement along Critical Chain willimprovement along Critical Chain will

likely result in improvements to the pro'ect as a whole7likely result in improvements to the pro'ect as a whole7

improvements elsewhere will notimprovements elsewhere will not

Project buferProject bufer* protects pro'ect commitment dates rom* protects pro'ect commitment dates rom8uctuations on the Critical Chain8uctuations on the Critical Chain

Feeding buferFeeding bufer* protects Critical Chain rom 8uctuations* protects Critical Chain rom 8uctuationson eeding tasks7 provides the possibility or Critical Chainon eeding tasks7 provides the possibility or Critical Chain

tasks to start earlytasks to start early

Resource buferResource bufer* protects the Critical Chain rom lack o* protects the Critical Chain rom lack o

availability o reuired resources7 also provides the possibilityavailability o reuired resources7 also provides the possibilityor Critical Chain tasks to start earlyor Critical Chain tasks to start early




37/43



2 " $

1 2 " $ Project (uffer

Tas/ 1

-riginal Critical .ath

-riginal Critical .ath with:ufer

(1aety removed rom individualtasks+


38/43



39/43





40/43



About :ufersAbout :ufers

5dentiy the points at which to place pro'ect) eeding)and resource bufers

:ufer sies determined appro;imately) based eitheron average task duration estimates) or acombination o average and worst*case duration

estimates 5ndividual bufer sies can be ad'usted based onintuitive assessment o risk

:ufer insertion may cause the Critical Chain) andhence the pro'ect completion date) to be pushedlater




41/43



The Critical Chain approach to schedulingThe Critical Chain approach to scheduling

helps minimie pro'ect duration and ,5.)helps minimie pro'ect duration and ,5.)

delay investment as ar as possible) anddelay investment as ar as possible) and

ma;imie the chance o on*timema;imie the chance o on*time

completioncompletion

+ast or#+ast or#


42/43


+ast or#+ast or#

% &he #e$elop'ent of ne project 'ana)e'ent&he #e$elop'ent of ne project 'ana)e'ent

techniues ha$e not re#uce# uncertainttechniues ha$e not re#uce# uncertaint

% Bence e nee# as toBence e nee# as to 'ana)e'ana)ean#an# not a$oi#not a$oi#

uncertaintuncertaint

% Critical Chain Mana)e'ent is a a to achie$eCritical Chain Mana)e'ent is a a to achie$e

thisthis


43/43

( &han" Lou

saji-ii"&.ac.in / saji.gopinath-yahoo.insaji-ii"&.ac.in / saji.gopinath-yahoo.in

01222324320122232432

14.15 Prof.saji Gopinath

Documents