PSP TSP XP CMMIPSP, TSP, XP, CMMI… eating the alphabet ... · Scrum and CMMI V1.3 6 5 Optimizing...

1

PSP TSP XP CMMIPSP, TSP, XP, CMMI…eating the alphabet soup!eating the alphabet soup!

D i l M RDaniel M. RoySSTC 2011

19 May 2011

Copyright © 2010 STPP (Software Technology, Process & People)

PSP, TSP, Personal Software Process and Team Software Process are service marks of CMUCMM and Capability Maturity Model are registered in the U.S. patent and trademark office

Report Documentation Page Form ApprovedOMB No. 0704-0188

Public reporting burden for the collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources, gathering andmaintaining the data needed, and completing and reviewing the collection of information. Send comments regarding this burden estimate or any other aspect of this collection of information,including suggestions for reducing this burden, to Washington Headquarters Services, Directorate for Information Operations and Reports, 1215 Jefferson Davis Highway, Suite 1204, ArlingtonVA 22202-4302. Respondents should be aware that notwithstanding any other provision of law, no person shall be subject to a penalty for failing to comply with a collection of information if itdoes not display a currently valid OMB control number.

1. REPORT DATE 19 MAY 2011 2. REPORT TYPE

3. DATES COVERED 00-00-2011 to 00-00-2011

4. TITLE AND SUBTITLE PSP, TSP, XP, CMMI...eating the alphabet soup!

5a. CONTRACT NUMBER

5b. GRANT NUMBER

5c. PROGRAM ELEMENT NUMBER

6. AUTHOR(S) 5d. PROJECT NUMBER

5e. TASK NUMBER

5f. WORK UNIT NUMBER

7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) Software Technology, Process & People Inc,20 Forest Rd,Bradford Woods,PA,15015

8. PERFORMING ORGANIZATIONREPORT NUMBER

9. SPONSORING/MONITORING AGENCY NAME(S) AND ADDRESS(ES) 10. SPONSOR/MONITOR’S ACRONYM(S)

11. SPONSOR/MONITOR’S REPORT NUMBER(S)

12. DISTRIBUTION/AVAILABILITY STATEMENT Approved for public release; distribution unlimited

13. SUPPLEMENTARY NOTES Presented at the 23rd Systems and Software Technology Conference (SSTC), 16-19 May 2011, Salt LakeCity, UT. Sponsored in part by the USAF. U.S. Government or Federal Rights License

14. ABSTRACT

15. SUBJECT TERMS

16. SECURITY CLASSIFICATION OF: 17. LIMITATION OF ABSTRACT Same as

Report (SAR)

18. NUMBEROF PAGES

31

19a. NAME OFRESPONSIBLE PERSON

a. REPORT unclassified

b. ABSTRACT unclassified

c. THIS PAGE unclassified

Standard Form 298 (Rev. 8-98) Prescribed by ANSI Std Z39-18

2

Agendag

Robust, fragile, agileRobust, fragile, agileFrom CMMI to TSPThe manifestoThe manifestoXP and SCRUM vs. CMMIPSP and TSP at workPSP and TSP at workEarned value for the rest of usSSynergySurf the next wave!


3Different tacks


4

Manifesto


5

XP 12 KPs

The Planning Game Pair ProgrammingThe Planning GameSmall Releases

h

Pair ProgrammingCollective OwnershipC di S d dMetaphor

Simple DesignCoding Standards40-hour Week

Incremental testingRefactoring

On-site Customer Continuous integrationRefactoring Continuous integration


6Scrum and CMMI V1.35 Optimizing

4 Q tit t

Continuous processimprovement

Organizational Performance ManagementCausal Analysis and Resolution

Level Focus Process Areas

Requirements DevelopmentTechnical SolutionVerification

4 Quantitat.Managed

3 Defined

QuantitativemanagementOrganizationengineering

Organizational Process PerformanceQuantitative Project Management

VerificationValidationOrganization Process FocusOrganization Process DefinitionOrganizational TrainingIntegrated Project Management

processstandardization

Integrated Project ManagementRisk managementDecision Analysis and ResolutionProduct Integration

2 M d R i t t2 Managed BasicProject management

Requirements managementProject PlanningProject Monitoring and ControlSupplier Agreement ManagementMeasurement and AnalysisProcess and Product Quality Assurance


Process and Product Quality AssuranceConfiguration Management

7

From CMMI to TSPThe CMMI

derives from Schewart Deming Juran Crosby and others’– derives from Schewart, Deming, Juran, Crosby and others work

– describes universally accepted system development t ti t th i ti l lmanagement practices at the organization level

– provides a roadmap for improvement based on PDSAWatts’ latest body of work (PSP/TSP)Watts latest body of work (PSP/TSP)

– builds on these principles and results in the field– demonstrates development practices and improvement

techniques at the individual and small team level– provides a framework for data-driven improvement


8The experience factoryProject organization Experience factory

Environment projectSet goals

Choose processPl

TailoringEnvironment, projectcharacteristics

Processes modelsPlan

Experience

Processes, modelstools, components

E t l

pbase

Data, lessons learnedExecute planCollect data

PackagingProject/risk analysis

From ‘The experimental paradigm in software eng.’,Rombach, Basili, Selby, Springer-Verlag, 1994

9What’s a PSP?

The PSP is a self improvement paradigm based on individual procedures and dataindividual procedures and data

A small set of scripts, standards and forms A i l b t hi hl ff ti t f kA simple but highly effective measurement frameworkSelf management based on individual metrics

i i li i i lFostering commitment to quality principlesIt is an agile level 5 process for individualsIt is a also a one person implementation of Basili’s

Quality Improvement Paradigm (experience factory)


10

From a great founding fatherg g

“At the Software Engineering Lab (NASA GSFC) weAt the Software Engineering Lab (NASA GSFC) we have applied evolutionary improvement concepts (PDSA) to the development domain…

With this [PSP] book, Watts Humphrey has developed an evolutionary improvement paradigm at the personal level by providing a mechanism for learning through experience,

t d f db k ”measurement and feedback.”Vic Basili foreword to Watts’ bookA discipline for software engineering


discipline fo softwa e enginee ing

11

The PSP Process Flow

PSP0 Process

Requirements

PSP0 Process

Development

Planning

pDesign

Code

ProcessScripts Guide Time

andDefectL

Block diagr

Schematics

Postmortem

Compile

Test

LogsPlan

SummarySimulation/Prototype

Postmortem

Project and Process data


Finished Product Project and Process data

12Time Estimating Accuracy - % ErrorError

250300350

Erro

r

100150200

mat

e E

MaxClass

500

50100

% E

stim Min

-100-50

1 2 3 4 5 6 7 8 9 1

%


Program Number STPP, Inc. data, 2005

13PSP in Lagos, Nigeria


14

Data driven improvementData driven improvementDefect Fix Time By Typey y

160180

80100120140

20406080

080 50 20 70 30 40 90 10 60 100


15Yield - All Students, All PPrograms

120 Defects found BEFORE compile

100

120

%

Defects found BEFORE compileDefects injected before compileYield =

60

80

Yiel

d % Max

Avg

Cross-reviews

20

40

Y MinPersonal reviewsIntroduced at day 7

01 2 3 4 5 6 7 8 9 10 11

y

Copyright © 2010 STPP (Software Technology, Process & People)Program Number STPP, Inc. data, 2005

16

PSP Return On InvestmentPSP Return On InvestmentThe following table shows projected integration and system Example from Microsoft

It took 250 engineers an entire year to removeg p j g ytest (I&S) rework costs per KLOC for engineers before and after PSP training.

It took 250 engineers an entire year to remove30,000 defects from Windows NT 4.0At 2,000 h per staff year that’s 500khOr 500/30=16.7 ~17h per defectU l l f h b i 10 40h/d f i I/SBefore PSP After PSP

Defects/KLOC into Unit Test 40 10

Yield for Unit Test 50% 50%

Usual rule of thumb is 10-40h/defect in I/S test

Yield for Unit Test 50% 50%

Defects/KLOC into I&S Test 20 5

I&S Defect fix time/KLOC 200 50

Integration and system test rework savings are 150 hours per KLOC (training costs recovered in less than 6 weeks of work!)


( g )

17PSP: An experience workshop?

PSP exercise PSP experienceExercise

Set goalsStudy process

Pl

ImprovementExercisecharacteristics

Process modelPlan

PSP elements

Process, modeltools, components

E t l

PSP elements,data & reports

Data, lessons learnedExecute planCollect data

Post MortemImmediate feedback

18What’s a TSP?

The Team Software Process uses the PSP principles toApply the PSP sound engineering discipline to project work– Apply the PSP sound engineering discipline to project work

– Form and help perform outstanding self directed teams– Produce an aggressive but realistic plan own by the teamCMMI SCAMPI A results since 2002:

maturity level 1 to 2 takes 14 months– Proactively track project progress against the plan– Continuously measure and improve the processes– To produce world class products

maturity level 1 to 2 takes 14 monthsmaturity level 2 to 3 takes 17 monthsmaturity level 3 to 4 takes 15.5 monthsmaturity level 4 to 5 takes 12.5 monthsp p

PSP/TSP accelerate CMMI maturation by a factor 2..3 (level 1 to 4 in 24 months)1

y

The TSP is an agile level 5 process for small teamsThe TSP starts with a project “launch”


1- Julie Switzer, NAVAIR TSP symposium 2008

19What’s a launch?Day 1 Day 2 Day 3 Day 4

1. Establish product and

business goals

4. Build top-down and next-phase

plans

7. Conductrisk

assessment

9. Holdmanagement

review

2. Assign rolesand define

5. Developthe quality

8. Preparemanagementbriefing and

Launchpostmortemteam goals plan briefing and

launch reportpostmortem

6. Build bottom-up and

consolidatedplans

3. Produce development

strategyand process


20Building High-Performance Teams

The TSP strategy is to

TeamManagement

Team communicationTeam coordination

Project trackingRisk analysis PThe TSP strategy is to

improve performance from the bottom up. Team

Goal settingRole assignment

Risk analysis

TS

P

Thi t t t t Team

Building

Process discipline

Tailored team processDetailed and balanced plans

PThis strategy starts with PSP training.

TeamMemberSkills

Process disciplinePerformance measures

Estimating and planning skillsQuality management skills

PS

P


21PSP/TSP impact Average Schedule Deviation - Range

140%160%

Average Effort Deviation - Range

100%

120%

40%60%80%

100%120%140%

20%

40%

60%

80%

100%

-20%0%

20%

Pre TSP/PSP With TSP/PSP-20%

0%

20%

Pre TSP/PSP With TSP/PSP

Defects/KLOC in Acceptance Test - Range

0.70.80.9

Post-Release Defects/KLOC - Range

1.2

1.4

0.20.30.40.50.6

0 2

0.4

0.6

0.8

1


00.1

Pre TSP/PSP With TSP/PSP0

0.2

Pre TSP/PSP With TSP/PSP

22Earned value for the rest of us

My weekPrepare presentationD i li l

Estim.5h3h

EV25%15%

Sum Done by5h Tuesday8h T dDesign a little

Attend useless meetingCode a littleTest a little

3h1.5h7h2 5h

15%7.5%35%12 5%

8h Tuesday9,5h Wednesday16.5h Friday19h FridayTest a little

Write status report2.5h1h20h

12.5%5%

19h Friday20h Friday


23First PSP/TSP project in PRC S b 2008 M h 2009September 2008..March 2009

Cumul at i ve Pl an/ Ear ned Val ueCumul at i ve Pl an/ Ear ned Val ue

100 0

120. 0

60 0

80. 0

100. 0

PVReplan: 3 5% per week

20. 0

40. 0

60. 0 EV

Actual: 35% in 10 weeks

Plan: 4% per weekReplan: 3.5% per week =>65/3.5=19 more weeks

0. 0

20. 0

W1 We W5 W7 W9 W11

W13

W15

W17

W19

W21

W23

W25

W27


24Predictable SchedulesSchedule Deviation Individual Value Control Chart -

Commercial Systems350

150200250300

on CMM introduced

050

100150

% D

evia

ti CMM introduced

TSP introduced

-150-100

-500

01/8801/89

01/9001/91

01/9201/93

01/9401/95

01/9601/97

01/98

Date of Project Start

Individual Data Points Mean Upper Natural Process Limit

Lower Natural Process Limit One Standard Deviation


[Source: AIS]

25Rigor continuumIronclad contract

SandboxhackingCMM-I

XP

Ri k d i d lRisk driven models

SCRUM

PSP/TSP


26Agile CMMI vs. “pure” agileSimilarities

Self directed teamsDifferences

True blue agile quite a bit more:Self directed teamsWell defined rolesWorking sw and biz value

True blue agile quite a bit more: – collegial– code drivenWo g sw a d b va ue

Incremental developmentManager as coach

– project centered– test focused– time boxed

Quick reaction/incrementalReviews/inspections

And quite a bit less worried about– detailed process documentation

l h l i / kiMeaningful meetingsPeople interactionCultural evolution

– early phases planning/tracking– product doc. (except code)– detailed data (but there is some)


Cultural evolution – CMM stuff!!!

27

It’s the People stupid!

LevelsPeople CMM Threads

Developing Building Motivating Shaping

p p

5

p gCapability &Competency

gWorkgroups &

Culture

g& ManagingPerformance

p gthe

Workforce

ContinuousCapability

ContinuousWorkforce

OrganizationalPerformanceOptimizing

4

CapabilityImprovement

Mentoring

WorkforceInnovation

CompetencyIntegration Quantitative

PerformanceOrganizational

Capability

PerformanceAlignment

Predictable

3

CompetencyBased Assets

CompetencyDevelopment

EmpoweredWorkgroups

WorkgroupDevelopment

PerformanceManagement

CompetencyBased Practices

CapabilityManagement

WorkforceDefined

2

CompetencyAnalysis

Training and

ParticipatoryCulture

Communication

CareerDevelopment

CompensationPerformanceManagement

Planning

Staffing


Managed Development & Coordination ManagementWork Environment

Staffing

28Process synergy: AIM

CMMI - Buildsi ti l yorganizational

capability

capa

bilit

y

TSP - Buildsquality products

on cost and s pr

oces

s

on cost and schedule

–en

able

s

PSP - Buildsproject team

skill & disciplineP-

CM

M –


29

AIM: The next waveIndia has successfully used the CMM

di i j f id bl i f“ The quality of aIn five years, I want Indian companies enjoy a formidable reputation for cost advantageB t i i li i t d d ill d thi d t

The quality of a software system is governed by

y ,the world to be asking, “How did India software industry

But rising living standards will erode this advantageCompetition is ramping up

Chi i i 40% d i h

is governed by the quality of th d

g,Mexico do it?”

60000

70000

80000

D i– China sw is growing 40% per year and is cheaper– Ex soviet union labor is plentiful and price competitive

Latin America countries want their share

the process used to develop and

l i ”30000

40000

50000

60000

$USMillions

Domestic

Exports

– Latin America countries want their share

Every competing country must now demonstrate a consistent measurable quality advantage

evolve it.”Watts Humphrey 1986

0

10000

20000

1997 1999 2001 2004 2005 2006 2007 2008 2009 2010


consistent measurable quality advantage—Watts Humphrey, 1986Fiscalyear

30Conclusion

Identify, mentor, nurture and lavishly praise local l d hileadership– Get a (preferably knowledgeable) champion

Work the management chain before and after the TSP like– Work the management chain before and after the TSP-like SPI planning session

Make the culture data hungry one byte at a timeMake the culture data hungry one byte at a timeYou’re not good at planning? Plan often!Surf the next wave: Go PSP/TSP (it is part of AIM)Surf the next wave: Go PSP/TSP (it is part of AIM)Watch failure cost of quality go down, show ROI and

celebrate… But keep going!


celebrate… But keep going!

31Questions?


PSP TSP XP CMMIPSP, TSP, XP, CMMI… eating the alphabet ... · Scrum and CMMI V1.3 6 5 Optimizing...

Documents

Transcript of PSP TSP XP CMMIPSP, TSP, XP, CMMI… eating the alphabet ... · Scrum and CMMI V1.3 6 5 Optimizing...