parametric project monitoring and control · 2017-05-19 · october 24, 2005 1 parametric project...

1October 24, 2005

Parametric Project Monitoring and Control:Parametric Project Monitoring and Control:PerformancePerformance--Based Progress Assessment and PredictionBased Progress Assessment and Prediction

Presented by:Mike Ross, Chief Engineer

Galorath Incorporated100 North Sepulveda Boulevard

Suite 1801El Segundo, California 90245

480.488.8366 (o) 480.488.8420 (f)[email protected]

Galorath Inc. 2005All Rights Reserved

October 24, 2005 2

SummarySummary

Software projects fail more often than notProject success � Good managementMeasurement objectifies managementSoftware projects are governed by dynamic properties• Properties currently accounted for in the Project Planning process

• Properties should also be accounted for in the ProjectMonitoring and Control process

Project Monitoring � Performance Measurement4-D Earned Value objectifies progressProject Control � Control LimitsRe-Baselining � Performance-Based ForecastingCommunication is essential to successful project management


October 24, 2005 3

Things are Getting Better; however,Things are Getting Better; however,There’s Still Room For ImprovementThere’s Still Room For Improvement

Success: The project iscompleted on time and onbudget, with all featuresand functionsChallenge: Over budget,over time, offers fewerfeatures than originallyspecifiedFailure: Project iscancelled prior tocompletion16%

53%

31%

27%

33%

40%

26%

46%

28%

28%

49%

23%

34%

51%

15%

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

1994 1996 1998 2001 2004

Project Resolution

FailureChallengeSuccess

Source: Standish Group International, Inc.“CHAOS” studieswww.standishgroup.com

$0

$50

$100

$150

Billions ofDollars

1995 1998 2004

The Price of Failure

OverrunsFailure

How does ineffectivemanagement of resources

(people, time, $) contribute tothis problem?


October 24, 2005 4

Mnemonic Aid forMnemonic Aid forSoftware Project ManagementSoftware Project Management

Planning – estimating, scheduling

Resourcing – interviewing, hiring, motivating

Organizing – establishing interpersonal communication pathsand rules, mapping resources to tasks

Training – teaching, mentoring

Equipping – acquiring and allocating equipment, tools,materials, supplies, products etc.

Controlling – directing, measuring, correcting and/or replanning

Transitioning – delivering, reviewing, analyzing, archiving

Michael A. ROSS Consulting & Training 2002All Rights ReservedReprinted by Galorath Inc. with permission from ROSS

October 24, 2005 5

Project Management ContextProject Management Context

Know

ledg

e &Ex

perie

nce

Baseline(s) &Final Actuals

To-date

ActualsMeasure

&Analyze

Monitor&

Control

Baseline

Estimate&

Plan


October 24, 2005 6

Process Focus (CMMI™)Process Focus (CMMI™)

Project Planning• Establish Estimates• Develop a Project Plan• Obtain Commitment to the Plan

Project Monitoring and Control• Monitor Project Against Plan• Manage Corrective Action to

ClosureMeasurement and Analysis• Align Measurement and Analysis

Activities• Provide Measurement Results


October 24, 2005 7

Desire Software

Labor

Software Development andSoftware Development andMeasurementMeasurement

StartFinish

Effort(person-months)

Size(work units)

Size(work units)

Defects(count)

Time(calendar months)

SoftwareDevelopment

Process


Staffing(people, time)

Cost(currency)

Technology

Friction

Effective Technology(coefficient)

October 24, 2005 8

Fundamental MeasuresFundamental Measures

SizeEffective TechnologyTimeEffort � Cost, StaffingDefects


October 24, 2005 9

Estimate DefinedEstimate Defined

es·ti·mate (es′ti mit), n.an approximate judgment or calculation, as of thevalue or amount of somethinga prediction that is equally likely to be above or belowthe actual result (Tom DeMarco)

A WELL FORMED ESTIMATEIS A DISTRIBUTION


October 24, 2005 10

3 Laws of Software Development3 Laws of Software DevelopmentDynamicsDynamics

Brooks’ Law (Software Equation)• Adding people to a late project makes it later.• Development time (duration) and development effort (labor) are not linearly

interchangeable.Paul Masson’s Law Applied to Software Development (Minimum TimeEquation)• No [software] before its time.• Each and every project, by its nature (technical difficulty), can effectively

handle only so much staffing acceleration; therefore, there exists, for eachand every project, some minimum achievable development time.

Parkinson’s Law (Optimal Effort Equation)• Work expands so as to fill the time available for its completion.• There exists, for each and every project, some point of maximum

productivity; i.e., some point that represents the most efficient use of laboron the project.


October 24, 2005 11

Software Development DynamicsSoftware Development Dynamics


For a given Size and Technology

Impo

ssible

Paul Masson’sLaw

Parkinson’sLaw

Inefficient Brooks’Law

Minimum Time

Optimal Effort

Elapsed Calendar Time (months)

Effo

rt (p

erso

n-m

onth

s)

Reasonable

October 24, 2005 12

Software Project ManagementSoftware Project ManagementOut of Control ProcessOut of Control Process

SoftwareDevelopment

ProcessEstimating Scheduling DirectingDesire


October 24, 2005 13

Software Project ManagementSoftware Project ManagementAd Hoc ProcessAd Hoc Process

SoftwareDevelopment


Measuring


October 24, 2005 14

Software Project ManagementSoftware Project ManagementPartially Managed ProcessPartially Managed Process

SoftwareDevelopment


Measuring


October 24, 2005 15

Software Project ManagementSoftware Project ManagementFully Managed ProjectFully Managed Project

SoftwareDevelopment


Measuring


October 24, 2005 16

Performance Measurement:Performance Measurement:Measures and MetricsMeasures and Metrics

Fundamental Cost of Work Measures• Baseline Budget – Budget at Completion (BAC)• Planned – Budgeted Cost of the Work Scheduled (BCWS)• Earned – Budgeted Cost of the Work Performed (BCWP)• Spent – Actual Cost of the Work Performed (ACWP)

Variances (Differences between Cost of Work Measures)• Schedule Variance (SV)• Cost Variance (CV)• Budget Variance (BV)• Time Variance (TV)


October 24, 2005 17

Performance Measurement:Performance Measurement:Measures and MetricsMeasures and Metrics

Performance Indices – (Ratios Between Cost of Work Measures)• Schedule Performance Index (SPI)• Cost Performance Index (CPI)• Budget Performance Index (BPI)• Time Performance Index (TPI)• Composite Performance Index (XPI)• To-Complete Performance Index (TCPI)

Status and Forecasting Metrics• Estimate at Completion (EAC)• Estimate to Complete (ETC)


October 24, 2005 18

Three Unit Systems forThree Unit Systems forPerformance Measurement ValuesPerformance Measurement Values

Monitary Value – units of currency; e.g.:• $• £• €

Effort Value – units of labor; e.g.:• person-hours, staff-hours, effort-hours, labor-hours• person-months, staff-months, effort-months, labor-months

Normalized Value – unitless• % of full scale


October 24, 2005 19

44--D Earned ValueD Earned Value

SDLC Primary Activity CompletionArtifact CompletionMilestone CompletionDefect Discovery / Removal

October 24, 2005 20

Schedule Accomplishments ChartSchedule Accomplishments Chart


BV / BPI

ScheduleSlip

SV / SPI

TV / TPI

CV / CPI

CostOverrun

October 24, 2005 21

Example Project: Metrics ChartsExample Project: Metrics Chartsat Project Start (Initial Plan)at Project Start (Initial Plan)


October 24, 2005 22

Example Project: Metrics ChartsExample Project: Metrics Chartsat System Design Reviewat System Design Review


October 24, 2005 23

Example Project: Metrics ChartsExample Project: Metrics Chartsat Software Requirements Reviewat Software Requirements Review


October 24, 2005 24

Example Project: Metrics ChartsExample Project: Metrics Chartsat Preliminary Design Reviewat Preliminary Design Review


October 24, 2005 25

Example Project: Metrics ChartsExample Project: Metrics Chartsat Critical Design Reviewat Critical Design Review


October 24, 2005 26

PerformancePerformance--Based Forecasting andBased Forecasting andReRe--BaseliningBaselining

1. Start a new estimate2. Update size estimate3. Update technology assumptions4. Update schedule assumptions5. Update staffing assumptions6. Update labor rate and FTE assumptions7. Time now calibration8. Communicate the results9. Re-Baseline the project


October 24, 2005 27

Example Project: Metrics ChartsExample Project: Metrics ChartsUpdate Size EstimateUpdate Size Estimate


October 24, 2005 28

Example Project: Metrics ChartsExample Project: Metrics ChartsUpdate Technology AssumptionsUpdate Technology Assumptions


October 24, 2005 29

Example Project: Metrics ChartsExample Project: Metrics ChartsUpdate Staffing AssumptionsUpdate Staffing Assumptions


October 24, 2005 30

Example Project: Metrics ChartsExample Project: Metrics ChartsReRe--Baseline the ProjectBaseline the Project


October 24, 2005 31

Example Project: Metrics ChartsExample Project: Metrics Chartsat Code & Unit Test Completeat Code & Unit Test Complete


October 24, 2005 32

Example Project: Metrics ChartsExample Project: Metrics ChartsUpdate Staffing AssumptionsUpdate Staffing Assumptions


October 24, 2005 33

Example Project: Metrics ChartsExample Project: Metrics ChartsReRe--Baseline the ProjectBaseline the Project


October 24, 2005 34

Example Project: Metrics ChartsExample Project: Metrics Chartsat Component Int. & Test Completeat Component Int. & Test Complete


October 24, 2005 35

Example Project: Metrics ChartsExample Project: Metrics Chartsat Program Test Completeat Program Test Complete


October 24, 2005 36

Example Project: Metrics ChartsExample Project: Metrics Chartsat Project Finish (Initial Delivery)at Project Finish (Initial Delivery)


October 24, 2005 37

ReviewReview

Software projects fail more often than notProject success � Good managementMeasurement objectifies managementSoftware projects are governed by dynamic properties• Properties currently accounted for in the Project Planning process

• Properties should also be accounted for in the ProjectMonitoring and Control process

Project Monitoring � Performance Measurement4-D Earned Value objectifies progressProject Control � Control LimitsRe-Baselining � Performance-Based ForecastingCommunication is essential to successful project management


parametric project monitoring and control · 2017-05-19 · october 24, 2005 1 parametric project...

Documents