Fall 2009 David Harris

MGMT631 Project ManagementMGMT631 Project Management

Slides Six

Fall 2009 2

Session ObjectivesSession Objectives

Planning & scheduling Work Breakdown Structure Realistic Estimating Critical Path Time Boxes & Critical Chain

Fall 2009 3

Fall 2009 4

Project Time Management ProcessesProject Time Management Processes

Project time management involves processes required to ensure timely completion of a project

Processes include: Activity definition Activity sequencing Activity duration estimating Schedule development Schedule control

Work Breakdown Structure (WBS)Work Breakdown Structure (WBS)

WBS breaks down the work of the project WBS breaks down the work of the project into manageable chunksinto manageable chunks

WBS can be graphical or outlineWBS can be graphical or outline Start at top, progressively break work Start at top, progressively break work

down (tree structure) into work packagesdown (tree structure) into work packages Packages provide clear work assignmentsPackages provide clear work assignments

Estimating: Work Breakdown StructureEstimating: Work Breakdown Structure “Project best understood by breaking it down

into its parts” Work Breakdown Structure (WBS)

powerful tool for doing this (not just a task list) defines the total scope of the project fundamental to much of project planning & tracking

Start at top, progressively break work down (tree structure) into work packages

Roll up the packages for bottom up estimating Packages give clear work assignments

Sample Intranet WBS Sample Intranet WBS Organized by PhaseOrganized by Phase

Intranet WBS in Tabular FormIntranet WBS in Tabular Form1.0 Concept

1.1 Evaluate current systems1.2 Define Requirements

1.2.1 Define user requirements1.2.2 Define content requirements1.2.3 Define system requirements1.2.4 Define server owner requirements

1.3 Define specific functionality1.4 Define risks and risk management approach1.5 Develop project plan1.6 Brief web development team

2.0 Web Site Design3.0 Web Site Development4.0 Roll Out5.0 Support

Intranet WBS and Gantt ChartIntranet WBS and Gantt Chart

Project 98 file

Intranet WBS & Gantt Chart Organized by PM GroupsIntranet WBS & Gantt Chart Organized by PM Groups

Fall 2009 11

Realistic Estimating (Frame)Realistic Estimating (Frame)

Lots of reasons for poor estimates inexperience, technical problems, changes

optimists, low-balling, politics

Bottom-up cost estimating rollup the WBS packages

Top-down or Parametric estimating from experience to complex models

Fall 2009 12

Realistic Estimating (cont.)Realistic Estimating (cont.)

Which technique is better? ideally use both early on don’t have WBS so must use

top-down accuracy of top-down depends on

availability/quality of historical data building complete WBS can be

expensive, but guesses can be even more costly

Fall 2009 13

Deliverables and MilestonesDeliverables and Milestones Deliverables

Tangible, verifiable work products Reports, presentations, prototypes, etc.

Milestones Significant events or achievements Acceptance of deliverables or phase

completion Cruxes (proof of concepts) Quality control Keeps team focused

The seeds of major software disasters are usually sown in the first three months of commencing the software project. Hasty scheduling, irrational commitments, unprofessional estimating techniques & carelessness of the project management function are the factors that tend to introduce terminal problems. Once a project blindly lurches forward toward an impossible delivery date, the rest of the disaster will occur almost inevitably.

T. Capers Jones

Fall 2009 15

Software Engineering MetricsSoftware Engineering Metrics

Lines of Code (LOC)Function Points

Fall 2009 16

Lines of Code (LOC)Lines of Code (LOC)

Most traditionally used metric for project sizing

Most controversial Count comments? Declaring variables? Efficient code vs. code bloat Language differences Easier to count afterwards than to estimate

Fall 2009 17

Function PointsFunction Points

Analysis based on evaluation of data and transactional types: Internal Logical File (ILF) External Interface File (EIF) External Input (EI) External Output (EO) External Inquiry (EQ)

Fall 2009 18

Function PointsFunction Points

Value Adjustment Factor based onDegrees of Influence (DI) aka Processing

Complexity Adjustment (PCA)General Systems Characteristics (GSC)

Total Adjusted Function Points (TAFP)transformed into development estimatesconverted in equivalent LOC

Backfiring

The Application Boundary for The Application Boundary for Function Point AnalysisFunction Point Analysis

Fall 2009 20

The Mythical Man-Month – Frederick BrooksThe Mythical Man-Month – Frederick Brooks

1.1. 0ur techniques of estimation are poorly 0ur techniques of estimation are poorly developed. More seriously, they reflect developed. More seriously, they reflect an unvoiced assumption which is quite an unvoiced assumption which is quite untrue i.e., that all will go well.untrue i.e., that all will go well.

2.2. Our estimating techniques fallaciously Our estimating techniques fallaciously confuse confuse effort with progresseffort with progress, hiding , hiding the assumption that men and months the assumption that men and months are interchangeable.are interchangeable.

Fall 2009 21

The Mythical Man-Month – Frederick BrooksThe Mythical Man-Month – Frederick Brooks

3.3. Because we are uncertain of our estimates, Because we are uncertain of our estimates, software managers often lack the courteous software managers often lack the courteous stubbornness of Antoines chefstubbornness of Antoines chef

4. Schedule progress is poorly monitored. 4. Schedule progress is poorly monitored. Techniques proven & routine in other Techniques proven & routine in other engineering disciplines are considered radical engineering disciplines are considered radical innovations in software engineeringinnovations in software engineering

Fall 2009 22

The Mythical Man-Month – Frederick BrooksThe Mythical Man-Month – Frederick Brooks

5. When schedule slippage is recognized, the 5. When schedule slippage is recognized, the natural tendency (& traditional) response it natural tendency (& traditional) response it to add more manpower. Like dousing a fire to add more manpower. Like dousing a fire with gasoline, this makes matters worse, with gasoline, this makes matters worse, much worse. More fire requires more much worse. More fire requires more gasoline & thus begins a regenerative cycle gasoline & thus begins a regenerative cycle which ends in disaster. which ends in disaster.

Fall 2009 David Harris

Brooks Law:

“Adding manpower to a late software project makes it

later.”

Fall 2009 24

Scheduling Tools & MethodsScheduling Tools & MethodsGantt ChartsCritical Path Method (CPM)Program Evaluation and Review

Technique (PERT)Time BoxesCritical Chain

(Theory of Constraints)

Fall 2009 25

Developing a Project ScheduleDeveloping a Project Schedule Base documents

Project charter

– start/end dates (tentative), budget Scope statement

– what will be done (& not done)

Activity definitions develop detailed WBS plus supporting explanations to understand

all work to be done

Fall 2009 26

Activity SequencingActivity Sequencing Review activities & determine

dependencies Mandatory dependencies: inherent in the

nature of the work; hard logic Discretionary: defined by the project team;

soft logic External: involve relationships between

project and external activities Must Must determine dependencies to use

critical path analysis

Fall 2009 27

Gantt ChartsGantt Charts Created 1800 Standard format for displaying

project schedules activities, durations, start/end finish

dates displayed in calendar format Advantages

enforces planning easy to create & understand preferred for summary/exec-level

information

Fall 2009 28

Gantt Charts (cont.)Gantt Charts (cont.)

Symbols include: black diamond: milestones Thick black bars: summary tasks Lighter horizontal bars: tasks Arrows: dependencies between tasks

Bar Charts Simplified version Serve similar function

Fall 2009 29

Sample Project Gantt ChartSample Project Gantt Chart

Sample Tracking Gantt ChartSample Tracking Gantt Chart

white diamond: slipped milestone

two bars: planned and actual times

Fall 2009 31

Critical Path Method (CPM)Critical Path Method (CPM) Developed 1957 CPM diagram shows activities,

durations, start/end dates & sequencesequence in which they must be completed

Critical path for project is the series of activities that determines the earliest earliest timetime by which project can be completed

Critical path is longest pathlongest path through network diagram, has least (zero) slack or float

Fall 2009 32

CPM (cont.)CPM (cont.)

Critical path helps you make schedule trade-offs

Slack or floatSlack or float amount of time activity can be delayed without delaying early start of dependent activities

Simple Example of Determining Critical PathSimple Example of Determining Critical Path consider following network diagram assume all times in days

2 3

4

5

A=2 B=5C=2

D=7

1 6

F=2

E=1

start finish

a. How many paths are on this network diagram?

b. How long is each path?

c. Which is the critical path?

d. What is the shortest amount of time needed to complete this project?

Activity-on-Arrow (AOA) Network Diagram

Fall 2009 34

Precedence Diagramming Method (PDM)Precedence Diagramming Method (PDM)

Activities are represented by boxes Arrows show relationships between

activities More popular than ADM method as

used by PM software Better at showing different types of

dependencies

Task Dependency TypesTask Dependency Types

Activity Analysis for AONActivity Analysis for AON

Activity Description Estimated Duration (Days)

Predecessor

A Evaluate current technology platform

2 None

B Define user requirements 5 A

C Design Web page layouts 4 B

D Set-up Server 3 B

E Estimate Web traffic 1 B

F Test Web pages and links 4 C,D

G Move web pages to production environment

3 D,E

H Write announcement of intranet for corp. newsletter

2 F,G

I Train users 5 G

J Write report to management 1 H,I

Fall 2009 37

Activity on the Node (AON) Network DiagramActivity on the Node (AON) Network Diagram

Possible Activity PathsPossible Activity Paths

Possible Paths Path Total

Path 1 A+B+C+F+H+J 18

2+5+4+4+2+1

Path 2 A+B+D+F+H+J 17

2+5+3+4+2+1

Path 3 A+B+D+G+H+J 16

2+5+3+3+2+1

Path 4 A+B+D+G+I+J 19*

2+5+3+3+5+1

Path 5 A+B+E+G+I+J 17

2+5+1+3+5+1

Fall 2009 39

Critical PathCritical Path Longest path Shortest time project can be completed

Zero slack (or float)The amount of time an activity can be delayed

before it delays the project Must be monitored and managed!

PM can expedite or crash Can fast track The CP can change Can have multiple CPs

Fall 2009 40

Techniques for Shortening a Project ScheduleTechniques for Shortening a Project Schedule

Shortening durations of critical tasks: add more resources change their scope

Crashing tasks by obtaining the greatest amount of schedule compression for the least incremental cost

Fast tracking tasks by doing them in parallel or overlapping them

Fall 2009 41

PERTPERT

Developed 1959 for Polaris project Similar to CPM but addresses

uncertainties in task durations Uses probabilistic time estimates –

optimistic, most likely, pessimistic estimates of activity durations

PERT Formula and ExamplePERT Formula and Example

PERT weighted average formula: optimistic time + 4X most likely time + pessimistic time

6

Example:

PERT weighted average =

8 workdays + 4 X 10 workdays + 24 workdays = 12 days

6

where 8 = optimistic time, 10 = most likely time, and 24 = pessimistic time

Normal DistributionNormal Distribution

PERT DistributionPERT Distribution

Activity Analysis for PERTActivity Analysis for PERTActivity Predecessor Optimistic

Estimates (Days)

Most Likely Estimates

(Days)

Pessimistic Estimates

(Days)

Expected Duration

(a+4b+c)

6

A None 1 2 4 2.2

B A 3 5 8 5.2

C B 2 4 5 3.8

D B 2 3 6 3.3

E B 1 1 1 1.0

F C,D 2 4 6 4.0

G D,E 2 3 4 3.0

H F,G 1 2 5 2.3

I G 4 5 9 5.5

J H,I .5 1 3 1.3

30 days 31.6 days

Fall 2009 46

Selecting Scheduling ApproachSelecting Scheduling Approach Consider project size, risk and

complexity Gantt

senior management smaller, less complex projects

CPM medium size/complexity/risk

PERT high risk projects medium to high complexity

Time Box SchedulingTime Box Scheduling

Deals with delivering in short time frames

Focuses on prioritizationFacilitator brings stakeholders

together to agree on prioritiesIt’s customer–developer partneringpartnering

win-win win-win Compromise on scope to achieve

early delivery

Critical Chain SchedulingCritical Chain Scheduling Addresses challenge of meeting or beating project

finish dates Application of the Theory of Constraints (TOC) Developed by Eliyahu Goldratt in his books The Goal The Goal

and Critical Chainand Critical Chain Method of scheduling that takes limited resources into

account when creating project schedule & includes buffers to protect completion date

Assumes resources do not multitask as it often delays task completions & increases total durations

Multitasking ExampleMultitasking Example

Fall 2009 50

Buffers and Critical ChainBuffers and Critical Chain

A buffer is additional time to complete a task

Critical chain schedule removes buffers from individual tasks & instead creates A project buffer, which is additional time

added before the project’s due date Feeding buffers, which are addition time

added before tasks on the critical path

Critical Chain Scheduling ExampleCritical Chain Scheduling Example

Fall 2009 52

Critical Chain (cont.)Critical Chain (cont.)

Simple view: in critical situation, don’t try to strengthen

all of the links in the chain focus on the weakest link

Some organizations view as best thing since sliced bread

But does it sound like common sense?

Fall 2009 53