SE 313A Operating Systems Unit 1

EP 704Unit 6

Project Time Management

Dr. J. Michael Bennett, P. Eng., PMP

UNENE, McMaster University, The University of Western Ontario

Version 2K6-X-08

2K6-X-08 Dr. J.M. Bennett, P.Eng., PMP EP 7046-2

Unit 6 Project Time Management

Revisions

2K6-X-08 Initial Creation



EP 704 Road MapUnit 1 Introduction to Project ManagementUnit 2 The Project Management ContextUnit 3 Project Management ProcessesUnit 4 Project Integration ManagementUnit 5 Project Scope ManagementUnit 6 Project Time ManagementUnit 7 Project Cost ManagementUnit 8 Project Quality ManagementUnit 9 Project Human Resource ManagementUnit 10 Project Communications ManagementUnit 11 Project Risk ManagementUnit 12 Project Procurement Management



Process Time Management

Here we estimate the time and sequencing of WBEs

Must have the WBS done

The material is presented as sequential but likely will be significant overlap

In smaller projects, activity sequencing and duration and schedule development will be a single process done by the PM



Project Time Management Processes

6.1 Activity Definition

6.2 Activity Sequencing

6.3 Activity Resource Estimating

6.4 Activity Duration Estimating

6.5 Schedule Development

6.6 Schedule Control




.1 Inputs

.1 EEF .2 OPA . 3 Scope Statement .4 WBS .5 WBS Dictionary .6 PMP.2 Tools and Techniques .1 Decomposition .2 Templates .3 Rolling Wave Planning .4 XJ .5 Planning Component.3 Output . 1 Activity list .2 Activity Attributes . 3 Milestone List .4 Requested Changes

6.3 Activity Resource Estimating.1 Inputs .1 EEF .2 OPA .3 Activity List .4 Activity Attributes .5 Resource Availability .6 PMP

.2 Tools and Techniques . 1 Expert judgment .2 Alternatives Analysis .3 Published Estimating Data .4 PM Software .5 Bottom-up Estimating.3 Output .1 Activity Resource Requirements .2 Activity Attributes (Updates) .3 Resource Breakdown Structure .4 Resource Calendars (Updates) .5 Requested Changes

6.2 Activity Sequencing.1 Inputs .1 Scope Statement .2 Activity List . 3 Activity Attributes .4 Milestone List .5 Approved Change Requests.2 Tools and Techniques .1 Precedence diagramming (PDM) .2 Arrow diagramming (ADM) .3 Schedule Network Templates .4 Dependency Determination .5 Applying Leads and Lags.3 Output .1 Schedule Network Diagrams .2 Activity Lists (Updates) .3 Activity Attributes (Updates) .4 Requested Changes

PMI Project Time Management


Unit 6 Project Time Management PMI Project Cost Management

6.6 Schedule Control.1 Inputs .1 Schedule Management Plan .2 Schedule Baseline

.3 Performance Reports

.4 Approved Change Requests.2 Tools and Techniques .1 Progress Reporting .2 Schedule CC System .3 Performance Measurement .5 PM software .6 Variance analysis .7 Schedule Comparison Bar Charts

.3 Output .1 Schedule Model Data (Updates) .2 Schedule Baseline (Updates) .3 Performance Measurements .4 Requested Changes .5 Recommended Corrective Actions .6 OPA (Updates) .7 Activity List (Updates) .8 PMP (Updates) .9 Activity Attributes (Updates)

6.4 Activity Duration Estimating.1 Inputs .1 EEF .2 OPA .3 Scope Statement .4 Activity List .5 Activity Attributes .6 Activity Resource Requirements .7 Resource Calendars .8 PMP - Risk Register - Activity Cost Estimates

.2 Tools and Techniques .1 Expert Judgment .2 Analogous Estimating .3 Parametric Estimating .4 Three-point Estimates .5 Reserve Analysis.3 Output .1 Activity Duration Estimations .2 Activity Attributes (Updates)

6.5 Schedule Development.1 Inputs .1 OPA 2 Scope Statement .3 Activity List .4 Activity Attributes .5 Schedule Network Diagrams .6 Activity Resource Requirements .7 Resource Calendars .8 Activity Duration Estimates .9 Risk Register.2 Tools and Techniques .1 Schedule Network Analysis .2 Critical Path Method .3 Schedule Compression .4 What-if Scenario Analysis .5 Resource Leveling .6 Critical Chain Method .7 PM Software .8 Applying Calendars .9 Adjusting Leads and Lags .10 Schedule Model.3 Output .1 Project Schedule .2 Schedule Model Data .3 Schedule Baseline .4 Resource Requirements (Updates) .5 Activity Attributes (Updates) .6 Project Calendars (Updates) .7 Requested Changes .8 PMP (Updates)





6.3 Activity

Resource Est.

6.4 Activity

Duration Est.


Enterprise Environmental

12.4 Select

Sellers

4.3 Develop

PMP

4.6 Int Change

Control

Scope Statement

Organizational Process Assets

5.2 Scope Definition

5.3 Create

WBS

9.2 Acquire

Project Team

7.1 Cost

Estimation

10.3 Performance

Reporting

4.7 Close Project

4.4 Project Execution

11.2-5 Risk

Planning

6.6 Schedule

Control

WBS+D

Res Avail

History, Calendar

Resource Availability

Resource Availability

Activity Cost Estimates

Risk Register

Performance Reports

Work Performance Information

OPA Updates

History, Calendar

PMP, SMP

Activity List, Attributes, Milestone List Requested Changes

Requested Changes

PMP Updates

Schedule Network Diagrams

Requested Changes

Requested Changes

Requested Changes

Requested Changes

Approved Change Reqs

Act Res Rqs, RBS, Res Cals

Activity Duration Estimates

Schedule; Model data; Sched Baseline; Res Reqs; Proj Cal UD

Perf Mease; UDs of AL, AA, ;PMP, Sked BL; Proj Sked




1 Decomposition2 Templates3 Rolling Wave4 XJ5 Planning Comp

1 Activity List

2 Activity Atts

3 Milestone List

4 Req’ed Changes

1 EEF2 OPA3 Scope Statement4 WBS5 WBS Dictionary6 PMP



1 Activity List

Comprehensive list of all project schedule activities

Includes schedule identifier and enough detail for the team to understand what to do

Usd in the schedule model and the PMP

AL ≠ WBS components



2 Activity Attributes

Such asActivity identifierActivity codesActivity descriptionPredecessor activitiesSuccessor activitiesLogical relationshipsLeads and lagsResource requirementsImposed datesAssumptions and constraints



3 Milestone List

All milestones must be IDed

Mandatory or optional



4 Requested Changes

AD can generate changes



6.1.2 Activity Definition: T&T

1 Decomposition

2 Templates

3 Rolling Wave

4 XJ

5 Planning Component



Recall Simon’s Tripartite Division

Level 1 – Milestones

Level 2 – WBSes

Level 3 – Schedule Activities



EHB Example

Milestone – sign-off of Driver module

WBS – completed requirements

SA – review requirements with customer Stakeholders - ½ day 4 people



1 Decomposition

Subdivides WPs into smaller components; called schedule activities

These are schedule driven not deliverable driven



2 Templates

Useful from previous projects

Used to estimate resource skills, hours of effort, risk ID, etc



3 Rolling Wave

Progressive elaboration

Detail near-term events

Leave the far-term adumbrated, to be detailed later at a more appropriate time

Early on, leave far-term at the milestone level



4 XJ

Mumble mumble



5 Planning Component

When insufficient definition of the scope is available to decompose the WBS, leave the last node as a PC

Control Account: a node to be developed further later

Planning Package: subcomponent above the WP but below the CA



6.1.3 Activity Definitions: Outputs

.1 Activity List

.2 Activity Attributes

.3 Milestone List

.4 Requested Changes



1 Activity List

Includes all schedule activities needed to be done on the project

IncludesActivity identifier

Work description in enough detail to schedule and understand the work

SAs are discrete units but are NOT WPs



2 Activity AttributesExpansion of characteristics of SAs including

Activity IDActivity codeActivity descriptionPredecessor activitiesSuccessor activitiesLogical relationshipsLeads and lagsResource requirementsImposed datesConstraintsAssumptions



3 Milestone List

List of all milestonesMandatory

Optional

Will be part of the PMP and used for Milestone Scheduling



4 Requested Changes

Since we are fleshing out interior details of the WPs, changes will be needed as more detail is unfolded.

Fed through ICC of course.




1 PDM2 ADM3 Sched NW Temps4 Dependency Det.5 Leads and Lags

1 Scope Statement2 Activity List3 Activity Attributes4 Milestone List5 App’d Change Reqs

1 Sched NW Diags2 Act. Lists (Up)3 Act. Atts (Up)4 Req’d Changes



Modalities of Scheduling

Gantt/bar charts

Milestone charts

NetworksADM

Precedence

PERT



Problems with Each

Gantts do not show interdependencies

PERTs et al are time intensive, too much detail

Each is useful in their own right; not the final solution



Example: Gantt, Milestone, PERT

1144141

21

53

76

4

1 2

54

6

3

7

2

4 53

6 7

3

2 2

4

1 3

1

14

7

2



Network Fundamentals

Shown through a diagram. Visualization of:Activity interdependence

Project completion time

Impact of early/late starts

Trade-off analysis

“what if” scenarios

Cost of crashing

Slippages in planning/performance

Performance evaluation



6.2.2 Tools & Techniques

1 Precedence Diagramming Method (PDM)

2 Arrow Diagramming Method (ADM)

3 Schedule Network Templates

4 Dependency Determination

5 Leads and Lags



1 Precedence Diagramming Method (PDM)

Also called AON (activity on Node)

Puts effort on the node

Most common today



PDM (pmbok)

Begin EndH

LK

JI

F G

EDC

A B



Definitions (activities on node: AON)

Event is start or end of group of activities (circle)

Activity is work required to move from event to event (node time)

6/2 9/3

Complete testing

Final report



Dependency Relations

Finish-to-Start (must finish before next can start)

Finish-to-Finish (must finish before next can finish)

Start-to-Start (must start before next can start)

Start-to-Finish (must start before next can finish)

All of these assume 100% completion; could have a percentage



Network Analysis

A# Dur

ES

ACTIVITY DESCRIPTION

EF

LS LF



Zero or 1?

Note do we start at day 0 or day 1?

Most folks start at 0

1 1

A 4

5 5

B 3

0 0

A 4

4 4

B 3



Forward Pass

1 1

A 1

2 2

B 3

5 5

D 5

2 3

C 2

10

10

F 3

4 7

E 3



Earliest-Latest Dates

1 1

A 1

2 2

B 3

5 5

D 5

2 3

C 2

10 10

F 3

4 7

E 3

12 12

7 9 3 4

1 1

4 4 9 9



Backward Pas

A 1

B 3

-8

D 5

C 2

0

F 3

E 3

0 0

0

-8

-3 -3

-11 -11

-12 -12



Critical Path

1 1

A 1

2 2

B 3

2 3

C 2

10 10

F 3

4 7

E 3

12 12

7 9 3 4

1 1

2 4

5 5

D 5

5 9



Comments

Idea of Critical Path

Note FP is done to estimate finish date

BP is done when finish date is fixed and you want to know when to start

Must be the same



1

0

ID

Customer

3

3 2

3Develop Questionnaire

13

10

6

38 Print Q

48

10

4

33Rev, Finalize Q

38

5

7

38 Develop

DA SW

50

12

5

38 Mail

labels

40

2

8

38Develop Test Data

40

2

1

48 Mail Qs, get Resps

9

113

65

a

50 Test

SW

55

5

3

13 Test Q

33

20

1

113 Input Res Data

b

120

7 c

120 Analyze Data

128

8 1

128 Prepare Report

d

183

10

-5-8 2555-5

3025

105100

10098

120112112105 130120



2 Activity on Arrow (Arrow Diagramming Method)

Event is start or end of group of activities (circle)

Activity is work required to move from event to event (arrow)

6 9

Complete testing

Final report

2 weeks



Sources and Sinks

6

23

33

8

77

11

10

9

Source (burst point)

Sink



Comments

Idea of Critical Path

Can use optimistic, normal or pessimistic time estimates (Ro6?)

Can use “dummy” variables to help in sequencing

PERT for high variance

CPM for low



Mathematical Choices

Critical Path Method (CPM)

Graphical Evaluation and Review Technique (GERT)

Program Evaluation and Review Technique (PERT)



PERT/GERT/Network Analysis

Basic Definitions

How to Crash Critical Paths (later)

Estimating ranges of completion times



Graphical Evaluation and Review Technique (GERT)

Permits an iterative looping in the schedule (none of the others do)

Uses probabilistic estimates



Program Evaluation and Review Technique (PERT)

Uses a weighted average like the Rule of Six

Good for calculating best, expected and worse case scenarios



The Beta Distribution EV=(BC+4ML+WC)/6

0

0.2

0.4

0.6

0.8

1

1.2

1 4 7 10 13 16 19 22 25

Beta Distribution

BC ML WCEV



Estimating the SD PERT-wise

σ = (WC-BC)/6

If you have many, you must add up the variances not the σs. (var = σ2)



Example

a b c d2,3,4 3,5,74,7,10

σ ab = 0.33 σbc = 1.0 σcd = 0.67

σad = √ σ2ab + σ2

bc + σ2cd

σad = √ (.332+1.02+0.672) = 1.25



Definitions

DependenciesHard – must be done first

Soft – may be necessary or not (I can start high level design before all requirements are done)

External – beyond PM’s control



Dummy Activities

1

D2

53

4

A

B

C

DUMMY

4<1,2,3



Slack Time

= time between scheduled completion date and required date (to meet CP)

TE is earliest time event can take place

TL is latest time

ST = |TE – TL|



PERT with Slack Time

2 4

53

6

1

36

7

2

TE=3TL=3

5

5

3

TE=0TL=0

TE=2 TL=5

TE=10TL=10

TE=15TL=15

TE=6TL=9



Can Refine

ES = earliest start

EF = earliest finish

LS = latest start

LF = latest finish



Full PERT

C(6,11) 5(14,19)

ACTIVITY IDENTIFICATION

ACTIVITY TIME

LATEST FINISH TIME

LATEST START TIME

EARLIEST FINISH TIME

EARLIEST START TIME



PERT with Full Slack Times

B(6,15)9(6,15)

D(6,12) 6(11,17)

A(0,6)6(0,6)

H(12,16) 4(18,22)

G(12,14)6(17,19)

E(15,18) 3(15,18)

F(15,22) 7(15,22)

C(6,11) 5(14,19)

I(18,21) 3(19,22)


Unit 6 Project Time Management PERTing Along

develop schedule

$?

?

res?

plan?

feedback

BL plans & schedule

no

no

no

noResource control

Management Approval



How to Feedback?

Transfer resources from sps to cps

Eliminate activities

Add more resources

Use less time-consuming activities

Parallelize more

Shorten CP

Shorten earliest activities



Shorten latest activities

Increase number of working hours/day

Use cheaper people



Parallelizing to Shrink Critical Paths

1 2 3

1

2

3

4

4

0

16

4

16

0



Nested PERTs

EB

C

F

D

A

FA 7 8

5

43

6

7

2

54

4

8

6

9



PERT Life Cycle

1 lay out list of activities

2 order them and add arrows

3 review with line managers

4 doers add time estimates (unlimited Res)

5 PM adds calendar dates (limitations)

6 Checks reality of calendar dates



Perturbation Analysis

Always check if times change dramatically

Primary Objectives are:Best time

Least cost

Least risk



Secondary Objectives

Alternatives

Optimum schedules

Effective use of resources

Communications

Refinement of the estimating process

Ease of project control

Ease of time/cost revisions



PERT Constraints

Calendar completion

Cash flow

Limited resources

Management approvals



PERTs and CPMs

PERTs are event-orientedGood for R&D

Hard to tell percentage complete

Payouts at milestones

CPMs are activity-oriented% complete along lines can be done

Good for well-defined activities



CPM best for

Well-defined projects such as construction

One dominant organization

Relatively small risk

One geographic location



Project Software Support

Level I (Excel)Level II (MS)Level III (Artemis)Level IV (in-your-dreams)

REMEMBERSOFTWARE DOESN’T MANAGE PROJECTS:

PEOPLE DO



SW Capabilities

System capacity

Network schemes (AD/PRE)

Calendar dates

Gantt charts

Flexible report generation

Updating

Cost control

Scheduled dates

Sorting

Filtering

Resource allocation

Plotting

Machine requirements

Cost



3 Schedule Network Templates

Mature organizations will have general templates to begin the work

Especially if portions are repetitive. Such as floors on a high-rise, clinical trials, software module construction.



4 Dependency Determination (3 kinds)

MandatoryNormal (see PDM 4 types)

DiscretionarySuch as a preferred way of executing a sequence of events when there are several OK paths

ExternalOutside the PM’s control such as delivery of necessary hardware, Y2K, laws, etc



5 Leads and Lags

Lead permits the acceleration of the successor task. Eg. Can start chapter 2 15 days before chapter 1 is complete (F2S with 15 day lead)

Lags delay next task. Concrete must cure for 15 days. Therefore a F2S with a 15 day lag

Leads and Lags can be negative (but why?)



Lag Time

Suppose that B lags A by 3

1 1

A 4

8 8

B 3




1 EEF2 OPA3 Activity List4 Act Atts5 Resource Avail6 PMP

1 Xpert Judgment2 Alternative Anal3 Published E Data4 PM Software5 Bottom-up Est

1 Act Res Reqs2 Act Atts (ups)3 Res BD Structure4 Res Cals (ups)5 Req’d Changes



6.3.1 ARE Inputs

EEFsUses infrastructure resource availability as per the nature of the company

OPAsPolicies for staffingRental or purchase of supplies, equipmentHistorical information

Resource AvailabilityIn the market now? Later?



6.3.2 Tools and Techniques

1 Expert Judgment

2 Alternative Analysis

3 Published Estimating Data

4 PM Software

5 Bottom-up Estimation



1 Expert Judgment

Experts in the area can tell us who we need

Any group or person having area-specific knowledge use\ful here



2 Alternatives Analysis

Make-or-buy decisions

Who can do the work

May be necessary to outsource some of the schedule activities

Can we cannibalize other work?



3 Published Estimating Data

There are commercially available books of production rates and unit costs for trades, material, equipment, in many countries or geographical areas



4 PM Software

PM software can use RBSs, resource availability, resource calendars, to allocate for us



5 Bottom-up Estimation

If we cannot estimate the needed resources, may need more decomposition into finer detail. Continue the decomposition until we can estimate and then roll back up



6.3.3 ARE Outputs

1 Activity Resource Requirements

2 Activity Attributes (ups)

3 Resource Breakdown Structure

4 Resource Calendars (ups)

5 Required Changes




IDs types and quantities of resources required for each schedule activity.

Can then roll up for total for the work package

Will lead to the estimation numbers in the next section



3 Resource Breakdown Structure

Same as WBS only for RBS



6.4 Activity Duration Estimation

1 EEF2 OPA3 Scope Statement4 Act List5 Act Attributes6 Act Res Reqs7 Res Cals8 PMP

1 Xpert judgment2 Analogous Est3 Parametric Est4 Three-point Ests5 Reserve Anal

1 Act Duration Ests2 Act Atts (ups)



General

This is big

We take all of the preceding and roll it all up into the best time estimate that we can manage.

We need to know the size of the work and the production rates to time it out



Other things

Need to know the expected working periods

Do you count weekends?

Also what is the normal metric for effort?ph?

pd?

pm?

py?



6.4.1 Inputs to Activity Duration Est

1 EEF2 OPA3 Scope Statement4 Activity List5 Activity Attributes6 Activity Resource Requirements7 Resource Calendars8 PMP



1 EEF

Historical data important

Also when durations are not driven by the work but by things like:

Curing time of concrete

Time to get approvals through government agencies



2 OPA

Recorded data from previous work important here

Team effort records (sw for example, fp/m rates of individuals)



Historical Information

Can come from the PM morgue

In many engineering areas, there are tables

Steel girders, for example

Unions have rates

Commercial databases



3 Scope Statement

Assumptions from the Scope; reporting periods can dictate maximum schedule durations

Review periods

Document submittals etc.




Trickery trickery trickery.

Will affect the schedule

EG; need 2 engineers to do the design

If only 1 is available, may take twice (or more likely more than twice) to complete

Applying n resources will not cut the time by 1/n. In fact may increase it.



Resource Requirements

Is it the case that the work can be paralleled?

For example, two people can do the work twice as fast as one

But be careful: the fallacy of linear scaling

There comes a time when adding more people to the work only causes it to take longer



5 Resource Capabilities

People do work at different rates

Senior people should be faster than juniors

Some areas are human-specificin coding, it has been measured, that holding all other variables constant, there can be a ten to one difference in coding rates (Weinstein, 1972)



8 PMP Inputs

Risk RegistryRisks are associated with resource availability and goodness of resources

Activity Cost EstimatesCan use activity cost estimates from the PMP here



Remember the Cops and the Donuts

We need two estimates

SIZE

EFFORT

We also want to specify the confidence levels of our numbers

Rule of 6 good here



Risk

Need to estimate the costs of risk

High risk means higher costs because of the risk oversight and possible mitigation



6.3.2 T&T for Activity Duration Est

0 Introduction to Estimation

1 Expert Judgment

2 Analogous estimates

3 Parametric Estimating

4 Three-Point Estimates

5 Reserve Analysis (contingency)



0 Estimating in General

General Idea

Rules of Thumbs and SWAGs



General Principles of Estimation

General Principles

Pitfalls of Estimation

General Volumetrics



General Estimating

Estimates are just that!

Example: how long does it take you to drive to work?



Distributions

How measurements might be distributed

Plot the length of 100 meter sticks

Plot the Julian birthday of every Canadian (JBD is the day of the year tat you were born. Jan01=1 and Dec 31=365)



The Normal Distribution

0

0.2

0.4

0.6

0.8

1

1.2

1 3 5 7 9 11 13 15 17 19

Normal Curve



Normal Distribution

0

0.2

0.4

0.6

0.8

1

1.2

-9S

-7S

-5S

-3S

-1S 1S 3S 5S 7S 9S

Normal Curve

MEAN

6Σ 3.4 in 106

5Σ 1.0 in 105

4Σ 1.0 in 104

3Σ 1.0 in 103

2Σ 1.0 in 102

1Σ 1.0 in 101



Six-Sigma

One and two tailed estimates

Ours are normally one-tailed

2 is 99%

3 is 99.9%

4 is 99.99%

5 is 99.999%

6 is 3 part in a million (99.9999)



Question?

You have a 2400 square foot house and you order your cleaner to clean it to within 6.

What is the size of the largest piece of dirt?A thimble?

A teacup?

A saucer?

A bathroom?



General Estimating cont.

normally, use the average

the (1+4+1) / 6 is good

be realistic (factor in time of year)



Rules of Thumbs

My Uncle's example

The Rule of 3

The Back of the Envelope



The Rule of 33 people in my house30 close neighbours300 on my jogging route3000 in my school draw30000 in my ward300,000 in London3,000,000 in Ontario-Toronto30,000,000 in Canada300,000,000 in NA (- Mexico)3,000,000,000 "consumers" in the world



Rules of Thumbs (jon bentley)

How much water flows out of the Mississippi River in one day (cu miles)?



Rule of 72

Exponential are difficult

Most of our problems ARE expos

If you invest a sum that must double in y years at an interest rate of r percent/yr then r*y = 72 holds. (RULE OF 72)

Example, how long will it take for $1,000 to double at 6%? 72/6=12 years ($2012)



Example

A program takes 10 seconds for size n=40

Increasing n by 1 increases time by 12% (expo)

Rule-of-72 says RT doubles when n increases by 6

By 60, then 1,000

By 160, 107 seconds



Help Ma!

How BIG is 107 anyway?

Actually dear, 3.155x107 seconds in a year

Or seconds in a nanocentury

264 = 100,000,000 donuts/sec for 5,000 years



The Delphi Approach

No clear way to estimate

Gather a group of Xperts

Give them the problem; they go away and independently estimate as well as they can

They meet and exchange information

Then they repeat the above

After 3-4 cycles they will normally converge on a unified answer



A Little Quiz (thanks Jon)(give 1+4+1 confidence limits)

1. Canadian population Jan 1,2004

2. Year of Napolean's birth

3. Length of the Great Lakes/St Lawrence watershed

4. Maximum takeoff weight of a 747 (pds)

5. Mass of the earth

6. Number of Fathers of Confederation



7. Latitude of London England

8. Number of airplanes in the air at this minute

9. Number of PCs in Canada

10.Number of bones in the adult human



General Estimating cont.

make sure you have a complete SOW

work out the WBS completely

hand off to the person responsible to estimate and cost

collect them in the PP

note that you do this at EACH of the three levels of report generation



Things to Avoid

warm fuzzies

too-new technologies

biggies

too-optimistic estimates

LINEARITY



Linear Scaling

1 person can do the work in 8 days

2 can do it in 4

4 can do it in 2

8 can do it in 1

16 can do it in ½ a day

32 in a ¼

64 in an hour etc etc



Examples

How many kilometers per year does a taxi driver drive if he works an 8 hours day 200 days a year?



Conclusions, Crystal Balling

It Works!

can easily tailor the tool to the organization’s process and culture

can instrument to collect metrics

can do the EV easily

can prompt the user for missing steps

can archive for the Morgue

can collect quality metrics



1 Expert Judgment

Remember my definition of xpert



2 Analogous Estimates

Compares against work already done

Is really a form of expert judgment

Are most reliable whenPreviously done activities are very similar

Experts really know the area



3 Parametric Estimating

When we know the rates

For example, function points and the Industrial Averages



4 Three-Point Estimates

Most likely

Optimistic

Pessimistic

Useful for worst-case, best-case scenarios

Rule-of-Six gives better estimates



4 Reserve Analysis (contingency)

Buffer in case of risky activities

Need to annotate the reasons for asking for one

Known-unknowns (contingency)

Unknown-unknowns (management reserves)



6.4.3 Outputs for Activity Duration Est

1 Activity duration estimates

2 Bases of estimates

3 Activity lists updates



1 Activity Duration Estimates

Need also to list the confidence levels of the estimates

Prob or SD good here




1 OPA2 Scope Statement3 Act List4 Act Atts5 Sched N/W Diags6 Act Res Reqs7 Res Cals8 Act Dur Ests9 Risk Register

1 Sked N/W Anal2 CPM3 Sked Compress4 What-if Anal5 Res Leveling6 Critical Chain 7 PM Software8 Applying Cals9 Adjusting L&Ls10 Schedule Model

1 Project Sched2 Sched Model Data3 Sched Baseline4 Res Reqs (ups)5 Act Atts (ups)6 Proj Cals (ups)7 Req’d Changes8 PMP (ups)



Inputs from OPAs

Project calendar may dictate days when no work can be done.

Shifts may be constrained



Input from Scope Statement

PSS can contain assumptions and constraints that affect schedule development; two main types1 Internally imposed dates

Agreed-upon contract datesWeather restrictionsGovernmental mandated compliance dates“Start no earlier than” and “Finish no later than” most commonly used

2 Externally imposed datesStakeholders can dictate important datesMilestones hat connect to external projects



7 Calendars

These show when resources and the project are available for work assignment

Resources have vacations, religious holidays etc

A labour contract may limit the days of the week a person can work



6.4.2 T&T for Schedule Development

1 Schedule Network Analysis2 Critical Path Method3 Schedule Compression4 What-if Analysis5 Resource Leveling6 Critical Chain Method7 PM Software8 Applying Calendars9 Adjusting Leads & Lags10 Schedule Model



1 Schedule Network Analysis

This generates the project schedule

Uses the following techniques

Checks for loop or open ends



2 Critical Path Method

Uses the earliest start and finish dates and the late start and finish dates without any regard to resource limitations

Worries about float

Does a forward pass and a backward pass

Uses a single estimate for each activity

Leads to a Critical Path and a deterministic schedule



Types of Float

Free Float – time a task can be delayed without delaying the early start date of its successorTotal Float - time a task can be delayed without delaying the project completion dateProject Float - time the project can be delayed without delaying the externally imposed project completion date (by customer, management, project manager etc.)



Passes

To compute the likely finish time plus critical path(s)Forward pass

Start at the beginning

Backward passStart at the customer’s wanted finish date and work backwards

Can have negative float!



Two ways to compress without descopingCrashing

Fast-tracking

3 Schedule Compression



Crashing

Try to compute the CP

Work out the cost per week to crash

Start with lowest



Crashing with CPM

A.4.

F.6

E.7C.2

D.2B.6



Crash Data

Normal Crash Norm$ Crash$ CC/wk

A 4 2 10K 14K 2K

B 6 5 30K 42.5K 12.5K

C 2 1 6K 9.5K 3.5K

D 2 1 12K 18K 6K

E 7 5 40K 52K 6K

F 6 3 20K 29K 3K



Crash details

Normal time

Crash timeNote; follows U-curve

CT is most compressed time

Compute (CC-NC)/(NT-CT)



Crashing

Try to compute the CP

Work out the cost per week to crash

Start with lowest



Crashing Problems

May not be possible

Will INCREASE costs for sure

Assumes that you can take people off one task and add them to another (true in construction for example; may well NOT be true in IT!)



Fast Tracking

Do CP tasks that were planned in series, in parallel

Problems:Often forces rework

Increases risk

Requires more communications

May cost more (need new people)



4 What-if Scenario Analysis

As seen

Often uses Monte Carlo techniques to check out worst-case, best-case, random-case examples



5 Resource Leveling

Critical path may over-allocate resources

Necessary to “level” them

An option in MSP



Resource Loading and Leveling

Resource loading: amount of individual resources an existing project schedule requires during specific time periodsResource histograms show resource loadingOver-allocation means more resources than are available are assigned to perform work at a given time



Resource Leveling

Resource leveling is technique for resolving resource conflicts by delaying tasks

Primary purpose of resource leveling: create a smoother distribution of resource usage & reduce over-allocation



Resource Histogram for Large IT Project



Histogram Showing an Over allocated Individual


Unit 6 Project Time ManagementResource Leveling Example



6 Critical Chain

A better way

Each activity has a mean of execution time, not a constant

CC says, start as soon as you finish

Suppose A B, A = 4±2, B=6±3

CMP says B starts on Day 5, regardless

CC says, start on Day 3 if lucky



CC comments

Idea is to take advantage of early finishes

What tends to happen in Anal Orgs is that the start date of each task is fixed

When CP task slips, whole project time slips

When it is early, people go fishing until the specified start date of the next task

CC starts ASAP and averages out under runs and overruns



7 PM Software

Can be used to do the preceding techniques

Can print out nice diagrams of the schedules

Can compress/expand as needed



8 Applying Calendars

Used to cover weather events etc. as seen

Can be used for scenarios such asOnly regular hours allocated

Two shifts

7x24 shifts



9 Adjusting Leads & Lags

Can be modified in the simulation to see how sensitive the schedule is to adjusting leads and lags



10 Schedule Model

This is the final sign-offed product that will be used for the duration of the project

Very important for ICC to check the effects of changes in scope, requirements



Coding Structure

Activities should have a code (database?) so that you can sort/extract on different attributes of the activities such as

ResponsibilityGeographic areaBuildingProject phaseSchedule levelWBS classification



6.5.3 Outputs from Schedule Development

1 Project Schedule2 Schedule Model Data3 Schedule Baseline4 Resources Requirements (ups)5 Activity Attributes (ups)6 Project Calendars (ups)7 Required Changes8 PMP (ups)



1 Project Schedule

Includes at least a start and end date for every schedule activity

Normally presented graphicallyNetwork Diagram

Bar (Gannt) charts

Milestone charts



2 Schedule Model Data

Supporting data for all activity attributes, all schedule activities, all assumptions and constraints

Also may includeResource requirements by time period

Alternative schedules (best, worse)

Schedule contingency reserves



3 Schedule Baseline

This is the approved-by-management schedule for tracking purposes




This must do 3 things1. Ensure that changes are agreed on

2. Determine that the schedule has changed

3. Managing the changes when they occur

Is another example of change control and if it is integrated properly, can be rolled up into it



Schedule Control

1 Sched Man Plan2 Sched BL3 Perf Reps4 App’d Change Reqs

1 Prog Reporting2 Sched CC System3 Perf Meas’t5 PM Software6 Variance Anal7 Sched Bar Charts

1 Sked model (ups)2 Sked BL (ups)3 Perf Meas’s4 Req’d Changes5 Rec’d Corrects6 OPA (ups)7 Act List (ups)8 PMP (ups)9 Act Atts (ups)



6.5.1 Inputs to Schedule Control

1 Schedule Management Plan

2 Schedule Baseline

3 Performance Reports

4 Approved Change Requests



3 Performance Reports

These flow out of communications

Indicate when we are falling behind and signal the need for change



4 Approved Change Requests

Once an item has been baselined, it is put under CCM

Should be a form which is filled out and this, when approved by the CCB, is put into the Schedule Control process



6.5.2 T&T for Schedule Control

1 Progress Reporting

2 Schedule CC System

3 Performance Measurement

5 PM Software

6 Variance Analysis

7 Schedule Bar Charts



1 Progress Reporting

Records actual start and finish dates (as opposed to planned)

Has all EV measurements

Should be in an org-wide template form



2 Schedule Change Control System

Formal procedure by which we do the changes

See unit 4.

Is a very important case of CCM

Is isolated here to stress its importance

Necessary approvals here important



3 Performance Management

Need to understand the metrics of PM and assess if change is needed immediately or can it wait?

If the activity is on the CP, do it now

If off the CP, could wait a bit



4 PM Software

Lets us know when corrective action is necessary

Could be a push technology (here be dragons!)



5 Variance Analysis

Critical for the EV portion of time

Float is key here

Need to sort sub-critical paths in terms of increasing float



6 Schedule Comparison Bar Charts

Also called “double Gantting”

Two bars; one the actuals, one the planned. Is another way to track progress. Note the fallacy compared with EVM



6.5.3 Outputs from Schedule Control

1 Schedule Model (ups)2 Schedule Baseline (ups)3 Performance Measurements4 Required Changes5 Required Corrects6 OPA (ups)7 Activity List (ups)8 PMP (ups)9 Activity Attributes (ups)



1 Schedule Model Updates

Any modification to the schedule

Must notify stakeholders

May trigger updates to other parts of the PMP



Schedule Baseline Revisions

These are changes to the project’s start and finish date

Are major

May require rebaselining (a Baaaad thing)

Rebaselining is a Last Resort



5 Corrective Action

Anything done to bring future performance in line with the planned estimates

Often involves expediting

Need to do a root cause analysis to avoid future deviations



6 OPA Updates (Lessons Learned)

REALLY important

“Those who ignore the failure lessons of history are doomed to repeat them” G. Santayana



Chapter Six: Time Management

2000 Edition






Third Edition









Case Study: Panama Canal (french)

Experts recommendedA sea-level canal (like Suez)

Reducing TL from 12 to 8 years

Reducing cost from $240M to $169M

Reducing the contingency from 25% to 10% (ignored cost of capital, cost of purchasing Panamanian railroad, administrative costs)

Estimators doubled the anticipated excavation volume by 100% (while doing the cost reduction)



De Lesseps goes to Panama

Spent 1 week there

Reduced the cost estimate to $132M

Discounted the deadly climate as “an invention of adversaries”



The Result:

20,000 Frenchmen died

Spent $287M for little work

Estimates were based on what they could SELL to the investors, not the actual cost

Philosophy: “get her started” and we’ll figure out something later

Big Dig!



WBS Structures the Project NetworkProject

ProjectProject

ProjectProject

Sub A11

Sub A11

Sub A11

Sub A11

Sub A11

Sub A11

Sub A11

Sub A11

Sub A11

Sub A11

Sub A11

Sub A11

Sub A11

SE 313A Operating Systems Unit 1

Technology

Transcript of SE 313A Operating Systems Unit 1