Administración de Proyectos

Page 1


Importance of Project Management • Projects represent change and allow organizations to effectively introduce new products, new process, new programs • Project management offers a means for dealing with dramatically reduced product cycle times • Projects are becoming globalized making them more difficult to manage without a formal methodology • Project management helps cross-functional teams to be more effective


Management of IT Projects • More than $250 billion is spent in the US each year on approximately 175,000 information technology projects. • Only 26 percent of these projects are completed on time and within budget. • The average cost for a development project for a large company is more than $2 million. • Project management is an $850 million industry and is expected to grow by as much as 20 percent per year. Bounds, Gene. “The Last Word on Project Management” IIE Solutions, November, 1998.


What Defines a Project? • • • • • • •

How does a project differ from a program?


Project Management versus Process Management “Ultimately, the parallels between process and project management give way to a fundamental difference: process management seeks to eliminate variability whereas project management must accept variability because each project is unique.” Elton, J. & J. Roe. “Bringing Discipline to Project Management” Harvard Business Review, March-April,

1998.


Measures of Project Success • • • • • • •

Was the movie “Titanic” a success?


Delayed Openings are a Fact of Life in the Foodservice, Hospitality Industry Disney's shipbuilder was six months late in delivering its new cruise ships, and thousands of customers who had purchased tickets were stranded. Even with that experience, their second ship was also delivered well after the published schedules. Universal Studios in Orlando, Fla. had been building a new restaurant and entertainment complex for more than two years. They advertised a December opening, only to announce in late November that it would be two or three months late. Even when facilities do open close to schedule, they are rarely finished completely and are often missing key components. Why do those things happen? With all of the sophisticated computers and project management software, why aren't projects completed on schedule? Frable, F. Nation's Restaurant News (April 12, 1999)


IT Project Outcomes More than 200% late

101-200% late

6% 16%

51-100% late

21-50% late

9%

29%

Cancelled

8% 6%

Less than 20% late

26%

On-Time

Source: Standish Group Survey, 1999 (from a survey of 800 business systems projects)


Why do Projects Fail? Studies have shown that the following factors contribute significantly to project failure: • Improper focus of the project management system • Fixation on first estimates • Wrong level of detail • Lack of understanding about project management tools; too much reliance on project management software • Too many people • Poor communication • Rewarding the wrong actions


Why do IT Projects Fail? • Ill-defined or changing requirements • Poor project planning/management • Uncontrolled quality problems • Unrealistic expectations/inaccurate estimates • Naive adoption of new technology Source: S. McConnell, Construx Software Builders, Inc.


Have You Ever Lost Sight of the Project Goals?

QuickTim e™ and a Photo - JPEG decompres sor are needed to s ee this picture.


Not all Projects Are Alike… “[in IT projects], if you ask people what’s done and what remains to be done there is nothing to see. In an IT project, you go from zero to 100 percent in the last second--unlike building a brick wall where you can see when you’re halfway done. We’ve moved from physical to non-physical deliverables….” J. Vowler (March, 2001)

Engineering projects = task-centric IT projects = resource-centric


Shenhar’s Taxonomy of Project Types Degree of Uncertainty/Risk Super HighTech

ERP implementation in multi-national firm

HighTech

New shrinkwrapped software

MediumTech LowTech

Advanced radar system

New cellphone Construction Assembly Projects

Auto repair System Projects

Array Projects

System Complexity/Scope

High


Required Resources

Project Life Cycle

Phase 1

Formation & Selection

Phase 2 Planning

Phase 3

Scheduling & Control

Phase 4

Evaluation & Termination

Time


Life Cycle Models: Pure Waterfall Concept Design Requirements Analysis Architecture Design Detailed Design Coding & Debugging System Testing

Source: S. McConnell Rapid Development (Microsoft Press, 1996)


Life Cycle Models: Code & Fix


DESIGN

Design, Cost, Time Trade-offs Required Performance

Target

M TI

E

LE U D E CH S (

Due Date

)

Budget Constraint

Optimal Time-Cost Trade-off

COST


Optional Scope Contracts Since it is widely accepted that you can select three of the four dimensions (or perhaps only two), what to do? Fixed Scope Contract

specifies

Optional Scope Contract

specifies

SCHEDULE, COST, SCOPE

SCHEDULE, COST, QUALITY

(general design guidelines may be indicated)


Importance of Project Selection “There are two ways for a business to succeed at new products: doing projects right, and doing the right projects.” Cooper, R.G., S. Edgett, & E. Kleinschmidt. Research • Technology Management, March-April, 2000.


Project Initiation & Selection • Critical factors 1) Competitive necessity 2) Market expansion 3) Operating requirement

• Numerical Methods 1) 2) 3) 4)

Payback period Net present value (NPV) or Discounted Cash Flow (DCF) Internal rate of return (IRR) Expected commercial value (ECV)

• Project Portfolio 1) Diversify portfolio to minimize risk 2) Cash flow considerations 3) Resource constraints


Payback Period Number of years needed for project to repay its initial fixed investment

Example: Project costs $100,000 and is expected to save company $20,000 per year Payback Period = $100,000 / $20,000 = 5 years


Net Present Value (NPV) Discounted Cash Flow (DCF) Let Ft = net cash flow in period t (t = 0, 1,..., T) F0 = initial cash investment in time t = 0

r = discount rate of return (hurdle rate)

T

NPV = t=0

Ft 1+rt


Internal Rate of Return (IRR) Find value of r such that NPV is equal to 0

Example (with T = 2): Find r such that F0 + F1 + F2 2 = 0 1+r 1+r


DCF Project Example* Phase I

Research and Product Development $18 million annual research cost for 2 years 60% probability of success

Phase II

Market Development Undertaken only if product development is successful $10 million annual expenditure for 2 years to develop marketing and distribution channels (net of any revenues earned in test marketing)

Phase III

Sales Proceeds only if Phase I and II verify opportunity. Production is subcontracted and all cash flows are after-tax and occur at year's end. The results of Phase II (available at the end of year 4) identify the product's market potential as indicated below:

Product Demand High M edium Low

Product Life

Annual Net Cash Inflow

Probability

20 years 10 years Abandon Project

$24 million $12 million None

0.3 0.5 0.2

*Hodder, J. and H.E. Riggs. “Pitfalls in Evaluating Risky Projects�, Harvard Business Review, Jan-Feb, 1985, pp. 128-136.


DCF Project Example (cont’d) Year 1 2 3 4 5 - 14 15 - 24

Expected Cash Flow (in $ million) -18 -18 0.6 (-10) = - 6 0.6 (-10) = - 6 .6 (0.3 x 24 + 0.5 x 12) = 7.92 .6 (0.3 x 24) = 4.32

What is the internal rate of return for this project?


DCF Example Continued What if you can sell the product (assuming that both Research and Product Development AND Market Development are successful) to a third party? What are the risks AT THAT POINT IN TIME?

Assume that discount rate r2 is 5% Probability What is 20 years of cash inflow at $24M/year? What is 10 years of cash inflow at $12M/year? Expected value of product at Year 4:

$299.09 $92.66 $136.06

0.3 0.5


DCF Example Continued Expected cash flows (with sale of product at end of year 4) are now:

Year 1 Year 2 Year 3 Year 4

Outflow $ 18.00 $ 18.00 $ 10.00 $ 10.00

Inflow

$

136.06

$ $ $ $

Net (18.00) (18.00) (10.00) 126.06

Probability 1 1 0.6 0.6

Expected Cash Flow $ (18.00) $ (18.00) $ (6.00) $ 75.63

What is the internal rate of return for this project?


Criticisms of NPV/DCF 1) Assumes that cash flow forecasts are accurate; ignores the “human bias� effect 2) Fails to include effects of inflation in long term projects 3) Ignores interaction with other proposed and ongoing projects (minimize risk through diversification) 4) Use of a single discount rate for the entire project (risk is typically reduced as the project evolves)


Expected Commercial Value (ECV) Probability = pc Commercial Success (with net benefit = NPV) Probability = pt Technical Success

Launch New Product

Probability = 1 - pc Commercial Failure (with net benefit = 0)

Develop New Product Probability = 1 - pt Technical Failure

Risk class 1

Risk class 2


DCF Example Revisited Product Demand High

0.3 Probability = pt Development Succeeds Research & Product Development

0.5 Market Development

0.2 Probability = 1 - pt Development Fails

Discount rate r1

Product Demand Medium

Product Demand Low

Drop project

Discount rate r2


Ranking/Scoring Models Profit abilit y/value 1) Increase in profitability? 2) Increase in market share? 3) Will add knowledge to organization that can be leveraged by other projects? 4) Estimated NPV, ECV, etc. Organizat ion's Strat egy 1) Consistent with organization's mission statement? 2) Impact on customers? Risk 1) Probability of research being successful? 2) Probability of development being successful? 3) Probability of process success? 4) Probability of commercial success? 5) Overall risk of project 6) Adequate market demand? 7) Competitors in market Organizat ion Costs 1) Is new facilit y needed? 2) Can use current personnel? 3) External consultants needed? 4) New hires needed? Miscellaneous Factors 1) Impact on environmental standards? 2) Impact on workforce safety? 3) Impact on quality? 4) Social/political implications


Scoring Attributes To convert various measurement scales to a (0, 1) range‌. xi - L LINEAR SCALE: value of attribute i is vi xi = U - L 1 - exp L - xi v x = . EXPONENTIAL SCALE: value of attribute i is i i 1 - exp L - U 1.00 0.90 0.80

Attribute Value

0.70 0.60 Linear Scale Exponential Scale

0.50 0.40 0.30 0.20 0.10 0.00 1

2

3

4 Response

5

6

7


Ranking/Scoring Example

Attribute 1) Does project increase market share?

Attribute Weight (wi)

Measurement Scale unlikely

1

2

2) Is new facility needed?

3 yes

3) Are there safety concerns?

likely

4

5

likely

15% 10%

no unsure

30%

no

4) Likelihood of successful technical development?

unlikely

1

2

3

4

5

likely

20%

5) Likelihood of successful commercial development?

unlikely

1

2

3

4

5

likely

25%


Ranking/Scoring Example (cont’d)

Attribute Project A Project B

Project Score (V j)

#1

#2

#3

#4

#5

4 2

yes no

likely unsure

4 3

1 4

0.75 0.25

0.25 0.75

0 0.5

0.75 0.5

0 0.75

0.413 0.525

0.97 0.64

0.64 0.97

0.00 0.88

0.97 0.88

0.00 0.97

0.581 0.845

Linear Scale Project A Project B

Exponential Scale Project A Project B


Analyzing Project Portfolios: Bubble Diagram Prob of Commercial Success High

Zero

High

Expected NPV

Low


Extent of Product Change

Analyzing Project Portfolios: Product vs Process

Extent of Process Change Source: Clark and Wheelwright, 1992


Key Elements of Project Portfolio Selection Problem 1. Multi-period investment problem 2. Top management typically allocates funds to different product lines (e.g., compact cars, high-end sedans) 3. Product lines sell in separate (but not necessarily independent) market segments 4. Product line allocations are changed frequently 5. Conditions in each market segment are uncertain from period to period due to competition and changing customer preferences


“Stage-Gate� Approach Initiation

Define

Design

Initiation Project Review Charter

Work Statement Risk Assessment Purchasing Plan Change Mgt

Detail Design Schedule & Budget Contingency Plan Product & Performance Reviews

Improve

Installation Plan Facility Prep Training Plan Implementation

Control

Production close-out Lessons learned Post-project audit

Source: PACCAR Information Technology Division Renton, WA


Project Selection Example

1 Project A Project B Budget Limit (B t )

Y e a r (t) 2

3

4

($40)

$10

$20

$20

($65)

($25)

$50

$50

$120

$20

$40

$55


Phases of Project Management n n

Project formulation and selection Project planning u u u u u

n

Project scheduling u u u u

n

Summary statement Work breakdown structure Organization plan risk management Subcontracting and bidding process Time and schedule Project budget Resource allocation Equipment and material purchases

Monitoring and control u u u

Cost control metrics Change orders Milestone reports


Project Planning n

Summary Statement u u u u u

n

Executive summary: mission and goals, constraints Description and specifications of deliverables Quality standards used (e.g., ISO) Role of main contractor and subcontractors Composition and responsibilities of project team

Organization Plan u u u u u u

Managerial responsibilities assigned; signature authority Cross impact matrix (who works on what) Relationship with functional departments Project administration Role of consultants Communication procedures with organization, client, etc.


Importance of Project Planning The 6P Rule of Project Management: Prior Planning Prevents Poor Project Performance “If you fail to plan, you will plan to fail” Anonymous


Work Breakdown Structure (WBS) 1) Specify the end-item “deliverables” 2) Subdivide the work, reducing the dollars and complexity with each additional subdivision 3) Stop dividing when the tasks are manageable “work packages” based on the following: • Skill group(s) involved • Managerial responsibility • Length of time • Value of task


Work Packages/Task Definition The work packages (tasks or activities) that are defined by the WBS must be: • Manageable • Independent • Integratable • Measurable


Design of a WBS “The usual mistake PMs make is to lay out too many tasks; subdividing the major achievements into smaller and smaller subtasks until the work breakdown structure (WBS) is a ‘to do’ list of one-hour chores. It’s easy to get caught up in the idea that a project plan should detail everything everybody is going to do on the project. This springs from the screwy logic that a project manager’s job is to walk around with a checklist of 17,432 items and tick each item off as people complete them….” The Hampton Group (1996)


Two-Level WBS

WBS level 1

WBS level 2

1.1 Event Planning

1. Charity Auction

1.2 Item Procurement

1.3 Marketing

1.4. Corporate Sponsorships


Three-Level WBS 1. Charity Auction

WBS level 1

WBS level 2

1.1 Event Planning

1.1.1 Hire Auctioneer

1.2 Item Procurement

1.3 Marketing

1.4 Corporate Sponsorships

1.2.1 Silent auction items

1.3.1 Individual ticket sales

1.2.2 Live auction items

1.3.2 Advertising

1.1.2. Rent space

WBS level 3 1.1.3 Arrange for decorations

1.2.3 Raffle items 1.1.4 Print catalog


Estimating Task Durations (cont’d) • Benchmarking • Modular approach • Parametric techniques • Learning effects


Beta Distribution Probability density function Completion time of task j

Time Optimistic Timetoj Expected duration = Âľ

M ost Likely Time = tm

Pessimistic Time tpj


Beta Distribution For each task j, we must make three estimates: toj most optimistic time tpj most pessimistic time tm j most likely time toj + tpj + 4tm j Expected duration Âľj = 6

Variance of task j =

Ďƒ2j

tpj - toj 2 = 36


Estimating Task Durations: Painting a Room Task: Paint 4 rooms, each is approximately 10’ x 20’. Use flat paint on walls, semi-gloss paint on trim and woodwork. Each room has two doors and four windows. You must apply masking tape before painting woodwork around the doors and windows. Preparation consists of washing all walls and woodwork (some sanding and other prep work will be needed). Only one coat of paint is necessary to cover existing paint. All supplies will be provided at the start of the task. Previous times on similar painting jobs are indicated in the table below. hours 27 38 33 17 26 22 14 30 28 21 23 27 23 37 17 17

min 25 25 12 44 7 1 2 27 30 13 59 44 15 6 54 13

hours 31 19 26 30 25 24 32 32 13 42 22 32 32 27 26 21

min 52 15 27 27 21 28 58 1 43 45 57 15 31 15 11 52

What is your estimate of the average time you will need? What is your estimate of the variance?


Estimating Task Durations with Incentives Task: Consider the painting job that you have just estimated. Now, however, there are explicit incentives for meeting your estimated times. If you finish painting the room before your specified time, you will receive a $10 bonus payment. HOWEVER, if you finish the painting job after your specified time, you will be fined $1000. Revised estimated time =


Estimating Task Durations with Incentives Task: Consider the painting job that you have just estimated. Now, however, there are explicit incentives for meeting your estimated times. If you finish painting the room before your specified time, you will receive a $10 bonus payment. If you finish the painting job after your specified time, there is no penalty. Revised estimated time =


Role of Project Manager/Team Client

Top Management Project Manager

Subcontractors Project Team Regulating Organizations

Functional Managers


Responsibilities of a Project Manager To the organization and top management • Meet budget and resource constraints • Engage functional managers

To the project team • Provide timely and accurate feedback • Keep focus on project goals • Manage personnel changes

To the client • Communicate in timely and accurate manner • Provide information and control on changes/modifications • Maintain quality standards

To the subcontractors • Provide information on overall project status


Project Team What is a project team? A group of people committed to achieve a common set of goals for which they hold themselves mutually accountable

Characteristics of a project team • • • •

Diverse backgrounds/skills Able to work together effectively/develop synergy Usually small number of people Have sense of accountability as a unit


“I design user interfaces to please an audience of one. I write them for me. If I’m happy, I know some cool people will like it. Designing user interfaces by committee does not work very well; they need to be coherent. As for schedule, I’m not interested in schedules; did anyone care when War and Peace came out?” Developer, Microsoft Corporation As reported by MacCormack and Herman, HBR Case 9-600-097: Microsoft Office 2000


Intra-team Communication M = Number of project team members L = Number of links between pairs of team members If M =2, then L = 1

If M =3, then L = 3


Number of Intra-team Links 450 400

300 250 200 150 100 50

N = No. of Team Members

N(N-1) 2

29

27

25

21

23

19

17

=

15

L = Number of Intra-team Links = N 2

11

13

9

7

5

3

0 1

L = No. of Intra-team Links

350


Importance of Communication On the occasion of a migration from the east, men discovered a plain in the land of Shinar, and … said to one another, “Come, let us build ourselves a city with a tower whose top shall reach the heavens….” The Lord said, …“Come, let us go down, and there make such a babble of their language that they will not understand one another’s speech.” Thus, the Lord dispersed them from there all over the earth, so that they had to stop building the city. Genesis 11: 1-8


Project Performance and Group Harmony What is the relationship between the design of multidisciplinary project teams and project success? Two schools of thought: 1) “Humanistic school” -- groups that have positive characteristics will perform well 2) “Task oriented” school -- positive group characteristics detract from group performance


Project Performance and Group Harmony (cont’d) Experiment conducted using MBA students at UW and Seattle U using computer based simulation of pre-operational testing phase of nuclear power plant* Total of 14 project teams (2 - 4 person project teams) with a total of 44 team members; compared high performance (low cost) teams vs low performance (high cost) teams Measured:

Group Harmony Group Decision Making Effectiveness Extent of Individual’s Contributions to Group Individual Attributes *Brown, K., T.D. Klastorin, & J. Valluzzi. “Project Management Performance: A Comparison of Team Characteristics”, IEEE Transactions on Engineering Management, Vol 37, No. 2 (May, 1990), pp. 117-125.


Group Harmony: High vs Low Performing Groups 6.00 5.80 5.60 Group Harmony

5.40 5.20 5.00 4.80 4.60 4.40 4.20 4.00 1

2

3

4

5

6

Week High Performance (low cost) Teams

Low Performance (high c ost) Teams

7


Extent of Individual Contribution: High vs Low Performing Groups 6.00

Extent of Indiv idual Contributions

5.80 5.60 5.40 5.20 5.00 4.80 4.60 4.40 4.20 4.00 1

2

3

4

5

6

Week High Performance (low cost) Teams

Low Performance (high c ost) Teams

7


Decision Making Effectiveness: High vs Low Performing Groups

Decision Making Effectiv eness

6.00 5.50 5.00 4.50 4.00 3.50 3.00 1

2

3

4

5

6

Week High Performance (low cost) Teams

Low Performance (high c ost) Teams

7


Project Organization Types • Functional: Project is divided and assigned to appropriate functional entities with the coordination of the project being carried out by functional and high-level managers • Functional matrix: Person is designated to oversee the project across different functional areas • Balanced matrix: Person is assigned to oversee the project and interacts on equal basis with functional managers • Project matrix: A manager is assigned to oversee the project and is responsible for the completion of the project • Project team: A manager is put in charge of a core group of personnel from several functional areas who are assigned to the project on a full-time basis


Project Organization Continuum

Functional Matrix

Functional Organization

Project fully managed by functional managers

Project Matrix

Balanced Matrix

Project Team Organization

Project fully managed by project team manager


A Business School as a Matrix Organization Dean Associate Dean for Undergraduate Program

Associate Dean for MBA Programs

Director of Doctoral Program

Accounting Department Chair

Larry

Zelda

Diane

Marketing Department Chair

Curly

Bob

Barby

Finance Department Chair

Moe

Gloria

Leslie


Matrix Organizations & Project Success • Matrix

organizations emerged in 1960’s as an alternative to traditional means of project teams

• Became • Still

popular in 1970’s and early 1980’s

in use but have evolved into many different forms

Basic question: Does organizational structure impact probability of project success?


Organizational Structure & Project Success • Studies by Larson and Gobeli (1988, 1989) • Sent questionnaires to 855 randomly selected PMI members • Asked about organizational structure (which one best describes the primary structure used to complete the project) • Perceptual measures of project success: successful, marginal, unsuccessful with respect to : 1) Meeting schedule 2) Controlling cost 3) Technical performance 4) Overall performance • Respondents were asked to indicate the extent to which they agreed with each of the following statements: 1) Project objectives were clearly defined 2) Project was complex 3) Project required no new technologies 4) Project had high priority within organization


Study Data • Classification of 547 respondents (64% response rate) 30% project managers or directors of project mgt programs 16% top management (president, vice president, etc.) 26% managers in functional areas (e.g., marketing) 18% specialists working on projects • Industries included in studies 14% pharmaceutical products 10% aerospace 10% computer and data processing products others: telecommunications, medical instruments, glass products, software development, petrochemical products, houseware goods • Organizational structures: 13% (71): Functional organizations 26% (142): Functional matrix 16.5% (90): Balanced matrix 28.5% (156): Project matrix 16% (87): Project team


ANOVA Results by Organizational Structure N

Controlling Cost Ave (SD)

Meeting Schedule Ave (SD)

Technical Performance Ave (SD)

Overall Results Ave (SD)

A

Functional Organization

71

1.76 (.83)

1.77 (.83)

2.30 (.77)

1.96 (.84)

B

Functional Matrix

142

1.91 (.77)

2.00 (.85)

2.37 (.73)

2.21 (.75)

C

Balanced Matrix

90

2.39 (.73)

2.15 (.82)

2.64 (.61)

2.52 (.61)

D

Project Matrix

156

2.64 (.76)

2.30 (.79)

2.67 (.57)

2.54 (.66)

E

Project Team

87

2.22 (.82)

2.32 (.80)

2.64 (.61)

2.52 (.70)

Total Sample

546

2.12 (.79)

2.14 (.83)

2.53 (.66)

2.38 (.70)

10.38*

6.94*

7.42*

11.45*

Organizational Structure

F-statistic Scheffe Results *Statistically significant at a p<0.01 level

A,B < C,D,E E<D A,B < C < D,E A,B < C,D,E A,B < C,D,E


Summary of Results • Project structure significantly related to project success • New development projects that used traditional functional organization had lowest level of success in controlling cost, meeting schedule, achieving technical performance, and overall results • Projects using either a functional organization or a functional matrix had a significantly lower success rate than the other three structures • Projects using either a project matrix or a project team were more successful in meeting their schedules than the balanced matrix • Project matrix was better able to control costs than project team • Overall, the most successful projects used a balanced matrix, project team, or--especially--project matrix


Subcontracting = Business Alliance n

When you subcontract part (or all) of a project, you are forming a business alliance....

Intelligent Business Alliances: “A business relationship for mutual benefit between two or more parties with compatible or complementary business interests and/or goals� Larraine Segil, Lared Presentations


Communication and Subcontractors What types of communication mechanism(s) will be used between company and subcontractor(s)?

WHAT a company communicates.....

HOW a company communicates.....

How is knowledge transferred?


Personality Compatibility Subcontractor Personality

Corporate Personality

Project

Individual Personality


Subcontracting Issues • What part of project will be subcontracted? n• What type of bidding process will be used? What type of contract? n• Should you use a separate RFB (Request for Bids) for each task or use one RFB for all tasks? n• What is the impact on expected duration of project? n• Use a pre-qualification list? n• Incentives? Bonus for finishing early? Penalties for finishing after stated due date? • What is impact of risk on expected project cost? n


Basic Contract Types n

Fixed Price Contract u

n

Cost Plus Contract u

n

Client pays a fixed price to the contractor irrespective of actual audited cost of project

Client reimburses contractor for all audited costs of project (labor, plant, & materials) plus additional fee (that may be fixed sum or percent of costs incurred)

Units Contract u

Client commits to a fixed price for a pre-specified unit of work; final payment is based on number of units produced


Incentive (Risk Sharing) Contracts General Form: Payment to Subcontractor = Fixed Fee + (1 - B) (Project Cost) where B = cost sharing rate Cost Plus Contract B=0

Fixed Price Contract Linear & Signalling Contracts

B=1


Why Use Incentive Contracts? Expected Cost of Project = $100M Two firms bid on subcontract Firm 1

Firm 2

Fixed Fee (bid)

$5 M

$7 M

Project Cost

$105 M

$95 M

(inefficient producer) What is result if Cost Plus Contract (B = 0) used?


Washington State Bid Code (WAC 236-48-093) n

n

n

n

n

n n n

n

WAC 236-48-093: A contract shall be awarded to the lowest responsible and responsive bidder based upon, but not limited to, the following criteria where applicable and only that which can be reasonably determined: 1) The price and effect of term discounts...price may be determined by life cycle costing if so indicated in the invitation to bid 2) The conformity of the goods and/or services bid with invitation for bid or request for quotation specifications depicting the quality and the purposes for which they are required. 3) The ability, capacity, and skill of the bidder to perform the contract or provide the services required. 4) The character, integrity, reputation, judgement, experience, and efficiency of the bidder. 5) Whether the bidder can perform the contract with the time specified. 6) The quality of performance on previous contracts for purchased goods or services. 7) The previous and existing compliance by the bidder with the laws relating to the contract for goods and services. 8) Servicing resources, capability, and capacity.


Competitive Bidding: Low-Bid System n

“In the low-bid system, the owner wants the most building for the least money, while the contractor wants the least building for the most money. The two sides are in basic conflict.� Steven Goldblatt Department of Building Construction University of Washington The Seattle Times, Nov 1, 1987


Precedence Networks Networks represent immediate precedence relationships among tasks (also known as work packages or activities) and milestones identified by the WBS Milestones (tasks that take no time and cost $0 but indicate significant events in the life of the project) Two types of networks: Activity-on-Node (AON) Activity-on-Arc (AOA) All networks: must have only one (1) starting and one (1) ending point


Precedence Networks: Activity-on-Node (AON)

A

C

Start

End

B

D


Precedence Diagramming Standard precedence network (either AOA or AON) assumes that a successor task cannot start until the predecessor(s) task(s) have been completed. Alternative relationships can be specified in many software packages: Finish-to-start (FS = α): Job B cannot start until α days after Job A is finished Start-to-start (SS = α): Job B cannot start until α days after Job A has started Finish-to-finish (FF = α): Job B cannot finish until α days after Job A is finished Start-to-finish (SF = α): Job B cannot finish until α days after Job A has started


Critical Path Method (CPM): Basic Concepts

Task A 7 months

Task B 3 months

Start End

Task C 11 months


Critical Path Method (CPM): Basic Concepts ESA = 0 LFA = 8 ESStart = 0 LFStart = 0

ESB = 7 LFB = 11 Task A 7 months

Task B 3 months

ESEnd = 11 LFEnd = 11

Start

End

Task C 11 months ESC = 0 LFC = 11

ESj = Earliest starting time for task (milestone) j LFj = Latest finish time for task (milestone) j


AON Precedence Network: Microsoft Project

Task A

Task B

2

7d

3

3d

Wed 12/20/00

Thu 12/28/00

Fri 12/29/00

Tue 1/2/01

Start 1

0d

Wed 12/20/00

Wed 12/20/00

End

Task C 4

11d

Wed 12/20/00

Wed 1/3/01

5

0d

Wed 1/3/01

Wed 1/3/01


Critical Path Method (CPM): Example 2 ES A = LFA =

Task A 14 wks ES START = 0 LFSTART = 0 ES B = LFB = START

Task B 9 wks

ES C = LFC =

Task C 20 wks

ES F = LFF = ES D = LFD =

Task F 9 wks

Task D 12 wks

ES END = LFEND=

END

ES E = LFE =

Task E 6 wks


Example 2: Network Paths

Path 1 2 3 4 5

Tasks START-A-D-F-END START-A-D-E-END START-B-D-F-END START-B-D-E-END START-C-E-END

Expected Duration (wks) 35 32 30 27 26


Example 2: CPM Calculations EARLIEST

Task or Milestone START

A B C D E F END

Duration ( ti ) 0 14 9 20 12 6 9 0

S tart Time (ES i) 0 0 0 0 14 26 26 35

Finish Time 0 14 9 20 26 32 35 35

L A TE S T

S tart Time 0 0 5 9 14 29 26 35

Finish Time (LFi) 0 14 14 29 26 35 35 35


Example 2: Calculating Total Slack (TSi) Total Slack for task i = TSi = LFi - ESi - ti

Task or Milestone

Duration ( ti )

Earliest Start Time (ES i)

START

0 14 9 20 12 6 9 0

0 0 0 0 14 26 26 35

A B C D E F END

Lastest Finish Time (LFi) 0 14 14 29 26 35 35 35

Total S lack (TSi)

Critical Task?

0

Yes

0

Yes

5 9

No No

0

Yes

3

No

0 0

Yes Yes


Slack (Float) Definitions (for task i) Total Slack (TSi)

= LFi - ESi - ti

Free Slack (FSi)

= ESi,min - ESi - ti

where ESi,min = minimum early start time of all tasks that immediately follow task i = min (ESj for all task j ∈ Si)

Safety Slack (SSi)

= LFi - LFi,max - ti

where LFi,max = maximum late finish time of all tasks that immediately precede task i = min (LFj for all task j ∈ Pi)

Independent Slack (ISi)

= max (0, ESi,min - LFi,max - ti)


Example #2: LP Model Decision variables: STARTj = start time for task j END = ending time of project (END milestone)

Minimize END subject to STARTj ≥ FINISHi STARTj ≥ 0

for all tasks i that immediately precede task j for all tasks j in project

where FINISHi = STARTi + ti = STARTi + duration of task i


Example #2: Excel Solver Model


Gantt Chart February ID

Task Name

1

Start

2

Task A

3

Task B

4

Task C

5

Task D

6

Task E

7

Task F

8

Task G

9

Task H

10

Task J

11

End

March 21 24

27

1

4

7

10 13

16

April 19 22 25 28

31

3

6

9

May

12 15 18

21

24

27 30

3

6

9

12

15 18

3/1 Workers[5] Workers[7] Workers[3] Workers[12] Workers[2] Workers Workers[2] Workers[5] Workers[6] 5/10

Microsoft Project 4.0


Project Budgeting • The budget is the link between the functional units and the project • Should be presented in terms of measurable outputs • Budgeted tasks should relate to work packages in WBS and organizational units responsible for their execution • Should clearly indicate project milestones • Establishes goals, schedules, and assigns resources (workers, organizational units, etc.) • Should be viewed as a communication device • Serves as a baseline for progress monitoring & control • Update on rolling horizon basis • May be prepared for different levels of aggregation (strategic, tactical, short-range)


Project Budgeting (cont’d) • Top-down Budgeting: Aggregate measures (cost, time) given by top management based on strategic goals

and constraints

• Bottom-up Budgeting: Specific measures aggregated up from WBS tasks/costs and subcontractors


Issues in Project Budgets • How to include risk and uncertainty factors? • How to measure the quality of a project budget? • How often to update budget? • Other issues?


Critical Path Method (CPM): Example 2 ES A = 0 LFA = 14

Task A 14 wks ES START = 0 LFSTART = 0 ES B = 0 LFB = 14

ES F = 26 LFF = 35 ES D = 14 LFD = 26

Task F 9 wks

Task D 12 wks

START

Task B 9 wks

ES C = 0 LFC = 29

Task C 20 wks

ES END = 35 LFEND= 35

END

ES E = 26 LFE = 35

Task E 6 wks


Project Budget Example Task or Milestone ST ART

A B C D E F END

Duration (tj) 0 14 9 20 12 6 9 0

Early Start Time (ESj)

No. of No. of Latest Start Resource A Resource B Time (LSj) work ers workers

Material Costs

Direct Labor Labor + Materials Cost/wk

0 0 0 0 14

0 0 5 9 14

2 4 3 0

0 12 14 8

-

-

$ $ $ $

340 125 200

$ $ $ $

800 8,800 9,600 4,800

$ $ $ $

1,140 8,925 9,600 5,000

26

29

1

0

$

560

$

400

$

960

26 35

26 35

4 -

10 -

$

90

$

7,600 -

$

7,690 -

-

Cost for Resource A worker = $400/week Cost for Resource B worker = $600/week


Project Budget Example (cont’d) Week

Early Start Times Task 1 A 1140 B 8925 C 9600 D E F

2

3

4

5

6

7

8

9

10

11

12

800 8800 9600

800 8800 9600

800 8800 9600

800 8800 9600

800 8800 9600

800 8800 9600

800 8800 9600

800 8800 9600

800

800

800

9600

9600

9600

19665 19665

19200 38865

19200 58065

19200 77265

19200 96465

19200 115665

19200 134865

19200 154065

19200 173265

10400 183665

10400 194065

10400 204465

Weekly Subtotals Cumulative

Week

Late Start Times Task A B C D E F Weekly Subtotals Cumulative

1

2

3

4

5

6

7

8

9

10

11

12

1140

800

800

800

800 8925

800 8800

800 8800

800 8800

800 8800 9600

800 8800 9600

800 8800 9600

800 8800 9600

1140 1140

800 1940

800 2740

800 3540

9725 13265

9600 22865

9600 32465

9600 42065

19200 61265

19200 80465

19200 99665

19200 118865


Cumulative Costs 450000 400000

300000

Range of feasible budgets

250000 200000 150000 100000 50000

Week Early Start Sc hedule

Late Start Schedule

33

31

29

27

25

23

21

19

17

15

13

11

9

7

5

3

0 1

Cumulativ e Cost

350000


Weekly Costs (Cash Flows) 25000

15000

10000

5000

Week Early Start Schedule

Late Start Schedule

33

31

29

27

25

23

21

19

17

15

13

11

9

7

5

3

0 1

Weekly Costs

20000


Managing Cash Flows • Want to manage payments and receipts • Must deal with budget constraints on project and organization requirements (e.g., payback period) • Organization profitability


Cash Flow Example Make payment of $5000

M1 Task A 2 mos

Task D 8 mos

Receive payment of $3000

Task C 4 mos

START

END Task B 8 mos

Task E 3 mos

M2

Receive payment of $3000


Cash Flow Example: Solver Model


Material Management Issues When to order materials? How much to order? Example: • Single material needed for Task B (2 units) and Task E (30 units) • Fixed cost to place order = S • Cost of holding raw materials proportional to number of unit-weeks in stock • Cost of holding finished product greater than the cost of holding raw materials • Project can be delayed (beyond 17 weeks) at cost of $P per week


Material Management Example LS A = 0

LS B = 4

Task A 4 wks

Task B 8 wks

LS C = 12 Task C 5 wks

2 units

Start LS D = 6

LS E = 12

Task D 6 wks

Task E 2 wks

30 units

End LS F = 14 Task F 3 wks


Lot-Sizing Decisions in Projects • To minimize holding costs, only place orders at Late Starting Times • Can never reduce holding costs by delaying project Time 1

2

3

4

Demand:

2

Order option #1:

32

Order option #2:

2

5

6

7

8

9

10

11

12 30

30

Choose the option that minimizes inventory cost = order cost + holding cost of raw materials


Time-Cost Tradeoffs


Time-Cost Tradeoff Example A

C

Start

End

B

D

Task

Normal Duration

A B C D

7 6 15 10

M arginal Cost to Crash One Normal Cost Week $60 $85 $55 $120

$8 $5 $10 $4


Time-Cost Tradeoff Example (cont’d) Project Duration (weeks)

Total Direct Cost

Critical Path(s)

Task(s) Reduced

22 21

Start-A-C-End

A

$320

20

Start-A-C-End

C

$338

19

Start-A-C-End

C

$348

18

Start-A-C-End

A, B

$361

Start-A-C-End Start-A-B-End Start-A-B-End Start-A-B-End Start-A-B-End

$328


Linear Time-Cost Tradeoff In theory, the normal or expected duration of a task can be reduced by assigning additional resources to the task Cost Crash Point

Crash cost = Ccj

Slope (bj) = Increase in cost by reducing task by one time unit

Normal Point

Normal cost = CN j

Crash time =

tcj

Time Normal time

N = tj


Balancing Overhead & Direct Costs Cost

Total Cost Indirect (overhead) Costs

Direct Costs

Crash Time

Minimum Cost Solution

Normal Time

Project Duration


Time-Cost Tradeoff (Direct Costs Only) Given Normal point with cost CNj and time tNj c

c and time tj

and Crash point with cost Cj

Ccj - CNj Assume constant marginal cost of crashing task j = bj = c N tj - tj Decision Variables:

Sj = Starting time of task j

END = End time of project tj = Duration of task j Minimize Total Direct Cost = Sj ≥ Si + ti tcj Š tj Š tNj END = Tmax tj , Sj ≥ 0

bj tj j

for all tasks i ∈ Pj for all tasks in project


General Time-Cost Tradeoffs Minimize Total Costs =

bj tj

+ I (END) + P L

j

where I = indirect (overhead) cost/time period P = penalty cost/time period if END is delayed beyond deadline Tmax L = number of time periods project is delayed beyond deadline Tmax QUESTION: HOW TO DEFINE L?


Software Project Schedules “Observe that for the programmer, as for the chef, the urgency of the patron may govern the scheduled completion of the task, but it cannot govern the actual completion. An omelet, promised in ten minutes, may appear to be progressing nicely. But when it has not set in ten minutes, the customer has two choices--wait or eat it raw. Software customers have the same choices. The cook has another choice; he can turn up the heat. The result is often an omelet nothing can save--burned in one part, raw in another.” F.P. Brooks, “The Mythical Man-Month”, Datamation, Vol 20, No 12 (Dec, 1974), pp. 44-52.


Coordination Costs (Software Development Project) n

n n

Assume you want to develop program that will require (approximately) 50,000 lines of PERL code A typical programmer can write approximately 1500 lines of code per week Coordination time is M (M-1)/2 weeks 12

No. of Weeks Coding 33.33 16.67 11.11 8.33 6.67 5.56 4.76 4.17 3.70 3.33 3.03

No. of Total Coordination Number of Weeks Weeks 0 33.33 1 17.67 3 14.11 6 14.33 10 16.67 15 20.56 21 25.76 28 32.17 36 39.70 45 48.33 55 58.03

10 No. of Programmers (Cost)

No. of Programmers 1 2 3 4 5 6 7 8 9 10 11

8

6 4

2 0 0

10

20

30

40

Total Number of Weeks

50

60

70


Brook’s Law

“Adding manpower to a late software project makes it later.” n

F.P. Brooks, “The Mythical Man-Month”, Datamation, Vol 20, No 12 (Dec, 1974), pp. 44-52.


Compressing New Product Development Projects

Traditional Method Design follows a sequential pattern where information about the new product is slowly accumulated in consecutive stages Stage 0

Stage 1

Stage N


New Product Development Process Overlapped Product Design Allows downstream design stages to start before preceding upstream stages have finalized their specifications‌. Stage 0 Stage 1 Stage N


Issues and Tradeoffs What are the tradeoffs when moving from a traditional sequential product design process to an overlapped product design process? • Increased uncertainty (that leads to additional work) • Can add additional resources to tasks to reduce duration--but costs are increased


Classic PERT Model Defined • Since task durations are now random variables, time of any milestone (e.g., end of project) is now RV • Assume all tasks are statistically independent • Use values of µj to identify expected critical path • Since time of event (e.g., ESk) is now sum of independent RV’s, central limit theorem specifies that ESk is approximately normally distributed with mean E[ESk] and variance Var[ESk] Expected early start time of task k = EES k = max s where there exists s paths to task k

µj tasks j on path s


Classic PERT Model (cont’d) Thus, expected project duration is defined as: µj

Expect Project Duration = E[ES END ] = tasks j on CP

σ 2j

Variance of Project Duration = Var[ES END ] = tasks j on CP

Using central limit theorem and standard normal distribution: P ESEND Š Tmax = P z Š

Tmax - E ESEND Var ESEND


PERT Example #1 Task B Programming Task E Implementation

S tart

Task A Requirements Analysis

Task C Hardware Acquisition

Task F Testing

End Task D User Training

Duration Estimates Task A B C D E F END

Description Requirements Analy sis Programming Hardware acquisition User training Implementation Testing End of project

Predecessors none A A A B, C E D, F

Optimistic 2 4 2 12 3 3 0

Pessimistic 14 12 13 18 7 7 0

Likely

Expected Duration

Variance

6 7 8 14 5 4 0

6.67 7.33 7.83 14.33 5.00 4.33 0.00

4.00 1.78 3.36 1.00 0.44 0.44 0.00


PERT Example #1 (cont’d) Task B Programming Task E Implementation

S tart

Task A Requirements Analysis

Task C Hardware Acquisition

Task F Testing

End Task D User Training

Task B,C,D E F End

Path Start-A Start-A-C Start-A-C-E Start-A-C-E-F-End

PERT Expected Duration = PERT Variance =

Expected Early Start 6.67 14.50 19.50 23.83 23.83 8.250

Variance

Due Date

Zi

4.00 7.36 7.81 8.25

6 15 20 25

-0.33 0.18 0.18 0.41

Expected CP

= {Start, A, C, E, F, End}

Pr(zi) 0.37 0.57 0.57 0.66


PERT Example #2 Task A

Task C

µA = 4

µC = 10

σ 2A

=2

σ 2C = 5

END

START Task B

Task D

µB = 12

µD = 3

σ B2 = 4

σ 2D = 1


Example #3: Discrete Probabilities Task A (8.0) Task D (9.3) Task B (10.0)

S TART

END

Task C (19.0)

Task A

Task B

Task C

Task D

Value

Prob

Value

Prob

Value

Prob

Value

Prob

7

0.333

2

0.2

5

0.2

3

0.3

8

0.333

12

0.8

15

0.2

12

0.7

9

0.333

25

0.6


Example #3 (cont’d) Task A Combination 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36

Value 7 7 7 7 7 7 7 7 7 7 7 7 8 8 8 8 8 8 8 8 8 8 8 8 9 9 9 9 9 9 9 9 9 9 9 9

Task B Prob 0.333 0.333 0.333 0.333 0.333 0.333 0.333 0.333 0.333 0.333 0.333 0.333 0.333 0.333 0.333 0.333 0.333 0.333 0.333 0.333 0.333 0.333 0.333 0.333 0.333 0.333 0.333 0.333 0.333 0.333 0.333 0.333 0.333 0.333 0.333 0.333

Value 2 2 2 2 2 2 12 12 12 12 12 12 2 2 2 2 2 2 12 12 12 12 12 12 2 2 2 2 2 2 12 12 12 12 12 12

Task C Prob 0.2 0.2 0.2 0.2 0.2 0.2 0.8 0.8 0.8 0.8 0.8 0.8 0.2 0.2 0.2 0.2 0.2 0.2 0.8 0.8 0.8 0.8 0.8 0.8 0.2 0.2 0.2 0.2 0.2 0.2 0.8 0.8 0.8 0.8 0.8 0.8

Value 5 5 15 15 25 25 5 5 15 15 25 25 5 5 15 15 25 25 5 5 15 15 25 25 5 5 15 15 25 25 5 5 15 15 25 25

Task D Prob 0.2 0.2 0.2 0.2 0.6 0.6 0.2 0.2 0.2 0.2 0.6 0.6 0.2 0.2 0.2 0.2 0.6 0.6 0.2 0.2 0.2 0.2 0.6 0.6 0.2 0.2 0.2 0.2 0.6 0.6 0.2 0.2 0.2 0.2 0.6 0.6

Value 3 12 3 12 3 12 3 12 3 12 3 12 3 12 3 12 3 12 3 12 3 12 3 12 3 12 3 12 3 12 3 12 3 12 3 12

Critical Prob 0.3 0.7 0.3 0.7 0.3 0.7 0.3 0.7 0.3 0.7 0.3 0.7 0.3 0.7 0.3 0.7 0.3 0.7 0.3 0.7 0.3 0.7 0.3 0.7 0.3 0.7 0.3 0.7 0.3 0.7 0.3 0.7 0.3 0.7 0.3 0.7

Path A, D A, D C A, D C C B, D B, D B, D B, D C C A, D A, D C A, D C C B, D B, D B, D B, D C C A, D A, D C A, D C C B, D B, D B, D B, D C C

Prob of

Length

CP 0.004 0.009 0.004 0.009 0.012 0.028 0.016 0.037 0.016 0.037 0.048 0.112 0.004 0.009 0.004 0.009 0.012 0.028 0.016 0.037 0.016 0.037 0.048 0.112 0.004 0.009 0.004 0.009 0.012 0.028 0.016 0.037 0.016 0.037 0.048 0.112

of CP 10 19 15 19 25 25 15 24 15 24 25 25 11 20 15 20 25 25 15 24 15 24 25 25 12 21 15 21 25 25 15 24 15 24 25 25

PATHS A,D

B, D

C

0.004 0.009 0.000 0.009 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.004 0.009 0.000 0.009 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.004 0.009 0.000 0.009 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000

0.000 0.000 0.000 0.000 0.000 0.000 0.016 0.037 0.016 0.037 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.016 0.037 0.016 0.037 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.016 0.037 0.016 0.037 0.000 0.000

0.000 0.000 0.004 0.000 0.012 0.028 0.000 0.000 0.000 0.000 0.048 0.112 0.000 0.000 0.004 0.000 0.012 0.028 0.000 0.000 0.000 0.000 0.048 0.112 0.000 0.000 0.004 0.000 0.012 0.028 0.000 0.000 0.000 0.000 0.048 0.112

6.8%

32.0%

61.1%


Example #3 (cont’d) Length of CP's 10 11 12 15 19 20 21 24 25

Cumulativ e Prob 0.004 0.004 0.004 0.108 0.019 0.019 0.019 0.224 0.599

Prob 0.00 0.01 0.01 0.12 0.14 0.16 0.18 0.40 1.00

Criticality Indices Task A 6.8%

Task B 32.0%

Task C 61.1%

Task D 38.8%

Expected Project Duration = 23.22


Monte-Carlo Simulation (PERT Example 1) Task A B C D E F END

Run 1 2 3 4 5 6 7

Task Duration (Uniform Dist) 4.99 4.75 3.38 12.20 5.94 5.34 0.00

Project Duration 31.07 27.41 23.97 28.93 26.85 28.82 28.77

197 198 199 200

30.37 29.78 25.33 29.70

Ave Var

27.13 16.777

Early Start

Latest Finish

Total Slack

Expected Duration

Variance

0 4.99 4.99 4.99 9.74 15.68 21.02

4.99 9.74 9.74 21.02 15.68 21.02 21.02

0.00 0.00 1.36 3.83 0.00 0.00 0.00

6.67 7.33 7.83 14.33 5.00 4.33 0.00

4.00 1.78 3.36 1.00 0.44 0.44 0.00

t(B) 1 0 1 0 1 0 0

t(C) 0 1 0 1 0 0 1

t(D) 0 0 0 0 0 1 0

t(E) 1 1 1 1 1 0 1

t(F) 1 1 1 1 1 0 1

0 1 1 0

1 0 0 1

0 0 0 0

1 1 1 1

1 1 1 1

48.5%

42.0%

Project Makespan 95% Confidence interval 99% Confidence interval

9.5%

Lower Limit 26.56 26.37

90.5%

Upper Limit 27.72 27.90

90.5%


Calculating Confidence Intervals For a confidence interval, we can use the sample mean X and the estimated standard error of the mean sX = sn where s is the sample standard deviation and n is the number of trials Using a normal approximation, a (1- Îą) twosided confidence interval is given by

+

X - zÎą/2 sX


New Product Development Projects

Lease Mfg/Office Space

Beta test fails (with probability of 0.25) and rework is needed

Design of physical unit Assemble prototype

Beta test prototype

START END Identify/hire staff

Electronics design

Software

Beta test fails (with probability of 0.25) and rework is needed


New Product Development Projects (cont’d)

Lease Mfg/Office Space

Design of physical unit

Prob = .75 Assemble prototype

START

Beta test prototype Prob = .25

Identify/hire staff

Electronics design

Beta test fails and rework is needed Software

END


Critical Chain and the Theory of Constraints (TOC) Project “Goal” (according to Goldratt): Meet Project Due Date

• Use deterministic CPM model with buffers to deal with any uncertainties, • Place project buffer after last task to protect the customer’s completion schedule, • Exploit constraining resources (make certain that resources are fully utilized), • Avoid wasting time slack time by encouraging early task completions, • Carefully monitor the status of the buffer(s) and communicate this status to other project team members on a regular basis, and • Make certain that the project team is 100 percent focused on critical chain tasks


Project Buffer Defined • Project Buffer is placed at the end of the project to protect the customer’s promised due date

Task B Programming Task E Implementation

Start

Task A requirements analysis

Task C Hardware acquisition

Task F Testing

Project Buffer

Task D User User training

PERT Example #1 Revisited with Project Buffer

End


Calculating Project Buffer Size For those “who want a scientific approach to sizing buffers....” For tasks k on critical chain, we can calculate project buffer using following formula that project will be completed within worst-case duration estimates around 90 percent of the time:

Buffer =

tasks k on critical chain

tpk - µk

2


Implications of Project Uncertainty Task A

END

START Task B

Assume that the duration of both tasks A and B are described by a normal distribution with a mean of 30 days What is the probability that the project will be completed within 30 days?


Uncertainty and Worker Behavior Consider a project with two tasks that must be completed serially The duration of each task is described by a RV with values Ti (i = 1, 2)

Start

Task 1

Task 2

End

Values of T 1

Prob

Values of T2

Prob

7 8 9

0.3 0.4 0.3 8.0

14 18

0.5 0.5 16


Parkinson’s Law (Expanding Work) “Work expands so as to fill the time available for its completion” Professor C.N. Parkinson (1957) Set a deadline D = 24 days So T(D) = project makespan (function of D) where E[T(D)] = E(T1) + E(T2) + E[max(0, D - T1 - T2)] Values of T1 7 7 8 8 9 9

Prob 0.3 0.3 0.4 0.4 0.3 0.3

Values of T2 14 18 14 18 14 18

Prob 0.5 0.5 0.5 0.5 0.5 0.5

E[T(D)] = 25 days

Project Makespan 24 25 24 26 24 27

Prob 0.15 0.15 0.2 0.2 0.15 0.15


Procrastinating Worker Set a deadline D = 24 days E’[T(D)] = E(T1) + E(T2) + E{max[0, D - T1 - E(T2)]}

Values of T 1

Prob

E[Delay] = max[0, D - T1 - E(T2)]

7 8 9

0.3 0.4 0.3 8

1 0 0 0.3

E[Makespan] 24 24 25 24.30

Can show that E[T(D)] ≥ E’[T(D)] ≥ D What are the implications for project managers?


Schoenberger’s Hypothesis An increase in the variability of task durations will increase the expected project duration‌.


Schoenberger’s Hypothesis Illustrated

Task A

END

START

Task B

Duration of Task A 12 14 16 14.0

Probability 0.1 0.8 0.1

Duration of Task B

Probability

10 15

0.5 0.5

12.5


Schoenberger’s Hypothesis Illustrated Realization 1 2 3 4 5 6

Task A Duration 12 14 16 12 14 16

Task B Duration 10 10 10 15 15 15

Probability 0.05 0.4 0.05 0.05 0.4 0.05

Max (A, B) 12 14 16 15 15 16

Expected duration equals 14.55 days Increasing the variance of Task A: Duration of Task A 12 14 16 14.0

Probability 0.3 0.4 0.3

Duration of Task B 10 15

Probability 0.5 0.5

12.5

Results in an increased expected duration = 14.65 days


Risk Management • All projects involve some degree of risk • Need to identify all possible risks and outcomes • Need to identify person(s) responsible for managing project risks • Identify actions to reduce likelihood that adverse events will occur


Risk Analysis Risk Exposure (RE) or Risk Impact = (Probability of unexpected loss) x (size of loss) Example: Additional features required by client Loss: 3 weeks Probability: 20 percent Risk Exposure = (.20) (3 weeks) = .6 week


How to Manage Project Risks? Preventive Actions • Actions taken in anticipation of adverse events • May require action before project actually begins • Examples?

Contingency Planning • What will you do if an adverse event does occur? • “Trigger point” invokes contingency plan • Frequently requires additional costs


Risk and Contracts

High

Low

Low

High

Degree of Risk Contractor Client Fixed Price Contract

Cost Plus Contract

Elements Cost Plus can be T&M with Firm price renegotiated Incentives with limits Incentives

Time & materials


Tornado Diagram Wage Rate Direct Labor Hours M aterial Units Needed

$1260 $1290

$1760 1760 $1700 $1720

$1265

$1680

$1310 1310 $1350

Early Completion Bonus Material Unit Cost

$1690

$1350

Interest rates Energy costs

$1640

$1380 $1400

Overhead $1200

$1300

$1620 $1625

$1400

$1500

$1600

Project Cost ($000's)

$1700 $1800


Sensitivity Chart Wage Rate Direct Labor Hours M aterial Units Needed Early Completion Bonus M aterial Unit Cost Interest rates Energy costs Overhead

0.85 0.73 0.62 -0.45 0.42 0.28 0.19 0.10 -0.5

0

0.5

1.0

Rank Order Correlation with Total Project Cost


Van Allen Company Task

Immediate Predecessors

Start A B C D E F G H I End

Start Start B A A C, D C, D G E, G F, H, I

Strike (wks)

Time

Cost

Time

Cost

0

-

0

-

3 1 5 2 2 5 4 1 1 0

$60 $50 $70 $60 $50 $90 $60 $40 $50 -

5 5 10 7 6 11 6 5 4 0

$40 $30 $40 $40 $30 $60 $30 $20 $20 -

Prob

3 0.45 4 0.3 5 0.25 E[Strike Duration]

Expected Duration 1.35 1.20 1.25 3.80


Resource Allocation & Leveling Resource Leveling: Reschedule the noncritical tasks to smooth resource requirements

Resource Allocation: Minimize project duration to meet resource availability constraints


Resource Allocation & Leveling Three types of resources: 1) Renewable resources: “renew� themselves at the beginning of each time period (e.g., workers) 2) Non-Renewable resources: can be used at any rate but constraint on total number available 3) Doubly constrained resources: both renewable and non-renewable


Resource Leveling Task C 9 wks

Task A 3 wks

Task D 5 wks

S TART

Task G 5 wks

Task B 2 wks

END

Task E 3 wks

Task F 2 wks

Task A B C D E F G

Workers 7 3 2 10 4 5 6

Duration (t j) 3 2 9 5 3 2 5

Early S tart 0 0 3 3 2 2 8

Late S tart 0 3 4 3 5 11 8


Resource Leveling: Early Start Schedule Early Start Schedule

Number of Workers Needed

25 20 Task G Task F Task E Task D Task C Task B Task A

15

10 5 0 1

2

3

4

5

6

7 Week

8

9

10

11

12

13


Resource Leveling: Late Start Schedule Late Start Schedule 18

Number of Workers Needed

16 14

Task G Task F Task E Task D Task C Task B Task A

12 10 8 6 4 2 0 1

2

3

4

5

6

7 Week

8

9

10

11

12

13


Resource Leveling: Microsoft Project Dec 17, '00 T

Dec 24, '00

W

T

10

10

F

S

S

Dec 31, '00

M

T

W

T

F

10

10

10

10

10

S

S

Jan 7, '01

M

T

W

T

F

10

10

16

16

16

S

S

M

T

W

T

F

16

16

21

21

21

25

20

15

10

5

Worke rs

10 Ov erallocated:

Allocated:

S


Renewable Resource Allocation Example (Single Resource Type) 3 workers

6 workers

Task A 4 wks

Task C 1 wk

Task E 4 wks

START

Task B 3 wks

Task D 5 wks

5 workers

8 workers

7 workers

Maximum number of workers available = R = 9 workers

END


Resource Allocation Example: Early Start Schedule Task C: 6 workers

Task A: 3 workers S tart

End Task B: 5 workers

Task E: 7 workers

Task D: 8 workers

Week

1

2

3

4

5

6

7

8

9

10

11

12

No. of Workers/wk Cumulative Workers "Wasted" worker-wks

8 8 1

8 16 1

8 24 1

11 35 -

14 49 -

8 57 -

8 65 -

8 73 -

7 80 -

7 87 -

7 94 -

7 101 -

Maximum number of workers available = R = 9 workers


Resource Allocation Example: Late Start Schedule

Task A: 3 workers S tart

Task C: 6 workers

End

Task B: 5 workers

Task E: 7 workers

Task D: 8 workers

Week

1

2

3

4

5

6

7

8

9

10

11

12

No. of Workers/wk Cumulative Workers "Wasted" worker-wks

5 5 -

5 10 -

5 15 -

11 26 -

11 37 -

11 48 -

11 59 -

14 73 -

7 80 2

7 87 2

7 94 2

7 101 2

Maximum number of workers available = R = 9 workers


Resource Allocation Heuristics n

Some heuristics for assigning priorities to available tasks j, where number of units of resource k used by task j

n

1) FCFS: Choose first available task

Rkj denotes the

Rkj

n

2) GRU: (Greatest) resource utilization =

n

3) GRD: (Greatest) resource utilization x task duration =

k

n

4) ROT: (Greatest) resource utilization/task duration =

Rkj / tj

n

5) MTS: (Greatest) number of total successors

n

6) SPT: Shortest processing time = min {tj}

n

7) MINSLK: Minimum (total) slack

n

8) LFS: Minimum (total) slack per successor

n

9) ACTIMj: (Greatest) time from start of task j to end of project = CP - LSj

n

10) ACTRESj: (max) (ACTIMj)

n

11) GENRESj: w ACTIMj + (1-w) ACTRESj where 0 ≤ w ≤ 1

Rkj tj

k

k


Resource Allocation Problem #2

Start

Task A1 6 days

Task A2 4 days

Task B1 3 days

Task B2 5 days

Task C1 2 days

Task C2 5 days

Gold Crew

Purple Crew

End


How to schedule tasks to minimize project makespan? Priority scheme: schedule tasks using total slack (i.e., tasks with smaller total slack have higher priority)

Gold Crew

Purple Crew

Task A1

Task B1

1

2

3

4

5

6

7

1

2

3

4

5

6

7

8

Task C1 9

10

11

12

9

10

11

12

Task A2 8

13

14

15

16

17

15

16

17

Task B2 13

14

18

19

20

Task C2 18

19

20


Resource Allocation Example (cont’d) But, can we do better? Is there a better priority scheme?

Gold Crew

Purple Crew

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20


Microsoft Project Solution (Resource Leveling Option)

Solution by: Microsoft Project 2000


Critical Chain Project Management • Identify the critical chain: set of tasks that determine the overall duration of the project • Use deterministic CPM model with buffers to deal with uncertainty • Remove padding from activity estimates (otherwise, slack will be wasted). Estimate task durations at median. • Place project buffer after last task to protect customer’s completion schedule • Exploit constraining resource(s) • Avoid wasting slack times by encouraging early task completions • Have project team focus 100% effort on critical tasks • Work to your plan and avoid tampering • Carefully monitor and communicate buffer status


Critical Chain Buffers Project Buffer

: placed after last task in project to protect schedule

Feeding Buffers

: placed between a noncritical task and a critical task

when the noncritical task is an immediate predecessor of the critical task

Resource Buffers resource type

: placed just before a critical task that uses a new


Critical Chain Illustrated Feeding Buffers Task C1 2 days

Task B1 3 days

Task A1 6 days

End

Start

Task C2 5 days

Resource Buffers

Task B2 5 days

Task A2 4 days


Non-Renewable Resources 12 units

Task B 5 wks 6 units

8 units

Task A 6 wks

START

Task D 2 wks

END

Task C 3 wks

10 units

Task A B C D

Duration 6 5 3 2

No. of Nonrenewable Resources Units Needed 6 12 10 8

Early S tart 0 6 6 11

Late S tart 0 6 8 11


Non-Renewable Resources: Graphical Solution

Cumulative Resources Supplied

40

Cumulative Re source s

36 32

Cumulative Resources Required

28 24 20 16 12 8 4 1

2

3

4

5

6

7

8

9

10

We eks

11

12

13

14

15

16

17

18

19

20


Resource Allocation Problem #3 Issue: When is it better to “team” two or more workers versus letting them work separately? • Have 2 workers, Bob and Barb, and 4 tasks: A, B, C, D • Bob and Barb can work as a team, or they can work separately • When should workers be assigned to tasks? Which configuration do you prefer?


How to Assign Project Teams? A

C

Start

End B

D

Configuration #1 Bob and Barb work jointly on all four tasks; assume that they can complete each task in one-half the time needed if either did the tasks individually

Configuration #2 Bob and Barb work independently. Bob is assigned to tasks A and C; Barb is assigned to tasks B and D


Bob and Barb: Configuration #1 TASK A Duration 6 5 4 Expected duration

Prob 0.33 0.33 0.33 5.0

TASK B Duration 9 6

Prob 0.667 0.333

TASK C Duration 12 7

8.0

Prob

TASK D Duration

0.6 0.4

10 6

10.0

Configuration #1 Bob and Barb work jointly on all four tasks. What is the expected project makespan?

Prob 0.25 0.75

7.0


Bob and Barb: Configuration #2 Bob and Barb work independently. Bob is assigned to tasks A and C; Barb is assigned to tasks B and D

Realization #

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

A 6 6 6 6 6 6 6 6 5 5 5 5 5 5 5 5 4 4 4 4 4 4 4 4

B 9 9 9 9 6 6 6 6 9 9 9 9 6 6 6 6 9 9 9 9 6 6 6 6

C 12 12 7 7 12 12 7 7 12 12 7 7 12 12 7 7 12 12 7 7 12 12 7 7

D 10 6 10 6 10 6 10 6 10 6 10 6 10 6 10 6 10 6 10 6 10 6 10 6

Bob A+C

Barb B+D

18 18 13 13 18 18 13 13 17 17 12 12 17 17 12 12 16 16 11 11 16 16 11 11

19 15 19 15 16 12 16 12 19 15 19 15 16 12 16 12 19 15 19 15 16 12 16 12

max (A+C, B+D) 19 18 19 15 18 18 16 13 19 17 19 15 17 17 16 12 19 16 19 15 16 16 16 12

Prob 0.03 0.10 0.02 0.07 0.02 0.05 0.01 0.03 0.03 0.10 0.02 0.07 0.02 0.05 0.01 0.03 0.03 0.10 0.02 0.07 0.02 0.05 0.01 0.03


Bob and Barb: Configuration #2 Bob and Barb work independently. Bob is assigned to tasks A and C; Barb is assigned to tasks B and D

max (A+C, B+D) 12 13 15 16 17 18 19

Prob 0.07 0.03 0.20 0.20 0.17 0.17 0.17

Cumulativ e Prob 0.07 0.10 0.30 0.50 0.67 0.83 1.00

Expected Project Makespan: 16.42


Parallel Tasks with Random Durations

Task A START

END

Task B

• Assume that both Tasks A and B have possible durations: 8 days with probability = 0.5 10 days with probability = 0.5 • What is expected duration of project? (Is it 9 days?)


Project Monitoring and Control n

“It is of the highest importance in the art of detection to be able to recognize, out of a number of acts, which are incidental and which are vital. Otherwise your energy and attention must be dissipated instead of being concentrated.� Sherlock Holmes


Status Reporting? One day my Boss asked me to submit a status report to him concerning a project I was working on. I asked him if tomorrow would be soon enough. He said, "If I wanted it tomorrow, I would have waited until tomorrow to ask for it!" New business manager, Hallmark Greeting Cards


Control System Issues n

What are appropriate performance metrics?

n

What data should be used to estimate the value of each performance metric?

n

How should data be collected? From which sources? At what frequency?

n

How should data be analyzed to detect current and future deviations?

n

How should results of the analysis be reported? To whom? How often?


Controlling Project Risks Key issues to control risk during projecct: (1) what is optimal review frequency, and (2) what are appropriate review acceptance levels at each stage? “Both over-managed and under-managed development processes result in lengthy design lead time and high development costs.” Ahmadi & Wang. “Managing Development Risk in Product Design Processes”, 1999


Project Control & System Variation Common cause variation: “in-control” or normal variation Special cause variation: variation caused by forces that are outside of the system According to Deming: • Treating common cause variation as if it were special cause variation is called “tampering” • Tampering always degrades the performance of a system


Control System Example #1 n

Project plan: We estimate that a task will take 4 weeks and require n

1600 worker-hours

At the end of Week 1, 420 worker-hours have been used

Is the task “out of control�?


Control System Example (cont’d) Week 2: Task expenses = 460 worker-hours Week 1 2

Planned Cost (BCWS) 400 400

Actual Cost 420 460

Cumulative Actual Cost (ACWP) 420 880

470

Cost (in worker-hours)

460 450 440 430 420 410 400 390 380 370 1

2

3

4

Week

Is the task “out of control”?


Control System Example (cont’d) Week 3: Task expenses = 500 worker-hrs Week

Planned cost (worker-hours)

Actual cost (worker-hours)

Cumulative cost (worker-hours)

1 2 3

400 400 400

420 460 500

420 880 1380

600

Worker-hours

500 400 300 200 100 0 1

2

3

4

Week

Is the task “out of control”?


Earned Value Analysis • Integrates cost, schedule, and work performed • Based on three metrics that are used as the basic building blocks:

BCWS: Budgeted cost of work scheduled ACWP: Actual cost of work performed BCWP: Budgeted cost of work performed


Schedule Variance (SV) Schedule Variance (SV) = difference between value of work completed and value of scheduled work

Schedule Variance (SV) = Earned Value - Planned Value = BCWP - BCWS


Cost Variance (CV) Cost Variance (CV) = difference between value of work completed and actual expenditures

Cost Variance (CV) = Earned Value - Actual Cost = BCWP - ACWP


Earned Values Metrics Illustrated

Worker-Hours

Present time

Planned Value (BCWS)

Actual Cost (ACWP)

BAC

Cost Variance (CV)

Earned Value (BCWP)

Schedule Variance (SV)

Week 1

Week 2

Week 3

Week 4

Week 5

Week 6


Relative Measure: Schedule Index Schedule Index

(SI ) =

BCWP BCWS

If SI = 1,

then task is on schedule

If SI > 1,

then task is ahead of schedule

If SI < 1,

then task is behind schedule


Relative Measure: Cost Index Cost Index (CI) =

BCWP ACWP

If CI = 1,

then work completed equals payments (actual expenditures)

If CI > 1,

then work completed is ahead of payments

If CI < 1,

then work completed is behind payments (cost overrun)


Example #2 W E E K 1

2

3

4

5

6

7

8

9

10

Task A (36 worker-hrs) 6

6

6

8

10

Task B (36 worker-hrs) 12

12

12 Task C (56 worker-hrs) 10

10

12

12

12

Weekly Scheduled Worker-Hrs Cumulative

6

6

6

20

22

22

10

12

12

12

Scheduled Worker-Hrs (BCWS)

6

12

18

38

60

82

92

104

116

128


Example #2 (cont’d) Progress report at the end of week #5: Cumulative Percent of Work Completed: Week

1

2

3

4

5

Task A Task B Task C

15%

30%

40%

60%

80%

25%

65%

Not started yet

Worker-Hours Charged to Project: Week

1

2

3

4

5

Task A Task B Task C

5

6

8

10 15

10 10

Not started yet


Example #2 (cont’d) Progress report at the end of week #5: W E E K 1

2

3

4

5

6

7

8

9

10

6

12

18

38

60

82

92

104

116

128

5

11

19

44

64

Earned Value (BCWP)

5.4

10.8

14.4

30.6

52.2

Schedule Variance (SV)

-0.6

-1.2

-3.6

-7.4

-7.8

Cost Variance (CV)

0.4

-0.2

-4.6

-13.4

-11.8

Cumulative Scheduled Worker-Hrs (BCWS) Actual WorkerHrs Used (ACWP)


Example #2 (cont’d) 140 BAC 120

BCWS

Performance Metric

100

80 Cost Variance

S chedule Variance

60 ACWP 40 BCWP 20

0 1

2

3

4

5

6 Week

7

8

9

10


Using a Fixed 20/80 Rule Cumulative Percent of Work Completed: Week

1

2

3

4

Task A Task B Task C

20%

20%

20%

5

20% 20% Not started yet

20% 20%

W E E K Cumulative S cheduled Worker-Hrs (BCWS ) Actual WorkerHrs Used (ACWP) Earned Value (BCWP) S chedule Variance (S V) Cost Variance (CV)

1

2

3

4

5

6

7

8

9

10

6

12

18

38

60

82

92

104

116

128

5

11

19

44

64

7.2

7.2

7.2

14.4

14.4

1.2

-4.8

-10.8

-23.6

-45.6

2.2

-3.8

-11.8

-29.6

-49.6


Using a Fixed 20/80 Rule 140

120

Cost (in Worke r-hours)

100

80

BCWS ACWP 60

40

BCWP 20

0 1

2

3

4

5

6 Week

7

8

9

10


Updating Forecasts: Pessimistic Viewpoint Assumes that rate of cost overrun will continue for life of project‌.

Estimate at Completion

(EAC) = ACWP BAC = 1 BAC . BCWP CI

= (64/52.2) 128 = 1.23 x 128 = 156.94 worker-hrs


Updating Forecasts: Optimistic Viewpoint Assumes that cost overrun experienced to date will cease and no further cost overruns will be experienced for remainder of project life‌

Estimate at Completion

(EAC) = BAC - CV = 128 + 11.8 = 139.8 worker -hrs .


Multi-tasking with Multiple Projects How to prioritize your work when you have multiple projects and goals? Consider two projects with and without multi-tasking Project A

A-1

B-1

Project B

A-2

B-2

A-3

B-3

A-4

B-4


Due-Date Assignment with Dynamic Multiple Projects • Projects arrive dynamically (common situation for both manufacturing and service organizations) • How to set completion (promise) date for new projects? • Firms may have complete control over due-dates or only partial control (i.e., some due dates are set by external sources) • How to allocate resources among competing projects and tasks (so that due dates can be realized)? • What are appropriate metrics for evaluating various rules?


What Does the Research Tell Us? • Study by Dumond and Mabert* investigated four due date assignment rules and five scheduling heuristics • Simulated 250 projects that randomly arrive over 2000 days • average interarrival time = 8 days • 6 - 49 tasks per project (average = 24); 1 - 3 resource types • average critical path = 31.4 days (range from 8 to 78 days) • Performance criteria: 1) mean completion time 2) mean project lateness 3) standard deviation of lateness 4) total tardiness of all projects • Partial and complete control on setting due dates * Dumond, J. and V. Mabert. “Evaluating Project Scheduling and Due Date Assignment Procedures: An Experimental Analysis” Management Science, Vol 34, No 1 (1988), pp 101-118.


Experimental Results • No one scheduling heuristic performs best across all due date setting combinations • Mean completion times for all scheduling and due date rules not significantly different • FCFS scheduling rules increase total tardiness • SPT-related rules do not work well in PM (SASP) • Best to use more detailed information to establish due dates


Project Management Maturity Models • Methodologies to assess your organization’s current level of PM capabilities • Based on extensive empirical research that defines “best practice” database as well as plan for improving PM process • Process of improvement describes the PM process from “ineffective” to “optimized” • Also known as “Capability Maturity Models”


PM Maturity Model Example* 1)

Ad-Hoc

The project management process is described as disorganized, and occasionally even

chaotic. Systems and processes are not defined. Project success depends on individual effort. Chronic cost and schedule problems.

2)

Abbreviated: Some project management processes are established to track cost, schedule, and performance. Underlying disciplines, however, are not well understood or consistently followed. Project success is largely unpredictable and cost and schedule problems are the norm.

3)

Organized: Project management processes and systems are documented, standardized, and integrated into an end-to-end process for the company. Project success is more predictable. Cost and schedule performance is improved.

4) Managed: Detailed measures of the effectiveness of project management are collected and used by management. The process is understood and controlled. Project success is more uniform. Cost and schedule performance conforms to plan.

5) Adaptive:

Continuous improvement of the project management process is enabled by feedback

from the process and from piloting innovative ideas and technologies. Project success is the norm. Cost and schedule performance is continuously improving. * source: The Project Management Institute PM Network (July, 1997), Micro Frame Technologies, Inc. and Project Management Technologies, Inc. (http://pm32.hypermart.net/)


Turn static files into dynamic content formats.

Create a flipbook
Issuu converts static files into: digital portfolios, online yearbooks, online catalogs, digital photo albums and more. Sign up and create your flipbook.