Schedule Management Techniques For Complex Projects

1
Schedule Management Techniques For Complex
Projects

• W. Scott Nainis
• Noblis, Inc.
• August 12, 2009

2
Todays Agenda

• Motivation for the topic
• Why do many projects get behind (and cost more)?
• How do we track project progress?
• Role of Earned Value
• Transition to Earned Schedule
• How can we be both Pessimistic and Optimistic at
  the Same Time?
• How can Event Chains (and similar simulation
  approaches) help us?
• Power of Synergy
• How can ES and Event Chain work together?
• How is the schedule management article related to
• other articles within the SIGMA PMO Edition?

3
Motivation for the Topic

• Historic value of quantitative methods for
  project management role of management
  science/operations research (MS/OR)
• PERT (Program Evaluation and Review Technique)
• CPM (Critical Path Method)
• Network and Optimization (linear programming,
  dynamic programming, simulation, etc.)
• What has MS/OR done for project management
  lately?
• Project management tools (e.g. MS Project) have
  incorporated many of the MS/OR quantitative
  methods
• Simulation (Monte Carlo analysis,
  simulation-based training, what-if analysis) has
  been an active area for development

4
Motivation for the Topic (Concluded)

• What is still one of the biggest problem areas in
  project management project schedule
• Projects come in very late or never (61
  IT projects fail / 78 are late or over budget)
• Project costs and project quality often suffers
• What techniques and approaches can support
  project schedule management?

5
Why do Many Projects get Behind (and cost more)?

• Overly optimistic project schedules
• Human nature
• Political pressures
• Lack of effective responses to project problems
  as they occur
• Need to anticipate
• Time and cost to implement

6
Why are We Overly Optimistic in Project
Estimation?

• Human nature tends to overestimate achievement
  and tends to forget negative outcomes
• Daniel Kahneman and Amos Tversky performed
  research into the psychological underpinning of
  such biases (Kahneman received the Nobel Prize in
  2002 partially for these theories)
• Research has shown that people estimate
  overly-optimistically even in spite of contrary
  evidence
• Political forces apply pressure for optimistic
  forecasts even if planners are aware of the risks
  and less optimistic
• Pressure from supervisors and peers
• Decision-making forces optimistic forecasting
  (e.g. competitive contracts)

7
Over-Optimism and Political Pressure Lead to
Unrealistic Project Schedules

• Project Managers take the optimistic, shortest
  estimate
• Project issues during execution are ignored
• Lengthen planned schedule
• Raise costs and lower cost-benefit assessment
• Raise issues that need to be resolved
• Not prepared ahead of time for many contingencies
• Dont Forget Plain Old Incompetence

8
Example Bostons Big Dig

• Boston wanted to submerge the Central Artery-
  an elevated highway that bifurcated the city for
  nearly 50 years.
• Serious planning started around 1980
• By 1985 the estimate for the work was
• Project length 10 years
• Project cost 2.8 Billion dollars
• Work concluded December 31st 2007
• Project length 22 years
• Project cost 14.6 Billion plus about
• 7 billion in interest for a total
• of nearly 22 billion
• Still not done, definitely not not the
  litigation!

9
Alternative Methods for Project Forecasting

• Concept of insider forecasting versus
  outsider forecasting
• Developed in 2006 to the concept of Reference
  Class Forecasts
• Use of real data from similar projects
• Become aware of what can actually go wrong with
  complex projects
• Take into account the distributional nature of
  project activities, impacts and results
• Allow for input and appraisal from those who are
  not too close to the project

10
Alternative Methods for Project Forecasting
(concluded)

• Parametric Software Project Cost and Schedule
  Estimating Techniques
• COCOMO II, CoStar, Cost Modeler, CostXpert,
  Knowledge Plan, PRICE S, SEER, SLIM, and SoftCost
• The above methods have aspects of being
  reference-based approaches
• How good is the data? Will it be used fairly?
• Heuristic Task-based versus Time/Support-based
  estimation Collective Wisdom
• Use of simulation-based project management tools

11
Heuristic Scheduling Example

• Small Project budget estimation
• Simple Data Analysis and Reporting Project of
  Four tasks
• Task-based Approach
• Task 1 Develop Data Collection Plan (Staff A and
  B - 40 hours each, Staff C 10 hours)
• Task 2 Collect Data (Staff B, D, and E - 80
  hours each)
• Task 3 Analyze Data (Staff A and B - 80 hours
  each)
• Task 4 Produce Results Presentation Report and
  Deliver Report (Staff A and B - 60 hours each,
  Staff C - 15 hours)
• Total Staff Hours 625 hours 10 contingency
  690 hours
• Placing Tasks End-to-End would result in 2.5
  month schedule, rounded up to 3 months.

Staff A Main Investigator B Right-hand
support C - Oversight Manger D - Date Collector E
Data Collector
12
Heuristic Scheduling Example (Concluded)

• Time/Support-based Approach
• Experience tells us this is no less than a four
  month project
• Staff A is the project leader day-to-day 70 of
  time required
• Staff B is the other main on-going support person
  50 of time required
• Staff C is the oversight senior manager 10 of
  time required
• Staff members D and E are focused on data
  collection 50 of time required over a 1.5
  month window
• Assume 158 hours available per staff per average
  month
• Allocation Staff A 440 hours, Staff B 320
  hours, Staff C 60 hours, Staff D and E 120
  hours each total 1,060 hours.
• About 50 greater hours than the Task-based
  approach, 33 -38 longer schedule

13
How do We Track Project Progress?

• Start with a base-line project schedule
• Project subtasks and milestones completed
• Keep track of project expenditures compared to
  project budgets and credit for task completed
• Keep track of change control status and map back
  to current schedule estimates may not be that
  apparent

14
Role of Earned Value Management

• Earned Value Management (EVM) has developed over
  the years as an important approach to management
  of both project budget and schedule
• Track project for budgeted versus actual
  expenditures
• Use the metrics from project financial measures
  to track project progress
• Required by OMB for most software projects
• OMB Circular A-11, Part 7 (ANSI/EIA Standard 748)
  7
• Time is typically not an explicitly tracked
  quantity

15
Earned Value and Schedule Performance
Earned Value (EV) 48 at week 10
16
Earned Value and Schedule Performance (Continued)
Earned Value and Cost Performance
CV 48 79 -31
CPI 48/79 0.61
CV 48 48 0
CPI 48/48 1.00
SV and SPI still as before.
17
Earned Value and Schedule Performance (Continued)
18
Earned Value and Schedule Performance (Continued)
SV reaches 45 and then goes to 0 at the end of
the project
19
Earned Value and Schedule Performance (Concluded)
20
Earned Value and Schedule Performance (Concluded)
Schedule Delay
SPI reaches a low of 0.72 but then tends back to
1.0 as the project completes 7.5 months late!
21
What is Earned Schedule?

• Simple, but elegant concept
• Uses EVM data to produce a more useful index of
  project schedule status
• Devised in 2003 by Walter Lipke, software project
  manager who has pioneered the use of EVM for
  software development project management
• Empirical studies found Earned Schedule (ES) to
  be a superior predictor of project schedule and
  completion
• www.earnedschedule.com

22
Calculating Earned Schedule (ES)
ES 7 (first 7 weeks of schedule progress)
Portion of week seven accomplished
(48-45)/(54-45)0.33 7.33 weeks
23
How Can We be Both Pessimistic and Optimistic at
the Same Time?

• Monte Carlo simulation analysis allows us to
  consider reference class forecasting
• Distributional impacts on activities duration and
  cost
• Takes into account the interaction of project
  activity events
• Leads to longer, more costly and pessimistic
  forecasts
• Need a way to counter-balance the pessimistic
  trends with Monte Carlo simulation
• Consider risk moderation responses
• What if? responses considered ahead of time

24
Basically We Need to Establish a Risk Analysis
Exercise During Project Planning and Continue It
During Project Execution
Source Jane Powanda, Noblis, Inc.
25
How can the Event Chain Method help us?

• An external event can occurs which impacts the
  status of one or more project activities
• In response to the first event subsequent events
  are triggered to respond to the effects of the
  first event
• Event Chains are established and simulation
  software is used to track and manage all the
  events across the project activities
• Interventions included in response events attempt
  to modify and manage the inherent risk to the
  project

26
Project Activities Can be Linked Through an Event
Chain

• Events can be external and
• autonomous a Triggering Event
• Event can be in response to a
• Triggering event

Excited State
27
Event Chains Can Initiate Mitigation Plans
Example - Trigger Event Machine tools found to
be out of specification, yielding lower quality
output and lower throughput.
Triggered Event Response Machine tools
inspected, recalibrated and repaired/replaced if
necessary.
28
Features Useful to Support Event Chain Method
Wish List

• Provide classic project management scheduling
  reporting and resource management capabilities
• Incorporate and interface with major project
  management scheduling software (e.g. MS Project,
  Primavera, etc.)
• Handle development and management of event chains
• Allow for interaction of triggering events and
  responsive events impacting one or more project
  activities and their associated resources
• Be capable of supporting Monte Carlo Analysis and
  statistical results reporting
• Support project resource utilization and activity
  completion accounting
• Support EVM maintenance
• Allow for project branching due to event
  occurrence
• Allow for re-baselining and maintenance of all
  project accounts for each baseline

29
Possible Software Candidates for Supporting Event
Chain Method

• Microsoft Project
• Standard for many users
• Does project scheduling and tracks activities and
  resources
• Supports critical path determination
• Does not support statistical simulation/Monte
  Carlo analysis directly
• _at_ Risk for Microsoft Project
• Add-on to Microsoft Project
• Performs simulation/Monte Carlo analysis to
  obtain distribution impact of project and
  resource variability
• Does not handle event chain methods
• Primavera Risk Analysis
• Works with Primavera PM Software
• Performs a fully capable risk analysis along with
  project scheduling and other PM functions
• Fully capable statistical simulation / Monte
  Carlo analysis with incorporated schedule
  analytics
• Full reporting with statistical information and
  all project financial assessment measures
• Works with Primavera EVM module
• Event chain methods can be formulated

30
Possible Software Candidates for Supporting Event
Chain Method (Concluded)

• ProChain
• Designed to work with MS Project and replace the
  MS Project scheduler
• Performs analysis to determine critical chain
  situations which are similar to event chains
  (Goldratt)
• No statistical simulation/ Monte Carlo analysis
• Risky Project
• Can be used stand-alone as a project management
  planning tool
• Can be used and interface with MS Project,
  Primavera and other PM software packages
• Designed for event chain modeling
• Supports statistical simulation/ Monte Carlo
  analysis
• Performs detailed resource and activity
  accounting and support EVM calculations

31
Power of Synergy How can Earned Schedule and
Event Chain Work Together?
Step 1. The project team develops the work
breakdown Structure (WBS) and lays out project
plan with resources and durations. EVM
accounting is put in place along with ES.
Step 2. A second team or sub-team group takes
plan and introduces risk elements to activities.
Identifies negative impact areas. Both teams
consider response events to mitigate or avoid
risk effects.
Step 3. Both teams work to develop an event chain
structure incorporating all information known to
date. New plan with incorporated event chains is
run to finalize the project schedule and costs.
Step 4. Original Project team continues to
monitor and manage project execution. Implements
planned event responses as necessary. Can
deviate and modify the event chains as events
unfold. Completed activities are documented.
32
Power of Synergy Schedule Management and the
other SIGMA PMO Articles

• Toward Best-Practice Management
• by Robert G. Vorthman, Jr.
• Many methods, templates and practices in PM are
  mentioned
• Some relate to schedule management, particularly
  risk analysis
• Monte Carlo and simulation cited as less useful,
  but what
• does this information from Bresner and Hobbs
  mean?

• The Modern Program Office New Goals,
• New Organization
• by Michael D. Nelson and Shawn J. Margolis
• 61 IT projects fail / 78 are late or over
  budget
• Project leadership differs from project
  management
• Schedule management needs both
• The PMO can be the source of expertise and
  knowledge to
• support improved schedule management
  approaches

• Toolkit for Federal Information Technology
• Project Managers
• by Brian H. Price and David W. Vera
• Devised an integrated approach to financial
  management/
• investment control and the SDLC
• Linkage to proper IT support roles
• Schedule management approaches must be
  consistent
• with financial and resource requirements

• Managing Mutiple Information Technlogy Projects
• Lessons Learned
• by Daphne B. Byron and Chip Steiner
• Project tracking knowledge and response
  essential
• Must understand how change control impacts
  schedule
• Existing EVM schedule indices not as useful,
  suggests ES

• The Case for Agile Management
• by John E. Freeman
• Plan-driven PMO may not be responsive enough
• Agile PM looks for internal initiative and
  controls, and
• flexible responses
• Schedule management can take advantage of
  pre-planned
• knowledge, yet be responsive to continuous
• learning and adjustment point of operation
  mid-way between
• agile management and the plan-driven PMO

• Using Six Sigma in Project Forensics
• by John K. Stevenson and Frederick W. James
• Looking for project defects after the fact
• Uses DMAIC framework
• Found project forecasting and unrealistic
  project schedule
• defects
• Also found project experience and requirements
• development defects

Design-Measure-Analyze-Identify-Control
33
The Reality of Project Management Practice
Besner, C. and Hobbs, B. (2004), University of
Quebec
34
The Reality of Project Management Practice
Besner, C. and Hobbs, B. (2004), University of
Quebec