Scenario Oriented Project Management Knowledge Reuse within a Risk Analysis Process

1
Scenario Oriented Project Management Knowledge
Reuse within a Risk Analysis Process

• Márcio de Oliveira Barros
• Cláudia Maria Lima Werner
• Guilherme Horta Travassos
• COPPE / UFRJ

2
Software Project Management

• Complex software development is a risk job
• Too many software projects are concluded costly
  and behind schedule than planned, specially large
  projects
• Project Management
• Project management is determinant to project
  success
• Most project management techniques that are used
  today date from the early 1950s
• There has been little advancements in PM
  techniques when compared to development techniques

3
Scenario Based PM

• Scenario analysis in project management
• Scenario based project management (SBPM) is a set
  of techniques that allow a manager to test a
  project sensibility to unexpected events and
  actions
• The manager defines project expect behavior, such
  as activities distribution through schedule,
  planned cost and quality (project model)
• The manager reuses and creates scenarios
  describing events, actions or policies that can
  affect the project (scenario models)
• Scenarios and project models are integrated to
  evaluate scenario impact upon the project

4
SBPM Architecture
builds
develops
Project Model
Scenarios
analyzes
Simulation Environment
5
Behavior Evaluation

• Models are described in system dynamics (SD)
• System dynamics is a modeling technique to
  describe complex systems, focusing on their
  structural aspects
• Models are mathematically formulated using
  extended system dynamics constructors
• Project models are evaluated through simulation,
  with or without integrated scenario models
• We are proposing some extension to SD
• Higher model representation level
• Element attributes representation
• Separating facts from assumptions

6
PM Knowledge Representation

• Scenarios hold project management knowledge
• Scenarios can be organized in an
  organization-wide knowledge base to project
  managers
• Scenarios allow formal documentation of
  assumptions and proven information about project
  management
• Senior project managers can create scenarios
  allowing less experienced managers to share their
  knowledge by retrieving and reusing such models
• During application development, the manager
  retrieves the scenarios that he believes can
  affect his project and evaluates their impact
  upon the project behavior

7
PM Knowledge Reuse

• Knowledge sharing resembles a reuse process
• Scenario models are the reusable artifact
• Two-staged development process
• Development for reuse and Development with
  reuse

8
PM Knowledge as Scenarios

• Advantages
• Scenarios are formal representations that can be
  directly evaluated upon a project model
• Due to system dynamics, scenarios capture
  non-linear relationships among project elements
• There is a framework for PM knowledge sharing and
  reuse (based on the generic reuse process)
• There is a technique for scenario evaluation and
  we are currently developing tools to support this
  technique
• Disadvantages
• Math is good for computer evaluation, but not for
  human analysis. Scenarios need complementary
  information to help selection and understanding

9
Risk Archetypes

• Standard information structure
• Describes a potential problem that can occur
  along several development projects
• The information structure documents the
  experience of forecasting, identifying and
  resolving such problems
• It wraps scenario models that are used to
  describe the impact of the problem and its
  resolution strategies
• It allows navigation among risk archetypes and
  has criteria for problem selection based on the
  application or development process
  characteristics

10
Development Sharing as Reuse

• Development for reuse
• Risk management for risk identification
• Identifying, modeling, and documenting risk
  archetypes
• Development with reuse
• Risk management for application development
• Selecting, understanding, and adapting risk
  archetypes
• Feedback
• Applications might need some archetypes that are
  not already developed, feeding the risk
  identification process with needs

11
Experimental Validation

• The proposed techniques are under evaluation
• Preliminary studies
• Observational study upon documenting the risks of
  an application domain using risk archetypes to
  (financial market risk management software)
• First major experiment
• Quasi-experiment to evaluate the application of
  the proposed techniques in software project
  management

12
Conclusions

• Scenarios are useful for PM knowledge
  representation
• Scenarios have higher representation capabilities
  than traditional PM models, such as PERT
• There is a framework to support scenario sharing
  and reuse
• Experimental results show promising results

13
Questions