Cocomo Lecture I EEE493 2000

1
Cocomo - Lecture I EEE493 2000
Royal Military College of Canada Electrical and
Computer Engineering

• Major Greg Phillips
• greg.phillips_at_rmc.ca
• 1-613-541-6000 ext. 6190

Dr. Scott Knight knight-s_at_rmc.ca 1-613-541-6000
ext. 6190
2
Refs

• Pressman, R.S., Software Engineering a
  Practitioners Approach 5th Ed., McGraw-Hill,
  2001, Chapter 4
• Boehm, Barry, Software Engineering Economics,
  Prentice-Hall, 1981

3
Teaching Points

• Basic COCOMO assumptions
• Basic COCOMO formula
• Phase distribution
• Maintenance Effort
• Modes of development

4
Review

• What are Adjusted Function Points?
• It is the case that the adjusted function point
  total no longer reflects the actual size of the
  system. Explain this.
• What is Backfiring?
• What is a major weakness of function points?

5
Recall Algorithmic Models

• Algorithmic models provide one or more
  mathematical algorithms which produce an estimate
  a a function of a number of variables considered
  to be major cost drivers. ?(x1, x2,... xn)
• Cost Drivers
• lines of source code
• programmer capability
• schedule constraints
• execution time constraints
• etc.

6
The COnstructive COst MOdel(COCOMO)

• Created by Barry Boehm
• Circa 1976
• There was no firm basis for explaining to a
  manager, customer, etc. that a proposal is
  unrealistic
• It was hard to make S/W - H/W trade-off analysis
• It was difficult to tell if development was
  proceeding I.A.W. plan

7
Basic COCOMO - Assumptions

• Primary cost driver is DSI - delivered source
  instructions
• The period covered is from the beginning of
  product design phase to the end of the
  integration and test phase
• Cost estimates cover only a defined set of
  activities (e.g. user training is not covered)
• COCOMO covers all direct-charged labour (e.g.
  analysts, project-managers, program librarians,
  designers, programers) but excludes indirect
  labour (e.g. secretaries, janitors,
  higher-management, etc.)

8
Basic COCOMO - Assumptions contd

• A COCOMO person-month
• 152 hours
• 19 person-days
• Assumes customer and developer enjoy good
  management
• Assumes Requirement spec. is not substantially
  changed
• Basic Intermediate COCOMO have phase
  independent cost drivers
• The phase costs include all costs incurred during
  the phase (I.e. the phases are temporal chunks of
  a project and not activity groups)

9
Basic COCOMO Formula

• PM 2.4(KDSI)1.05 person-months
• TDEV 2.5(PM)0.38 months
• This model is for small-to-medium size projects
  developed in a familiar, in-house, organic
  software development environment

10
Implications of the Model
11
Example

• A medium size project
• 32 KDSI
• PM 2.4(32)1.05 91 pm
• Productivity 32/91 352 DSI/pm
• TDEV 2.5(91)0.38 14 months
• Average Staff 91/14 6.5 fsp
• pm (person months)
• fsp(full time s/w personnel)

12
Phase Distribution
13
Interpolation

• Use linear interpolation for intermediate values
  of phase distribution effort and schedule
• Example
• a 12.8 KDSI project
• 20 between 8 KDSI and 32 KDSI
• therefore programming effort
• (62 -65)(.2) 65 64.4

14
Example

• For the previous project example
• Programming Phase
• effort (.62)(91 pm) 56 pm
• sched (.55)(14 months) 7.7 months
• 56pm/7.7months 7.3 fsp
• Note
• large projects spend less effort (time) on
  programming
• all have roughly the same design numbers
• large projects spend more effort (time) on
  integration test
• large projects have a higher labour distribution
  curve peak (Rayleigh curve)(easier to see on
  profiles slide)

15
Basic Project Profiles
16
Maintenance Effort Estimation

• Annual Change traffic (ACT)
• ACT DSI-changed/year DSI added/year Initial
  DSI
• PMAM 1.0(ACT)PMD
• FSPAM PMAM/12 months

17
Consider

• All software is not developed under the same set
  of circumstances
• What about developing
• new kinds of systems
• in new development environments
• on hardware being developed at the same time
• using new algorithms
• in an embedded environment
• under contract
• Introduce 3 Modes of Development

18
Organic Mode

• As in previous discussions, small-to-medium sized
  product development in a familiar, in-house
  development environment. A generally stable
  development environment, minimal need for
  innovative data processing architectures and
  algorithms. Relaxed schedule constraints.

19
Semidetached Mode

• Team members have an intermediate level of
  experience with related systems. Perhaps a
  mixture of experienced and inexperienced people.
  Parts of the project may require rigorous
  interfaces.

20
Embedded Mode

• Need to operate within tight constraints.
  Product must operate within a strongly coupled
  complex of H/W, S/W, regulations, and operational
  procedures. Cost of changing other parts of the
  complex is high. Tighter requirements and more
  inflexible scheduling.

21
Next ClassCocomo - Lecture II