Contents. Today Project Management. Project Management. Last Time - Software Development Processes. What is Project Management?

(1)

3/9 - 08 Programvarukonstruktion - Jonny Pettersson, UmU

Introduction

Software Development Processes

Project Management

Requirements Engineering

Software Construction

Group processes

Quality Assurance

Software Management and Evolution

Last Time - Software Development

Processes

What is Software Engineering?

Software Development Processes

What is that?

The Waterfall Model

The Spiral Model

The V-Model

Iterative and Incremental Development

Agile Development

CMM

Systematic Process Improvement

Today – Project Management

What is Project Management?

Cost Estimation

Project Planning and Scheduling

(2)

What is Project Management?

Planning

Evaluation

inp

ut t

o

Think first.

Check results.

produce

Activity

Result

improve

imp

ro

ve

Project Management Activities

Project acquisition

Feasibility study (initial project planning)

Resource assessment

Option (and risk) analysis

Cost estimation

Project planning and scheduling

Work breakdown structure

Effort estimation and distribution

Resource assignment

Project tracking and control

Risk management

Reporting

Project Resources

People

Required skills

Availability

Duration of tasks

Start date

Time

Budget

Office/lab space

…

Hardware and

software tools

Description

Availability

Duration of use

Delivery date

(3)

Option Analysis

Compare different development

alternatives

Evaluate their risks

Select best alternative

Tools

Polar graph

Forms

Decision tree

...

Polar Graphs

Portabilty Efficiency Schedule Reliability Cost Reuse Alternative A Alternative B Alternative C

Analysis Example

Alternative A B C Relative efficiency 0.8 1 0.6 Tools investment 250.000 500.000 500.000 Schedule (years) 2 1.3 2 Staff utilization (%) 85 70 100 Risk 0.75 0.6 0.9 Cost (SEK) 7.500.000 8.500.000 7.000.000

(4)

Analysis Example (Polar Graph)

Staff utilisation Schedule Efficiency Cost Tools investment Risk Alternative A Alternative B Alternative C 0.5 1 y 5 M 0 1 1 0 100% 250K 2 y 500K

Analysis Example (Forms)

Objectives Constraints Alternatives Risks Risk resolution strategy Results Plans Commitment

Develop a software components catalogue

Within one year

Must support all existing component types Must cost less than 1 MSEK

Buy existing information retrieval (IR) software

Buy a database and develop the catalogue using the query language Develop a special-purpose catalogue

May be impossible within the given constraints Catalogue functionality may be inappropriate Develop a prototype to clarify requirements Commission a consultants report on existing IR systems Relax the time constraints

IR systems are too inflexible Identified requirements cannot be met

The prototype using a DBMS may be enhanced to a complete system Special-purpose catalogue development is not cost effective Develop the catalogue using the existing DBMS by enhancing the prototype and building a GUI

Fund further 12 months of development

Analysis Example (Decision Tree)

Goal: Develop System X 3.8 MSEK 4.5 MSEK 2.75 MSEK 3.1 MSEK 4.9 MSEK 2.1 MSEK 4 MSEK 3.5 MSEK 5 MSEK Build Reuse Buy Contract Simple (0.3) Difficult (0.7) Minor changes (0.4) Major changes (0.6) Simple (0.2) Complex (0.8) Minor changes (0.3) Major changes (0.7) Without changes (0.6) With changes (0.4)

(5)

Project Management

What is Project Management?

Cost Estimation

Project Planning and Scheduling

Running and Documenting a Project

Cost Estimation

General approach

Decompose problem

Check for experiences/

data on sub problems

Make qualified estimations

(Make at least two

independent estimates)

General problems

What are good measures?

Do the estimates effect

the result?

Does the type of software

effect the result?

Does the project

environment effect the

result?

...

Î

Use empirical and historical data

Î

Algorithmic cost modelling

COCOMO (based on

LOC

)

FP (based on

function points

)

COCOMO

Constructive Cost Modelling [Boehm 81, Boehm 00]

Based on publicly available historical data of 63 TRW

projects

Basic assumptions:

Requirements change only slightly during the project

There is good project management

The historical data is representative

Assigning more resources to the project does NOT result in

linear decreasing development time

Basic model:

Effort = a KDSI

b

KDSI = Kilo Delivered Source Instructions (≈ LOC - comments) The a and b factors vary depending on the type of project Effort is measured in PM (Person Months = 152h of work)

(6)

COCOMO Project Types

OM: Organic Mode projects

Small teams which are familiar with the type of

application

Development in a familiar environment

EM: Embedded Mode projects

Large and inexperienced teams

Many constraints

SDM: Semi Detached Mode projects

Between OM- and EM projects

COCOMO Basic Model

PM: Person Months

2.4 KDSI

1.05

for OM projects

3 KDSI

1.12

for SDM projects

3.6 KDSI

1.20

for EM projects

TDEV: Time for DEVelopment

2.5 PM

0.38

for OM projects

2.5 PM

0.35

for SDM projects

2.5 PM

0.32

for EM projects

N: Number of personnel

PM / TDEV

This is Too Simplistic!?

There are many cost drivers that effect

effort

Programming language

Development methods

Tools and environments

Experience and capabilities of the development

team

Available time

Requirements volatility

(7)

COCOMO Intermediate Model

Takes into account 15 cost drivers, which are ranked on a

scale from very low to extra high

Product attributes (e.g., required reliability)

Platform attributes (e.g., time/space constraints)

Personnel attributes (e.g., programming language experience)

Project attributes (e.g., tool usage)

PM: Person Months

3.2 KDSI

1.05

× ΠCifor OM projects

3 KDSI

1.12

× ΠCi for SDM projects

2.8 KDSI

1.20

× ΠCi for EM projects

ΠCi ∈ [0.09..9.42]

TDEV and N as before

Cost Drivers

1.10 1.04 1.00 1.08 1.23 Required development schedule

0.83 0.91 1.00 1.10 1.24

Application of software engineering methods 1.24 1.10 1.00 0.91 0.82 Use of software tools

Project attributes 1.14 1.07 1.00 0.95 Programming language experience 1.21 1.10 1.00 0.90 Virtual machine experience

0.70 0.86 1.00 1.17 1.42

Software engineer capability 1.29 1.13 1.00 0.91 0.82 Applications experience 1.46 1.19 1.00 0.86 0.71 Analyst capability

Personnel attributes 0.87 1.00 1.07 1.15

Required turnabout time 0.87 1.00 1.15 1.30

Volatility of the virtual machine environment

1.56 1.21 1.06 1.00

Memory constraints 1.00 1.11 1.30 1.66

Run-time performance constraints Platform attributes 1.65 1.30 1.15 1.00 0.85 0.70

Complexity of the product 0.94 1.00 1.08 1.16 Size of application database

1.40 1.15 1.00 0.88 0.75 Required software reliability

Product attributes Extra High Very High High Nomin al Low Very Low Ratings Cost Drivers

Intermediate COCOMO Summary

Works quite well in practice

TRW data is publicly available

Needs KLOC as input

Problems:

Estimating KLOC in early project stages

Comparison of projects using different LOC counts

Outdated project data base for basic constants (70s)

Solutions:

Cross-check using an other estimation technique

Standardised LOC counts

Continuous model calibration

Use COCOMOII

(8)

COCOMO II

Stage 1: Application Composition (resolving high-risk issues)

Estimation base: Object points

Single standard project type

Stage 2: Early Design

(exploration of architectures and

concepts)

Estimation base: Function points or KLOC

Six project type factors

Stage 3: Post-architecture

(development has started)

Estimation base: Function points or KLOC

Six project type factors

Formulas more complicated; takes care of more variation

See simple

online tool

(

http://www.cms4site.ru/utility.php?utility=cocomoii

)

Function (Feature) Points

Estimate functionality captured in requirements

Very detailed and complex counting rules (IFPUG)

# User inputs x (3,4,6) = … # User outputs x (4,5,7) = … # User inquiries x (3,4,6) = … # Files x (7,10,15) = … # External interfaces x (5,7,10) = … (# Algorithms x (3,4,6) = … ) ---Count-total … Feature points only

FP = Count-total x [ 0.65 + 0.01 x

Σ

F

i

]

Adjustment factors (Fi ∈ {0,...,5}) i=1 14

Project Management

What is Project Management?

Cost Estimation

Project Planning and Scheduling

(9)

Activity

Major work unit

Start and end dates

Consumes resources

Results in work products

(and milestones)

Define Work Breakdown Structure

Project function

Continue throughout the

project

Not bound to specific

phases

Project Phase 1 Phase 2 Phase N Step 1 Step 2 Step 1 Step 2 Step 1 Step 2 Activity 1 Activity 2 Activity 3 Activity 1

Activity 2 Task 1_{Task 2}

Task 3

Schedule Activities

Almost all activities depend on the completion of

some other activities

Many activities can be performed in parallel

Track usage of resources

Organisation necessary to balance work-load,

costs, and duration

PERT chart (activity network/task graph)

Critical path

Project time-line (Gantt chart)

Resource table

PERT Charts

Program Evaluation and Review Technique

Graph

Nodes = activities/tasks and estimated duration

Edges = dependencies

Compute

Slack time = available time - estimated duration

Critical path

A path is critical when it contains an activity that,

if delayed, will cause a delay of the whole project.

(10)

PERT Chart “Syntax”

There is a nice tutorial at

http://www.getahead-direct.com/gwprpert.htm

An Example PERT Chart

Designelectrical and pneumatic renovations

4 4d

5/4/00 5/9/00

Schedule down-time with production supervisor

5 2d

5/10/00 5/11/00

Renovate electrical and pneumatic

13 2d

5/17/00 5/18/00

Get final report from PSU

1 5d

4/27/00 5/3/00

Schedule contractor for electrical and pneumatic

8 1d

5/12/00 5/12/00

Inform PSU about pickup

7 2d

5/5/00 5/8/00

Remove old hardware

11 2d

5/15/00 5/16/00

Shut down production line

10 1d

5/12/00 5/12/00

Install new hardware

14 3d

5/19/00 5/23/00

Schedule truck to pickup hardware at PSU

6 1d

5/4/00 5/4/00

Get blueprints for old hardware 2 1d 4/27/00 4/27/00 Time-motion study of existing hardware 3 2d 5/4/00 5/5/00

Store old hardware (just in case)

12 10d

5/17/00 5/30/00

Transport hardware from PSU

9 2d

5/9/00 5/10/00

(11)

Another Gantt Chart

A Resource Table

Can be (partly) generated from Gantt and or PERT

charts

Resource assignment

Useful for project tracking

Task name earliest latestStart earliest latestEnd ElapsedtimeResource 1Utilisationtime 1 Resource nUtilisationtime n

...

Project Management

What is Project Management?

Cost Estimation

Project Planning and Scheduling

(12)

Running a Project

To know where you are you need

A (detailed) plan/goal

Measurements of project status

time in, e.g., weeks

progress

100% plan

actual

Example Effort Distribution

0 50 100 150 200 250 300 350 400 Lecture s Meeting s Plannin g Require ments GUI de sign Analysi s DesignTestingCoding

Presentatio ns Subcon tract Code su bc. Evaluat ion of proto

type_Manual Final re port weekly reports and me eting m inutesOther 0 50 100 150 200 250

week35 week36 week37 week38 week39 week40 week41 week42 week43 week44

to tal ti me in h actual plan

Risk Management

Risk management is concerned with

identifying risks and drawing up plans to

minimise their effect on a project

A risk is a probability that some adverse

circumstance will occur

Project risks affect schedule or resources

Product risks affect the quality or performance

of the software being developed

Business risks affect the organisation

(13)

Risk Management Activities

Investigate potential risk factors

Risk

Likelihood of occurrence

Impact

Define mitigation strategies (i.e. be prepared)

What can be done to avoid the problem(s) in the first

place

What can be done to solve the problem(s) in case they

occur

Monitor risks

Determine if predicted risk occurs

Properly apply risk aversion steps

Collect info for future risk analysis

Example

Assume

Risk = High staff turnover

Likelihood of occurrence = 70%

Impact = Increase project time by 15%, project cost by

12%

Mitigation strategy

Identify high turnover causes

Reduce causes before project starts

Develop techniques to assure work continuity in light of

turnover

• For example ...

• or ...

Risk Management

Risk management Risk assessment Risk control Risk identification Risk analysis Risk prioritisation Checklist Decomposition Assumption analysis Decision driver analysis System dynamics Performance models Cost models Network analysis Decision analysis Quality risk factor analysis Risk exposure Compound risk reduction Buying information Risk avoidance Risk transfer Risk reduction leverage Development process Risk element planning Risk plan integration Risk mitigation Risk monitoring and reporting Risk reassessment

Contents. Today Project Management. Project Management. Last Time - Software Development Processes. What is Project Management?

Contents





Introduction

Introduction





Software Development Processes

Software Development Processes



Project Management



Requirements Engineering



Software Construction



Group processes



Quality Assurance



Software Management and Evolution

Last Time - Software Development

Processes



What is Software Engineering?



Software Development Processes

What is that?

The Waterfall Model

The Spiral Model

The V-Model

Iterative and Incremental Development

Agile Development

CMM

Systematic Process Improvement

Today – Project Management



What is Project Management?



Cost Estimation



Project Planning and Scheduling

What is Project Management?

Planning

Evaluation

inp

ut t

o

Think first.

Check results.

produce

Activity

Result

improve

imp

ro

ve

Project Management Activities



Project acquisition



Feasibility study (initial project planning)

Resource assessment

Option (and risk) analysis



Cost estimation



Project planning and scheduling

Work breakdown structure

Effort estimation and distribution

Resource assignment



Project tracking and control



Risk management



Reporting

Project Resources

