Towards Automated Software Project Planning Extending Palladio for - - PowerPoint PPT Presentation

towards automated software project planning
SMART_READER_LITE
LIVE PREVIEW

Towards Automated Software Project Planning Extending Palladio for - - PowerPoint PPT Presentation

KPD Symposium 2013 Position Paper Towards Automated Software Project Planning Extending Palladio for the Simulation of Software Processes Oliver Hummel & Robert Heinrich sdq.ipd.kit.edu SOFTWARE DESIGN AND QUALITY GROUP INSTITUTE FOR


slide-1
SLIDE 1

KIT – University of the State of Baden-Wuerttemberg and National Research Center of the Helmholtz Association SOFTWARE DESIGN AND QUALITY GROUP INSTITUTE FOR PROGRAM STRUCTURES AND DATA ORGANIZATION, FACULTY OF INFORMATICS

www.kit.edu sdq.ipd.kit.edu

Towards Automated Software Project Planning

Extending Palladio for the Simulation of Software Processes

Oliver Hummel & Robert Heinrich

KPD Symposium 2013 Position Paper

slide-2
SLIDE 2

Software Design and Quality Group Institute for Program Structures and Data Organization 2 / 11 2013-11-27 Towards Automated Software Project Planning Oliver Hummel

Motivation

FISCUS Inpol-neu “It is very difficult to make a vigorous, plausible, and job-risking defense of an estimate that is derived by no quantitative method, supported by little data, and certified chiefly by the hunches of the managers.”

Fred Brooks

[Wikipedia]

slide-3
SLIDE 3

Software Design and Quality Group Institute for Program Structures and Data Organization 3 / 11 2013-11-27 Towards Automated Software Project Planning Oliver Hummel

Overview

  • 1. Problem: Software project planning is difficult
  • 2. Idea: Simulate the development process
  • 3. Benefit
  • 4. Previous work
  • 5. Proposed approach and necessary next steps
slide-4
SLIDE 4

Software Design and Quality Group Institute for Program Structures and Data Organization 4 / 11 2013-11-27 Towards Automated Software Project Planning Oliver Hummel

Problem

Software projects are still planned like some twenty years ago

using relatively primitive techniques like bar charts or activity networks

Effort is estimated with the help of parametric models such as COCOMO

based on relatively coarse-grained parameters and assumptions

No possibility to model concrete

tasks, roles and their dependencies

[Sommerville]

Effort = A * SizeE * C

slide-5
SLIDE 5

Software Design and Quality Group Institute for Program Structures and Data Organization 5 / 11 2013-11-27 Towards Automated Software Project Planning Oliver Hummel

Idea

A look over the fence of software engineering reveals that other disciplines simulate their production processes

Can‘t we simulate software development processes, too?

[Tecnomatix]

slide-6
SLIDE 6

Software Design and Quality Group Institute for Program Structures and Data Organization 6 / 11 2013-11-27 Towards Automated Software Project Planning Oliver Hummel

Expected Benefit

Simulating the software development process will allow to –

derive better project plans (automatically)

thus, helps to educate project managers and allows a better project controlling

better understand the dependencies within a project

and hence reduce the risks of overlooking them

simulate various project flows in terms of –

selected development process e.g. agile vs. waterfall personell where and when does it make sense to add new staff

In a nutshell: it will help to minimize planning risks

slide-7
SLIDE 7

Software Design and Quality Group Institute for Program Structures and Data Organization 7 / 11 2013-11-27 Towards Automated Software Project Planning Oliver Hummel

Previous Work

… consists mainly of estimation tools that –

are based on COCOMO

  • r similar models

derive probability distributions for schedule and/or effort

activities and their dependencies

can‘t be modelled

The group of Ludewig (U Stuttgart) has developed a project management simulation for software projects (SESAM)

requires several hours of playing time

To our knowledge: no tool that allows e.g. the comparison of different process models

  • r the effects of a delayed activity
slide-8
SLIDE 8

Software Design and Quality Group Institute for Program Structures and Data Organization 8 / 11 2013-11-27 Towards Automated Software Project Planning Oliver Hummel

Proposed Approach

Based on a recent extension of PCM…

allowing to represent human actors and activities in a business process

…the ambitious long-term vision is a tool that –

creates an optimized project plan for a given set of requirements automatically

based upon a selected process model and available resources

allow to manipulate individual tasks and assignments

in order to see their influence on the project outcome

…the medium-term approach aims on –

1.

using (an extended) KAMP to derive a work breakdown structure (WBS) for an architectural model

2.

deriving a project plan and a Palladio model for the WBS

  • based on some desired process model

3.

simulating the derived model(s) in order to compare them

slide-9
SLIDE 9

Software Design and Quality Group Institute for Program Structures and Data Organization 9 / 11 2013-11-27 Towards Automated Software Project Planning Oliver Hummel

Necessary Extensions

1.

Extend the PCM+BPM with missing concepts –

Artefacts resp. Documents

in the business process model

Task Types (as identified by F. Brooks)

extend simulation behaviour with appropriate traversal strategies

Teams

add a layer to the organization environment model

Deadlines

add an automatic detection of deadline violations

2.

Extend and adapt KAMP

in order to allow the creation of a full WBS

3.

Come up with a user-friendly user interface

including easy to use templates for SE process models

slide-10
SLIDE 10

Software Design and Quality Group Institute for Program Structures and Data Organization 10 / 11 2013-11-27 Towards Automated Software Project Planning Oliver Hummel

Summary

slide-11
SLIDE 11

Software Design and Quality Group Institute for Program Structures and Data Organization 11 / 11 2013-11-27 Towards Automated Software Project Planning Oliver Hummel

Fin

Thank you for your attention! Time for some questions…

… or maybe you want to contact me later