Presentation from September 9, 2008 Dinner Meeting

8/9/2019 Presentation from September 9, 2008 Dinner Meeting

1/34

Art PysterDistinguished Research Professor

Stevens Institute of Technology and

Co-Director, Applied Systems Thinking Institute

September 9, 2008

[email protected]

When Systems and Software Engineering

Education Collide: Integrating the Education of

Software and Systems Engineers


2/34

2

Agenda

How the world has changed

The current state of software engineeringeducation

Bringing software and systemsengineering education together


3/34

3

A History Lesson - 1996

Resolved: Software Should Lead in SystemsEngineering

Jim Armstrong vs. Art Pyster

The systems engineering community has long debatedthe extent to which software disciplines, processes,and practitioners should influence systemsengineering. In August 1996, the authors held a livelydebate at a meeting of the Washington Metropolitan

Area Chapter of INCOSE on the proper role of softwareengineering within systems engineering. The particularissue debated was the proposal that software ideas,process, and people should be in the lead whenbuilding complex systems. Pyster favored that view

while Armstrong opposed it.


4/34

Software will be the center of

systems design.

Eberhardt Rechtin

4


5/34

Twenty years from now, softwarepeople will be sitting at the table

and the other disciplines will besitting around the sides of theroom.

Eberhardt Rechtin

5


6/34

Demand for software in complex

systems is growing exponentially

0

2

4

6

8

10

12

14

16

18

1960 1970 1980 1990 2000 2010 2020

Sea Systems Missiles/Space Ground Systems Aircraft

Patriot

PAC-3

DDX

Virginia SSN

FCS

Polaris A3Aegis System

Sources: CARD Data, SEI, CSIS Analysis

ESLOCinMill i

ons

F-35 Aircraft

and Ground

ACS

SBIRS

Software Content of Sample Major DoD Weapon Systems 1960 - 2020

F-22

*Software Industrial Base Phase I Study, CSIS, October 2006 6


7/34

7

The term software-intensive insoftware-intensive system is

redundant

*A system in which software represents thelargest segment in one or more of thefollowing criteria:

system development cost

system development risk

system functionality

development time

*DAU Glossary of Defense Acquisition Acronyms and Terms, 2005

Everything interesting is software-intensive.

Even bullets have software programmable munitions to optimizefuzing


8/34


9/34

9

What do we teach for a masters

degree in software engineering?

The last effort to create a reference curriculum forgraduate software engineering education was by

the SEI in the early 1990s.

There are, in effect, no current community-endorsed recommendations on what to teachsoftware engineers nothing that recognizes howthe world has changed.

Response: create a project to create a newreference curriculum in software engineering


10/34

10

The Integrated Software andSystems Engineering Curriculum

Project

Begun in May 2007 at Stevens Institute of Technology

Sponsored by DoD Director of Systems and Software

Engineering Three products planned:

1. A modern reference curriculum for a masters degree in software engineeringthat integrates an appropriate amount of systems engineering

2. A modern reference curriculum for a masters degree in systems engineeringthat integrates an appropriate amount of software engineering

3. A truly interdisciplinary degree that is neither systems nor software engineering it is both


11/34

11

1st Project Graduate Software

Engineering Reference Curriculum1. Understand the current state of SwE graduate education

(November 2007)

2. Create GSwERC 0.25 with a small team, suitable forlimited review (February 2008)

3. Publicize effort through conferences, papers, website,etc (continuous)

4. Obtain endorsement from INCOSE, NDIA, ACM, IEEE, andother professional organizations (continuous)

5. Create GSwERC 0.50 suitable for broad communityreview and early adoption (October 2008)

6. Create GSwERC 1.0 suitable for broad adoption (2009)


12/34

12

The evolving author team Rick Adcock, Cranfield University and INCOSE

Edward Alef, General Motors

Bruce Amato,Department of Defense

Mark Ardis,Rochester Institute of Technology

Larry Bernstein, Stevens Institute of Technology

Barry Boehm, University of Southern California

Pierre Bourque, Quebec University and SWEBOK

volunteer

John Bracket,Boston University

Murray Cantor,IBM

Lillian Cassel, Villanova and ACM volunteer

Robert Edson,ANSER

Richard Fairley, Colorado Technical University

Dennis Frailey,Raytheon & Southern MethodistUniversity

Gary Hafen,Lockheed Martin and NDIA

Thomas Hilburn,Embry-Riddle Aeronautical

University

Greg Hislop,Drexel University and IEEE

Computer Society participant

Dave Klappholz, Stevens Institute of Technology

Philippe Kruchten, University of British Columbia

Phil Laplante,Pennsylvania State University, Great

Valley

Qiaoyun (Liz) Li, Wuhan University, China James McDonald, Monmouth University

John McDermid, University of York, UK

Ernest McDuffie,National Coordination Office for

NITRD Bret Michael,Naval Postgraduate School

William Milam,Ford

Ken Nidiffer, Software Engineering Institute

Art Pyster, Stevens Institute of Technology

Doug Schmidt, Vanderbilt University

Mary Shaw, Carnegie Mellon University

Robert Suritis, IBM Richard Thayer, California State University at

Sacramento

Barrie Thompson, Sunderland University, UK

Richard Turner, Stevens Institute of Technology

Joseph Urban, Texas Technical University

Ricardo Valerdi, MIT & INCOSE

David Weiss,Avaya Mary Jane Wilshire, Colorado Technical University


13/34

13

Methodology to understandcurrent state of SwE education

Diverse set of universities with Masters programs in SWE

- Vary in size, geography, maturity, resources, target market,

- Focused on programs with degree in SWE or Computer Science with aSWE specialization - not degrees in information technology and relatedareas

Used Software Engineering Body of Knowledge (SWEBOK) as theprimary framework for SWE competencies

Collected data from school websites

- Degree, faculty size, student population, target market,

- Degree structure, individual course descriptions- Map between courses and SWEBOK

Validated data with one or more professors from each school

Analyzed for commonalities and uniqueness


14/34

14

Schools studied

1. Air Force Institute of Technology

2. Brandeis University

3. California State University Fullerton

4. California State University

Sacramento5. Carnegie Mellon University

6. Carnegie Mellon University West

7. DePaul University

8. Drexel University

9. Dublin City University (Ireland) *10. Embry-Riddle Aeronautical

University

11. George Mason University

12. James Madison University

13.Mercer University

14. Monmouth University

15. Naval Postgraduate School

16. Penn State University GreatValley

17. Quebec University (Canada) *

18. Rochester Institute of Technology

19. Seattle University

20. Southern Methodist University

21. Stevens Institute of Technology

22. Texas Tech University

23. University of Alabama

Huntsville24. University of Maryland University

College

25. University of Michigan Dearborn

26. University of Southern California

27. University of York (UK) *

28. Villanova University

* Non-US Schools


15/34

Observations from 28 schools

1. SWE is largely viewed as a specialization of ComputerScience - much as systems engineering was oftenviewed as specialization of industrial engineering oroperations research years ago

2. Faculty size is small - few dedicated SWE professors,

making programs relatively brittle3. Student enrollments are generally small compared to CS

and to other engineering disciplines

4. Many programs specialize to specific markets such asdefense systems or safety critical systems

5. The target student population varies widely - anyonewith Bachelors and B average to someone with CSdegree and 2+ years of experience

6. Online course delivery is popular

15


16/34

More observations7. Objective for graduates vary widely - software developer to

researcher to software manager

8. Wide variation in depth and breadth of SWEBOK coverage inrequired and semi-required* courses

9. Many programs have required or semi-required courses thatcover material that is either not in the SWEBOK at all or is notemphasized in the SWEBOK

10. Some significant topics are rarely mentioned - agility,software engineering economics, systems engineering

11. Some topics are ubiquitous - formal methods and architecture

12. Object-oriented is the standard development paradigm -creating a clash with many systems engineering programsthat emphasize structured methods

16*A student has a 50% or greater probability of taking a semi-requiredcourse.


17/34

17

Diverse focuses

1. Development of defense systems

2. Acquisition of defense systems

3. Embedded real-time systems

4. Entrepreneurial technology companies

5. Quantitative software engineering

6. Software economics

7. Safety critical systems

8. Secure software engineering

9. Highly dependable software systems

No focusdominated


18/34

18

Entrance requirements

%

Programs

Most programs offerleveling courses forstudents lacking entrancerequirements

Many programs routinelywaive academic requirementsfor students with industrialexperience


19/34

19

Methodology to Assess SWEBOKcoverage in required and semi-required courses

0: No coverage of topic

1: Some coverage but nodedicated course

2: One dedicated course

3: Two or more dedicated courses

REQ Sw RequirementsDES Sw DesignCST Sw ConstructionTST Sw TestingMNT Sw MaintenanceCNF Sw Config. Mgmt.MGT SwE ManagementPRC SwE ProcessTLS SwE Tools and

MethodsQLY Sw Quality

Required: Student must take thecourse

Semi-Required: 50% or moreprobability that course will be

taken

sample


20/34

20

SWEBOK coverage inrequired and semi-required courses


21/34

21

Some novel required andsemi-required courses

1. Reverse Engineering (Drexel)

2. Software Evolution and Re-engineering (Rochester)

3. Software Documentation (PennState Great Valley)

4. Software Risk Assessment in DoD(NPS)

5. Refactoring (Mercer)

6. Structured Document Interchangeand Processing (DePaul)

7. Avoiding Software Project Failures(Carnegie Mellon West)

8. Mathematical Foundations ofSoftware Engineering (Monmouth)

9. Global Software Development(Carnegie Mellon)

10. Management of OutsourcedDevelopment (Carnegie Mellon West)

11. Professional, Ethical and LegalIssues for Software Engineers (Cal.

State Univ. Fullerton)

12. Managing Software Professionals(Carnegie Mellon West)

13. Lean & Agile Software Processes(Mercer)

14. Artificial Intelligence (Michigan -Dearborn)

15. Software Engineering Economics(GMU, USC)

16. Computer Game Design &Implementation (Michigan -

Dearborn)17. Service Oriented Architecture


22/34

22

The approach

GSwERC 0.25, publicity, endorsements1. Understand the current state of SWE graduate education

(November 2007)


3. Publicize effort through conferences, papers,website, etc. (continuous)

4. Obtain endorsement from ACM, IEEE, INCOSE, NDIA,and other professional organizations (continuous)

5. Create GSwERC 0.50 suitable for broad community reviewand early adoption (Summer 2008)



23/34

23

The approach GSwERC 0.50

1. Understand the current state of SWE graduate education(November 2007)


3. Publicize effort through conferences, papers, website, etc.(continuous)

4. Obtain endorsement from ACM, IEEE, INCOSE, NDIA, andother professional organizations (continuous)

5. Create GSwERC 0.50 suitable for broad communityreview and early adoption (October 2008)



24/34

24

Expectations at entry

1. The equivalent of an undergraduate degree incomputing or an undergraduate degree in anengineering or scientific field and a minor incomputing

2. The equivalent of an introductory course insoftware engineering

3. At least two years of practical experience insome aspect of software engineering orsoftware development.


25/34

25

Outcomes 1 to 4 at graduation

1. Mastered the Core Body of Knowledge

2. Mastered at least one application domain, such asfinance, medical, transportation, ortelecommunications, and one application type, such

as real-time, embedded, safety-critical, or highlydistributed systems. That mastery includesunderstanding how differences in domain and typemanifest themselves in both the software itself and intheir engineering, and includes understanding how to

learn a new application domain or type.3. Mastered at least one knowledge area or sub-area

from the CBOK to at least the Bloom Synthesis level.

4. Demonstrated how to make ethical professionaldecisions and practice ethical professional behavior.


26/34

26


1. Understand the relationship between softwareengineering and systems engineering and beable to apply systems engineering principles andpractices in the engineering of software.

2. Be able to work effectively as part of a team,including teams that may be international andgeographically distributed, to develop qualitysoftware artifacts, and to lead in one area ofproject development, such as projectmanagement, requirements analysis,architecture, construction, or quality assurance.

3. Show ability to reconcile conflicting projectobjectives, finding acceptable compromises

within limitations of cost, time, knowledge,


27/34

27


8. Understand and appreciate the importance offeasibility analysis, negotiation, effective workhabits, leadership, and good communicationwith stakeholders in a typical software

development environment.

9. Understand how to learn new models,techniques, and technologies as they emerge,

and appreciate the necessity of such continuingprofessional development.

10.Be able to analyze a current significant softwaretechnology, articulate its strengths and

weaknesses, and specify and promote


28/34

28

Curriculum architecture


29/34

29

Flexible course structure


30/34

30

Comparison of 4 masters programs

to GSwERC 0.25 (not 0.50)

Scale: 1 = does not implement at all, 5 = fully implements


31/34

31

Reminder where we are today

1. Understand the current state of SWE graduate education(November 30, 2007)

2. Create GSwERC 0.25 with a small team, suitable for limitedreview (February 2008)

3. Publicize effort through conferences, papers, website, etc.(continuous)

4. Obtain endorsement from ACM, IEEE, INCOSE, NDIA, andother professional organizations (continuous)

5. Create GSwERC 0.50 suitable for broad community reviewand early adoption (October 2008)



32/34

32

Integrating Stevens systems andsoftware engineering masters

programs

1. Stevens has a large systems engineering masters program

2. Stevens has a moderate size software engineering program

3. Historically, these two programs were in separate schools withvirtually no overlap or ties

4. In September 2007, software engineering program moved toSchool of Systems and Enterprises in same academic unit assystems engineering

5. Concluded in fall 2007 that some integration between programs

was needed and possible - customers asked for it6. In spring 2008, began effort to integrate introductory courses


33/34

33

First integrated course combines

introductory material

Fundamentals of

Systems Engineering

Introduction to Software

Engineering

Introduction to Systems and

Software Engineering

Software Requirements

Analysis and Engineering


34/34

34

Next steps

Now combining systems architecture course andsoftware architecture course into one architecturecourse for both systems and software engineers

Will probably add advanced architecture coursesfor software and systems engineers to select

Now updating undergraduate survey of softwareengineering course to include systemsengineering as well

Piloting new courses this fall

Date post:	30-May-2018
Category:	Documents
Upload:	incosewma
View:	216 times
Download:	0 times

Presentation from September 9, 2008 Dinner Meeting

Documents