SGER A Framework for Adapting Decision-Based Scientific Principles in Engineering Design

to The National Science Foundation Directorate for Engineering Division of Design, Manufacture and Industrial Engineering Engineering Design Program by Department of Mechanical Engineeri

to

The National Science Foundation

Directorate for Engineering Division of Design, Manufacture and Industrial Engineering

Engineering Design Program

by

Department of Mechanical Engineering Stevens Institute of Technology

SGER: A Framework for Adapting Decision-Based Scientific Principles in Engineering Design

June 25, 2002

Approved by:

_

Professor Bernard Gallois Dean, School of Engineering Stevens Institute of Technology

Department of Mechanical Engineering Stevens Institute of Technology Castle Point on Hudson Hoboken, NJ 07030 Tel.: (201) 216-5559 Fax: (201) 216-8315 E-Mail: sesche@stevens-tech.edu

While modern engineering curricula should

reflect the emerging view that engineering

design involves decision making under

conditions of uncertainty and risk, student

exposure to these concepts is virtually non

existent in current educational programs A

pilot project is proposed, which will establish

an information-based approach as the

emerging paradigm in engineering design

education by demonstrating in two mechanical

engineering courses that concepts of data

uncertainty, decision theory and optimization

can be integrated effectively into

undergraduate engineering curricula.

Vision

Engineering design represents a process of

decision making under conditions of uncertainty

and risk, but today’s undergraduate engineering

curricula rarely include any principles of decision

theory.1 Even though value or utility theory are

crucial components of the decision making

process, these subjects are typically heavily

underrepresented in engineering curricula and

often treated incorrectly by the engineering

community at large Probability theory, which

establishes the basic mathematical tools needed for

the proper assessment of uncertainty and risk, is

often not put into learning-enhancing context such

as engineering design This unsatisfactory

situation calls for a revolutionary shift in design

education where practical examples of real design

cases are used to illustrate the application of sound

scientific principles

This program will prompt a strategic initiative for

the development, implementation and assessment

of novel approaches in engineering design

education at Stevens Institute of Technology

(SIT) Starting with a thorough feasibility study in

two courses taken in the junior year by mechanical

engineers, it will be demonstrated that the concepts of uncertainty in data, decision theory and optimization can be integrated effectively into undergraduate engineering design education Upon successful completion of this pilot project, this approach will be implemented immediately into the capstone design sequence in the mechanical engineering department Later, it will be propagated to the entire engineering curriculum at SIT through a major revision of the entire curriculum

Current State

SIT is a private technological university dedicated

to study and research Founded in 1870, SIT offers baccalaureates, master's and doctoral degrees in engineering, science and management It has earned an excellent reputation for its pioneering broad-based engineering curriculum.2 Based on its history, SIT provides a fertile ground for future curricular innovations SIT has often been an early adopter of emerging pedagogical approaches and educational technologies (e.g., student-owned PC/laptop, fully networked campus, WebCampus.Stevens distance learning program3), and major funding was committed to the development of state-of-the-art design laboratories Recently, SIT implemented a new undergraduate engineering curriculum This curriculum is designed to reflect the nationwide trend towards enhancement of traditional lecture-based courses with a significant design thread that propagates through the entire undergraduate educational program.4 Each design laboratory is integrated with one of the major engineering courses as outlined in Table 1

Currently, ME 322: Engineering Design VI, is the first discipline-specific design course in the mechanical engineering curriculum It embraces a holistic view of design that encompasses design activities spanning the entire product development cycle The topics covered include product conception, identification of customer needs,

product specifications, concept generation, concept

selection, concept testing, product architecture,

industrial design, design for manufacturing and

product development economics The associated

course, ME 345: Modeling and Simulation,

introduces modeling and simulation

methodologies and tools including model-block

building, logical and data modeling, validation

with applications in simulating product

performance, assembly, process and

manufacturing modeling and entity flow modeling

including conveyors, transporters and guided

vehicles

Table 1: Integrated design spine

Term Design Course Accompanying Course

1 E 121 Design I: design

concepts, product

dissection, professional

practice

E 120 Engineering Graphics: projections, dimensioning, tolerances

2 E 122 Design II: design

of structures, design for

environment, aesthetics

E 126 Mechanics of Solids: particle statics, force analysis, stress, strain

3 E 231 Design III:

design of energy

conversion systems and

chemical reactors

E 234 Thermodynamics

& Energy Conversion:

heat & work, 1st & 2nd

law, processes & cycles

4 232 Design IV: design

of filters, amplifiers and

embedded controllers

E 246 Electronics &

Instrumentation: signal acquisition &

processing, sensors, micro-controllers

5 E 321 Design V:

open-ended design projects

on processing of

materials to produce

products

E 344 Materials Processing: engineering properties of materials, scientific understanding

of properties and methods of controlling them

6 ME 322 Design VI:

discipline-specific

integrated product &

process design

ME 345 Modeling and Simulation:

methodologies and tools, applications 7/8 ME 423/424 Design

VII/VIII:

discipline-specific capstone design

project

ME 421 Engineering Economics: economic analysis, project management, marketing

of products

Plan of Action

Over a span of two years, the two mechanical engineering courses taught concurrently in the junior year will be completely redesigned to serve

as a test bed for the development, implementation and assessment of this novel approach to engineering design education A variety of design examples will be developed that can be used effectively to introduce concepts of decision making in the presence of uncertainty and risk as well as probability theory and optimization in the context of real design scenarios

This project will involve the tight integration of the two course syllabi ME 322 will introduce the theoretical concepts in the framework of a comprehensive group design project, and ME 345 will focus on the use of pertinent software tools

MS Excel will be used for data, regression and economic analyses as well as optimization Various MATLAB® programs will be implemented, which will be used by the students for the statistical analysis of data, the probabilistic modeling using the Monte Carlo method and the consideration of broad ranges of system options by optimization techniques

Objectives and Outcomes

Using two mechanical engineering courses as test bed, this project will demonstrate that methods for handling uncertainty in data, decision making based on value theory and optimization techniques can be integrated effectively into undergraduate engineering curricula The educational materials developed and the experiences gained herein will form the basis for more comprehensive curricular changes and cross-fertilization of related research programs at SIT.

Strategic Objectives

Successful completion of this program will prompt

a strategic initiative that aims at generating a

heightened awareness of the probabilistic nature of

engineering design attributes and establishing

science-based engineering design practices in

undergraduate education Through the

development, implementation and assessment of

educational modules utilizing rigorous

science-based techniques, this initiative will promote the

application of sound scientific principles in the

engineering design process It will thus form the

basis for a fundamental paradigm change in

undergraduate engineering design education and

encourage similar curriculum developments at

institutions nationwide

Project Objectives

This project is to be understood as a catalyst for

more comprehensive curricular changes at SIT and

later dissemination to other institutions It will

initially be of limited scope Examples and

methods for handling uncertainty in data, decision

making based on rigorous value theory and

optimization techniques will be implemented into

two undergraduate courses These topics will be

placed into the context of the engineering design

process that has traditionally been more

experience-based and problem-solving oriented A

successful implementation of this focused

approach will demonstrate that the above

mentioned concepts can be effectively presented to

and successfully applied by undergraduate

engineering students

Project Outcomes

Upon completion of the project, two junior-year

courses in mechanical engineering will have been

carefully redesigned and appropriate instructional

materials and tools will have been developed The

proposed changes include the reduction or

replacement of some of the currently taught topics

and the close integration of the syllabi

These two courses will provide the justification for comprehensive curricular revisions throughout the mechanical engineering program and across all engineering disciplines at SIT Additional external funding for these initiatives will be solicited -possibly from NSF - based on the experiences gained in the process of course development, implementation and piloting

Significance and Justification

The activities proposed herein are well aligned with the educational philosophy at SIT and hold strong potential for synergies with ongoing and planned research activities While the testing of the underlying educational hypothesis represents a high-risk proposition, successful project completion offers strong potential for catalyzing rapid and innovative advances in design education.

Significance

The activities proposed herein take into consideration the special character of the educational philosophy at SIT, which clearly reflects a strong awareness for the importance of design skills for engineering practitioners In particular, this project will form the framework and foundation for a wide range of future improvements of the engineering education at SIT

by providing practice-relevant context to fairly theoretical concepts In addition, the project objectives hold strong potential for synergies with the PI’s ongoing and planned research activities on modeling systems with uncertainty.5 , 6 , 7 , 8 , 9 , 10

Justification

Engineering design practice based on decision making strategies that resort to utility theory and take into account the probabilistic nature of design attributes represents an emerging paradigm Corresponding pedagogical approaches suitable for teaching these scientific concepts effectively in

the undergraduate engineering curriculum have

not been developed and assessed yet While this

lack of precedence renders the proposed

exploratory project a high-risk proposition, upon

its successful completion it offers strong potential

for catalyzing rapid and innovative advances in

engineering design education with significant

impact on engineering design practice thereafter

Plan of Work

Project Phases

Phase I of this two-year project will be devoted

to the detailed project planning and

development of the educational materials In

Phase II, the two courses will be piloted and

the project outcomes assessed In Phase III, the

follow-on project for the propagation of the

approach to the entire SIT engineering

curriculum will be planned and corresponding

funding solicited.

This two-year project will be carried out in three

phases (see Table 2 for an estimated timeline)

Phase I: Project Planning and Materials

Development

In Phase I, the detailed planning of this

exploratory project will be conducted Detailed

course syllabi for the two courses will be created

and a complete listing of the educational materials

to be developed in support of the pilot program

will be compiled Furthermore, these materials

will be reviewed and improved on the basis of

discussions and visitations with other experts in

the field such as Professor Saari11 and Professor

Howard12, two world-renowned specialists in the

field of decision theory

Table 2: Project schedule for two-year project duration

Tasks Au

02 Sp03 Su03 Au03 Sp04 Su04 I-1 Detailed Project

Planning I-2 Educational Materials Development

I-3 Reporting/Dissemination II-1 Pilot Courses

II-2 Assessment of Project Outcomes

II-3 Reporting/Dissemination III Planning/Proposal for Follow-on Project

The educational materials to be developed will include:

 lecture notes for the two courses, which synthesize the current state of the art in decision and utility theory at a level that is appropriate for undergraduate education and relate to specific design/product development activities,

 realistic examples that will place the theoretical concepts into practical context,

 small student projects that will allow for exploration-based learning modes in ME 345,

 a comprehensive design project for ME 322,

 homework problems and pertaining solutions,

 MATLAB® scripts (basic statistical analysis and graphical representation of data, Monte Carlo algorithms for probabilistic modeling of engineering systems, modeling of decision event trees and modeling of a comprehensive case study, optimization)

 MS Excel scripts (data analysis, regression analysis, financial analysis, optimization)

Phase I will be concluded by writing the progress report to the program manager and preparing initial presentations on this project at the ASEE Conference and Exposition and at the ASME Design Conference

Phase II: Pilot Courses and Outcomes Assessment

In Phase II, the two courses will be taught

concurrently The course outcomes will be

assessed according the following metrics: (i)

student satisfaction as measured by questionnaires

and attitudinal surveys and (ii) student

performance as measured by grades in the pilot

courses as well as in the capstone design The

student feedback will be collected using online

forms implemented with D*cide™, an assessment

software application developed at Stevens.13

Phase II will be concluded by disseminating the

project results through presentations at the ASEE

and ASME Design Conference, as well as through

preparation of peer-reviewed publications in the

ASEE Journal of Engineering Education14 and

ASME Journal of Mechanical Design.15

Phase III: Planning of Follow-on Project

In Phase III, the possibility for the propagation of

this novel approach to design engineering

education into the entire curriculum at SIT will be

analyzed Potential collaborators from other

engineering departments will be sought to

carefully plan more comprehensive curricular

implementations at SIT At the same time,

prospective external funding sources for the

corresponding developments such as the

Educational Materials Development (EMD) track

of the NSF-CCLI program16 will be reviewed and

targeted with proposals

Implementation

Two existing mechanical engineering courses

will be redesigned with a focus on

decision-based design engineering They will include a

comprehensive design project covering the

main phases of product development as well as

a laboratory component on related software

tools Collaborative learning through

self-discovery will be the preferred learning mode.

ME 322: Engineering Design VI and ME 345: Modeling and Simulation were selected as a test bed for the novel approach to design education proposed, in which the traditional teaching mode based on problem-solving is replaced by a focus

on decision making taught through collaborative student self-discovery These mechanical engineering courses are taken concurrently in the junior year for 2 and 3 credits, respectively Both classes include a lecture component, but while ME

322 features a comprehensive semester-long group design project covering the main phases of the engineering design process from product conception to economic analysis, ME 345 features

an accompanying laboratory session with a series

of smaller projects assigned to student teams with

3 to 4 group members Engineering reports and group presentations will complement the exams as assessment methods for student performance Compared with the current course implementations, coverage of the following topics will be reduced significantly or discontinued completely: product architecture, industrial design, assembly and production modeling and entity flow modeling These educational modules will be replaced by the topics listed in columns 2 and 3 of Table 3, which are integrated with the design activities listed in column 1 The primary textbooks17 , 18 selected for the two courses will be supplemented by material from the sources referenced in the table

Key Personnel and Success Measures

This project will be carried out by Dr Esche and Dr Chassapis with the support of one graduate student Its success will be measured based on pedagogical metrics as well as funding, publications and peer recognition.

Key Personnel

Dr Esche, Assistant Professor, will be the PI of the project who will be in charge of the plan of

work described above Dr Chassapis, Associate

Professor and Department Director, will act as

Co-PI and work in an advisory function He is the

Director of the Integrated Product Development

Program19 and Associated Director of the Design

and Manufacturing Institute.20 In addition to their

experience in teaching various design courses,

Drs Esche and Chassapis also have a strong

record in collaborating on the development,

implementation and administration of new

courses, educational programs and student

laboratories at SIT They are members of the

undergraduate curriculum committee of the

mechanical engineering department and have been

instrumental in a variety of strategic initiatives to

adapt innovative pedagogical approaches and

techniques such as project-based learning,21 , 22 , 23

assessment methodologies,24,25 remote

laboratories26 , 27 , 28 , 29 , 30 and their implementation into

the mechanical engineering curriculum at SIT

Table 3: Topics and tools to be integrated into ME

322 and ME 345

Design

Subject

ME 322 Topic ME 345 Topic Product

conception Decision making, design

options31,32,33,34

Model building, logical and data modeling35,36

Statistics

review

Sets, probability,

distributions, Bayes’

formula37,38

Data analysis with MATLAB39

and MS Excel40

Identification

of customer

needs

Forecasting41 Regression

analysis with MATLAB42 and

MS Excel43

Product

specs,

concept

generation &

selection

Rationality, utility

functions,44 Arrow’s

theorem,45,46,47

decision making in

presence of risk,

Borda Count,48,49 von

Neumann-Morgenstern utility50

Monte Carlo simulation of probabilistic models with MATLAB,51

optimization with

MS Excel, MATLAB52

Concept

testing53 Decision

analysis54,55,56,57 Monte Carlo

simulation of decision event trees with MATLAB58

Product

development Discounting, present value, interest, Monte Carlo simulation of

economics inflation/deflation59 case study with

MATLAB,60

financial analysis with MS Excel61

Measures of Success

The success of this exploratory project will be measured according to the following metrics: (i) student performance and student satisfaction during the pilot project and in the capstone design sequence following directly thereafter, (ii) early indications that other faculty will adopt the proposed pedagogical approach and the related educational materials in their courses (iii) number

of conference presentations and publications in conference proceedings and peer-reviewed journals, and (iv) follow-on funding by external sponsors (e.g., non-profit foundations, corporate sources, governmental agencies) for additional educational developments or cross-fertilized research activities

SGER: A Framework for Adapting Decision-based Scientific Principles in Engineering Design

Submitted to NSF on June 25, 2002 by S Esche & C Chassapis, Stevens Institute of Technology

L I S T O F R E F E R E N C E S

Goals of the Engineering Design Program, URL: http://www.eng.nsf.gov/dmii/Message/EDS/ED/ed.htm

2 Stevens Institute of Technology, URL: http://www.stevens-tech.edu/history/

3 WebCampus.Stevens, URL: http://attila.stevens-tech.edu/gradschool/distance_learning/

4 Sheppard K & Gallois, B (1999) The design spine: revision of the engineering curriculum to include a design experience each semester, Proceedings of the 1999 ASEE Annual Conference and Exposition, Charlotte, North Carolina, June 1999

5 Yu, Q & Esche, S K (2002) Modeling of grain growth kinetics with Read-Shockley grain boundary energy by a modified Monte Carlo algorithm Accepted for publication in Materials Letters

6 Yu, Q & Esche, S K (2002) A new perspective on the normal grain growth exponent obtained in

two-dimensional Monte Carlo simulations Submitted for publication in Modeling and Simulation in Materials Science and Engineering

7 Yu, Q & Esche, S K (2002) A Monte Carlo algorithm for single phase normal grain growth with improved accuracy and efficiency Submitted for publication in Computational Materials Science

8 Esche, S K., Chassapis, C & Manoochehri, S (2001) Concurrent product and process design in hot forging Concurrent Engineering: Research and Applications, Vol 9, No 1, pp 48-54

9 Esche, S K., Fidan, I., Chassapis, C & Manoochehri, S (2000) Knowledge-based part and process design for metal forging SAE Transactions - Journal of Materials and Manufacturing, Vol 108, No 5, pp 92-99

10 Esche, S K., Hadim, H & Chassapis, C (2002) Prototype for a wireless web-based building services monitoring and control system Proposal submitted to NSF-ESS program

11 Saari, D G., University of California at Irvine, Department of Mathematics, URL:

12 Howard, R A., Stanford University, Department of Management Science and Engineering, URL:

http://www.stanford.edu/dept/MSandE/faculty/rhoward/

13 D*cide for Educational Assessment™ by Choice Logic Corporation, URL: http://www.choicelogic.com/

14 ASEE Journal of Engineering Education, URL: http://www.asee.org/publications/jee/

15 ASME Journal of Mechanical Design, URL: http://www-jmd.engr.ucdavis.edu/jmd/

16 NSF Course, Curriculum, and Laboratory Improvement (CCLI) Program, URL:

http://www.ehr.nsf.gov/due/programs/ccli/

17 Ulrich, K T & Eppinger, S D (2000) Product Design and Development 2nd ed., McGraw-Hill Companies, Inc., 2000

18 Law, A M & Kelton, D W (1999) Simulation Modeling and Analysis, 3rd edition, McGraw Hill Companies, Inc., 1999

19 Integrated Product Development Program, URL: http://www.soe.stevens-tech.edu/ipd/

20 Design and Manufacturing Institute, URL: http://www.dmi.stevens-tech.edu/

21 Esche, S K (2002) Project-based learning in a course on mechanisms and machine dynamics Submitted for publication in World Transactions on Engineering and Technology Education

22 Esche, S K & Hadim, H A (2002) Introduction of project-based learning into mechanical engineering courses Proceedings of the 2002 ASEE Annual Conference and Exposition, Montréal, Quebec, Canada, June 16 - 19, 2002

23 Hadim, H A & Esche, S K (2002) Enhancing the engineering curriculum through project-based learning Accepted for presentation at the 32nd ASEE/IEEE Frontiers in Education Conference, Boston, Massachusetts, USA, November 6 - 9, 2002

24 Esche, S K., Pochiraju, K & Chassapis, C (2001) Implementation of assessment procedures into the mechanical engineering curriculum Proceedings of the 2001 ASEE Annual Conference and Exposition, Albuquerque, New Mexico, USA, June 24 - 27, 2001

25 Esche, S K (2002) Assessment of an open laboratory approach using experimental stations with remote access Abstract submitted for presentation at the Second ABET National Conference on Outcomes Assessment for Program Improvement, Pittsburgh, Pennsylvania, USA, October 31 - November 1, 2002

26 Esche, S K (2002) On the integration of remote experimentation into undergraduate education Submitted for publication in ASEE Journal of Engineering Education

27 Esche, S K., Chassapis, C., Nazalewicz, J W & Hromin, D J (2002) A scalable system architecture for remote experimentation Accepted for presentation at the 32nd ASEE/IEEE Frontiers in Education Conference, Boston,

Massachusetts, USA, November 6 - 9, 2002.

28 Esche, S K (2002) Remote experimentation - one building block in online engineering education Accepted for presentation at the 2002 ASEE/SEFI/TUB International Colloquium on Global Changes in Engineering Education, Berlin, Germany, October 1 - 4, 2002

29 Esche, S K., Prasad, M G & Chassapis, C (2000) Remotely accessible laboratory approach for undergraduate education Engineering Education Beyond the Millennium, Proceedings of the 2000 ASEE Annual Conference and Exposition, St Louis, Missouri, USA, June 18 - 21, 2000

30 Esche, S K & Chassapis, C (1998) An Internet-based remote-access approach to undergraduate laboratory education Engineering Education without Boundaries, Proceedings of the 1998 Fall Regional Conference of the Middle Atlantic Section of ASEE, Washington, DC, USA, November 6 - 7, 1998, pp 108-113

31 Hazelrigg, G A (1996) Systems Engineering: An Approach to Information-based Design, Ch 1 & 2 Prentice Hall, 1996

32 Hazelrigg, G A (1989) Comments on the engineering method ASME PVP, Vol 177, pp 153-158

33 Hazelrigg, G A (1996) Systems engineering: a new framework for engineering design ASME DSC, Vol 60, pp 39-46

34 Hazelrigg, G A (1999) Framework for decision-based engineering design Journal of Mechanical Design, Vol

120, No 4, pp 653-658

35 Kelton, W D., Sadowski, R P & Sadowski, D A (2001) Simulation with ARENA, 2nd ed., McGraw-Hill, 2001

36 Hazelrigg, G A (1999) On the role and use of mathematical models in engineering design Journal of Mechanical Design, Vol 121, No 3, pp 336-341

37 Denker, M., Ycart, B., Woyczynski, W A & Balakrishnan, N (1998) Introductory Statistics and Random Phenomena: Uncertainty, Complexity and Chaotic Behavior in Engineering and Science Springer Verlag, 1998

38 Hazelrigg, G A (1996) Systems Engineering: An Approach to Information-based Design, Ch 3 Prentice Hall, 1996

39 Childers, D G (1997) Probability and Random Processes: Using MATLAB with Applications to Continuous and Discrete Time Systems Irwin Professional Publishing, 1997

40 Harnett, D L & Horrell, J F (1998) Data, Statistics, and Decision Models with Excel John Wiley & Sons, 1998

41 Hazelrigg, G A (1996) Systems Engineering: An Approach to Information-based Design, Ch 8 Prentice Hall, 1996

42 Palm, W J (2000) Introduction to MATLAB 6 for Engineers McGraw-Hill, 2000

43 Bloch, S C (1999) Excel for Engineers and Scientists John Wiley & Sons, 1999

44 Edwards, W (ed.) (1992) Utility Theories: Measurements and Applications Kluwer Academic Publishers, 1992

45 Arrow, K J (1963) Social choice and individual values 2nd ed., John Wiley & Sons, 1963

46 Hazelrigg, G A (1996) Implications of Arrow’s impossibility theorem on approaches to optimal design

engineering Journal of Mechanical Design, Vol 118, No 2, pp 161-164

47 Saari, D G (1998) Connecting and resolving Sen’s and Arrow’s theorems Social Choice and Welfare, Vol 15,

pg 239-261

48 Borda, J C (1781) Mémoires sur les elections au scrutin L’Histoire de L’Académie Royale des Sciences, Paris, 1781

49 Saari, D G (1990) The Borda dictionary Social Choice and Welfare Vol 7, pg 279-317

50 Hazelrigg, G A (1996) Systems Engineering: An Approach to Information-based Design, Ch 7 Prentice Hall, 1996

51 Hazelrigg, G A (1996) Systems Engineering: An Approach to Information-based Design, Ch 7 Prentice Hall, 1996

52 Hazelrigg, G A (1996) Systems Engineering: An Approach to Information-based Design, Ch 5 Prentice Hall, 1996

53 Box, G E P., Hunter, J S & Hunter, W G (1978) Statistics for Experimenters: An Introduction to Design, Data Analysis, and Model Building John Wiley & Sons, 1978

54 Hazelrigg, G A (1999) On irrationality in engineering design Journal of Mechanical Design, Vol 119, No 2, pp 194-196