ENGINEERING SYSTEMS ANALYSIS FOR DESIGN

ESD.71, 1.146, 3.56, 16.861, 22.821
MIT School-Wide Elective

  Professor Richard de Neufville

Web site : http://ardent.mit.edu/real_options/ROcse_MIT_latest

Course Syllabus                                      Latest revision: Sept.10, 2009

Important notes: 
  • The subject and the syllabus has been completely revamped!                It has been completely reorganized to follow the sequence of chapters in the ated perspective on the material.  The class now follows the outline of the new book that the MIT Press will publish as an initial volume in its new series on Engineering Systems.  Prof. de Neufville has been writing it in collaboration with Prof. Stefan Scholtes from the University of Cambridge. 
  • The working title of the book is "Flexibility in Design".  Draft copies will be used in the course and made available electronically.
  • The syllabus emphasizes the development of "application portfolios".  These assignments show how the course material applies to each student's own professional interests (see Assignments section below).  Students have routinely incorporated these applications in their research, thesis, and professional presentations.
  • The subject assumes that participants are competent in the material covered by  the 3-unit micro-subject ( ESD 70J ) that covers Excel © spreadsheet analyses.  Go to that site for a self-assessment of whether you need to audit that subject (experience indicates you probably do).

Organization of Syllabus

This subject focuses on the flexible design of engineering systems.  It recognizes that the future is necessarily uncertain, and that we need to deliver good value for the range of possible scenarios.  Flexible designs enable system operators to adapt their project to future circumstances, to avoid downside risks and to take advantage of upside opportunities.

Many researchers and practitioners believe that flexible design will revolutionize the way we develop and manage complex projects. Indeed, case studies indicate that the approach can improve expected performance by around 25 percent.  These are significant opportunities.

The course shows how design flexibility can substantially increase the expected value of a system.  It presents an integrated approach that

  • properly deals with evaluation over time in risky situations, and
  • pinpoints the value of specific design elements.    

The syllabus consists of 6 major blocks.  The first two set the scene, the next four cover the essential elements of understanding the uncertainty, identifying valuable candidate flexibilities, evaluating and choosing preferred alternatives; and implementing flexible designs in practice. 

Part 1 Overview of Flexibility in Design

This introductory block sets the scene: it tells you why the material in this course is valuable and introduces the garage case example that shows you the essentials.

Conceptually, it shows why this approach represents a paradigmatic shift in concept from standard systems engineering and design. The crux is that we explicitly recognize the reality of uncertainty, and design our systems to perform well over the range of possibilities, and thereby increase expected value significantly -- often by 25% or more.

Part 2   Basic Concepts of Valuation

This section establishes the basic concepts for valuing projects.  Understanding this material is essential for any well-informed analysis and choice of preferred designs.

It first underscores the essential concepts of engineering economics and reviews the mechanics.  It then presents the fundamental notions of technical and economic efficiency, with particular emphasis on economies of scale and their significance as drivers of phased developments.

Part 3  Forecasting and Dynamic Modeling

Forecasting and modeling the uncertain, variable future is a crucial part of the approach founded on this reality.  The section develops decision analysis, dynamic  modeling, lattices and dynamic programming as core tools.

Part 4  Identification of Candidate Flexibilities

The identification of which flexibilities might be most effective in dealing with the plausible uncertain futures is not obvious.  This section develops basic approaches to this problem.

Part 5  Evaluation and Choice of Preferred Alternatives

When dealing with uncertain futures, it is almost meaningless to think of  "optimal" solutions.  This is because the preference for any alternative depends on trade-offs between risk and reward about which people differ.  We need to think of "preferred" alternatives rather than solutions that are demonstrably superior.

This section presents the concepts and methods for identifying and selecting preferred solutions.

Part 6  Implementation

The course closes with the vital discussion of the theory and practice of effective implementation of attractive solutions.  It is not enough to have a good idea, one has to know how to make good ideas happen.

Prerequisites

This subject builds upon a basic knowledge of calculus and probability.

Students should be proficient in the use of spreadsheet programs at the level covered by ESD 70J , the 3-unit intensive short course designed to bring them up to speed in Excel©.  They should click on the link to the subject, to assess whether they are already at the required level.  In general, everyone will benefit from participating in this offering.

Course Materials

The primary text for the class consists of draft chapters from the new text being written for the MIT Press.  These are posted on the class web site, and can be accessed by clicking on them in the schedule

This material is supplemented by chapters from Applied Systems Analysis (R. de Neufville, McGraw-Hill, 1990).  Participants can download these from the class web site.  Book versions are unavailable except in the library.  

Extensive references and papers are available from the detailed class syllabus and from the course web site.

All students will be expected to have an up-to-date full version of Excel©.   They may also want to try the following optional add-ins:

TreeAge Pro © can be downloaded for a 21-day trial from
http://www.treeage.com/products/download.html
and student licences can  be obtained from: http://server.treeage.com/treeagepro/purchase/index.asp
 

Crystal Ball ©  student licenses can be ordered from http://www.decisioneering.com/academic_home.html .

Participants can view and download pdf copies of the PowerPoint slides on the class web site. Prof. de Neufville will try to post each new presentation before the class.  Note that he also routinely improves and updates older presentations throughout the semester, so that the web versions of the slides change from time to time. Students may wish to verify that they have the latest version before downloading the slides for reference in class. 

Computer Environment

The class web site is the primary means of distributing course material. It contains the:

  • Syllabus and schedule of classes,
  • Pdf versions of slides used in the lectures,
  • Assignments of readings and homework,
  • Electronic versions of major problems sets and solutions,  
  • Additional software that students can use in the course,
  • Reference papers and theses.,and
  • Good examples of Application Portfolios from previous students.

Students are expected to use Excel © to work on assignments.

Stellar Bulletin Board

Stellar is the MIT course management system. Students should use it regularly.  It provides an electronic bulletin board the instructors will use it to post important information such as course updates and hints for assignments, and to respond to student questions.

Students should use the Stellar bulletin board to contact the instructors for general questions. By doing so, we can give a single answer to issues that may interest many students.  To reinforce this practice, we will answer emails using the bulletin board -- unless the question needs a private confidential answer.

Stellar also gives students access to their own grades.

Recitations

There will be regular weekly problem-solving sessions. Prof. de Neufville will schedule these in consultation with the class during the opening session.

Office Hours

Prof. de Neufville will be available for at least half an hour after class for immediate questions and comments. Students can also make appointments with him for personal discussions.

Assignments
 
Graded assignments for the class mainly involve the"application portfolio".  This is a collection of applications, on topics individually selected by each student, using the methods presented in class on an issue.  This has two phases
  • Individual assignments throughout the semester
  • A final portfolio that builds upon the individual assignments and the feedback during the semester.  It will present an integrated whole so that each student will each have a completed case study demonstrating how they can use design flexibility  to increase the expected value of their systems.
Ungraded problems  drawn the Applied Systems Analysis text are distributed throughout the semester.  Students are advised to complete them as a way of insuring that they know how to complete the applications.   Solutions to the problem sets will be posted so that students can grade themselves and get immediate feedback.

Grading

Grades depend principally on the application portfolio and tests, but include an allowance for class participation. The approximate weights are

  • Individual and Final Application Portfolio: 65%
  • Class Participation and graded assignments 10%
  • Mid-semester quiz: 25%

Students should complete assignments on time. The teaching assistant will mark down unexcused late assignments.

The instructors will modulate the final grade in appreciation of the participant's progress throughout the semester.  Those who finish strongly and demonstrate that they have, at the end, mastered the material will receive more credit for the final grades.

Previous mid-semester quizzes are posted in the "Exercises" section of the web page. They indicate the kind of questions likely to appear on future exams. However, their content may not match the current syllabus which has been completely reworked.

Absences

Students should plan on being at the mid-semester quiz and the final examination. Students who have conflicts should discuss them with Prof. de Neufville at the beginning of the semester to see what arrangements might be made. No one should expect special treatment and the extra effort to write equivalent examinations and give them at alternative times.

Prof. de Neufville will, of course, accept reasonable excuses (family emergencies, sickness) when presented near the event.

Academic Honesty

To avoid confusion that might result from different expectations in other contexts or establishments, please note the standards that apply in this subject:

  • Assignments turned in for grading must be done individually.  The instructors understand that students will discuss the course and often learn best collectively. However, students should prepare their reports for each assignment individually, in their own format and words. 
  • Demonstrated evidence of copying (exactly the same presentation, same wording of sentences, etc.) will result in zeros for each paper with this evidence.
  • Anyone found cheating in the mid-semester quiz (copying from another student or using unauthorized materials, etc.) will receive a zero for the event.
  • Material copied from other sources (articles, reports, web pages, etc) must be properly cited.  Evidence that material has been taken from elsewhere without citation will be treated as plagiarism.
  • Prof. de Neufville will place a note in the student record of each person associated with plagiarism or cheating. MIT views these infractions seriously and routinely expels students who seriously violate our academic standards. 
Any questions about this policy should be addressed to the instructors.