Course Syllabus Latest revision: Sept.10, 2009Important notes:
Organization of Syllabus
focuses on the flexible design of engineering systems. It
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.
The course shows how design flexibility can
substantially increase the expected value of a system.
It presents an integrated approach that
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
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.
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.
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.
builds upon a basic knowledge of calculus and probability.
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
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.
references and papers are available from the detailed class syllabus
and from the course web site.
will be expected to have an up-to-date full version of Excel©. They may also want to try the
following optional add-ins:
© can be downloaded for a 21-day trial from
Ball © student licenses can be ordered
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
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.
The class web site is the primary means of distributing course material. It contains the:
expected to use Excel
© to work on
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.
There will be
regular weekly problem-solving sessions. Prof. de Neufville will
schedule these in consultation with the class during the opening
Neufville will be available for at least half an hour after class for
immediate questions and comments. Students can also make appointments
him for personal discussions.
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
Grades depend principally on the application portfolio and tests, but include an allowance for class participation. The approximate weights are
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.
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.
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.
Neufville will, of course, accept reasonable excuses (family
emergencies, sickness) when presented near the event.
To avoid confusion that might result from different expectations in other contexts or establishments, please note the standards that apply in this subject: