Friday, October 1, 2010

Systems engineering class tackles complex real-world design issues - SDM Pulse Fall 2010

By Qi Van Eikema Hommes

Editor's note: In this article, Qi Van Eikema Hommes describes Systems Engineering, a course she taught this summer in MIT's System Design and Management Program (SDM).

Systems Engineering is one of the required core courses for MIT's System Design and Management Program (SDM). The course's objective is to help students develop a systems thinking capability by introducing classical and advanced systems engineering theory, methods, and tools. In this class, students learn to:
  • Develop a systems engineering plan for a realistic project.
  • Judge the applicability of any proposed systems engineering process, strategy, or methodology, using fundamental concepts from such disciplines as probability, economics, and cognitive science.
  • Understand the roles and responsibilities of systems engineers, as well as of systems engineering organizations.
  • Apply systems engineering tools (e.g., requirements development and management, robust design, design structure matrix, or DSM), to realistic problems.
  • Recognize the value and limitations of modeling and simulation.
  • Formulate an effective plan for gathering and using data.
  • Proactively design for, and manage, a system's lifecycle targets.
The subject of systems engineering sits at the intersection of engineering design, social science, and business economics. Topics including requirements management, system modeling and simulation, risk management, decision analysis, robust design, reliability, safety, and security all fall within this discipline. The first challenge in developing this course is how to help students get a good grasp of the fundamentals of systems engineering in a nine-week summer term.

In addition, systems engineering is deeply rooted within the engineering discipline. Many systems engineering tools and methods evolved to solve the engineering design problems found in increasingly large and complex engineering systems. However, these methods and tools, and the philosophy of systems thinking, extend to other systems that are not traditionally viewed as engineered systems-such as the enterprise system, the health-care system, and the energy system. So, the second challenge that I faced in teaching this course was helping students make the connection between the traditional systems engineering methods and tools, and the problems that are relevant today.

With the aforementioned challenges in mind, I designed several features into this course. First, this class introduced students to a range of advanced, cuttingedge research topics-including the roles of leadership and innovation in systems engineering, requirementsdriven predictive design structure matrices, one-factor-at-a-time vs. orthogonal arrays in design of experiments, and trade-space exploration techniques. Many of these topics emerged from the latest MIT Engineering Systems Division faculty research-and we were fortunate to have several faculty members themselves give the relevant lectures. Students had the privilege to learn material directly from Professor Dan Frey, Dr. Madhav Phadke, Dr. Donna Rhodes, and Dr. Adam Ross. This is one of the ways SDM's course in systems engineering stands out from what might be offered at another school.

To help students to organize these diverse topics in their minds, the class was structured in the form of the Systems Engineering Vee.

The second feature of this course is a series of five context-based opportunity sets, designed to reinforce materials learned in class. When I was preparing for this class in the early spring of 2010, Professor Steven Eppinger, SDM's MIT Sloan faculty codirector, encouraged me to put the class exercises in a context to help students better ground what they learn from the lectures. Toyota's safety recall was a heated topic in the news this summer, so I decided to focus on that.

Choosing that Toyota case study had two advantages: first, the publicly accessible information about cars and; second, a car is a complex system that every student can actually get his or her hands on. Therefore, the first four assignments were designed to engage students in applying a variety of methods and tools to the Toyota safety recall investigation, including stakeholder analysis and requirements definition, axiomatic design and DSM methods, TRIZ (Russian acronym for Theory of Inventive Problem Solving), robust design, and design of experiments. Students learned to look at system design problems through the lens of systems engineering methods and tools.

No collaboration was required for the first four assignments, so I designed the fifth assignment as an opportunity for students to share what they had learned about the Toyota case. Students formed teams of five and prepared presentations as if Toyota or the National Highway Traffic Safety Administration had called upon them to help with their investigation.

In this exercise, students shared with and learned from one another what they found in the news media, and how they analyzed the case. In the classroom, they presented their analyses to several systems engineering experts, including Dr. Richard John, director emeritus of the Volpe National Transportation Systems Center; Frank Serna, director of systems engineering at Draper Laboratory; Tony Lockwood, Ford hybrid systems engineering manager; and MIT's own expert on automotive system design, Dr. Daniel Whitney, senior research scientist at the MIT Center for Technology, Policy and Industrial Development. These experts provided additional comments and suggestions that enhanced the students' learning experiences.

So that students could calibrate their thinking against what the experts in the field were saying, I invited MIT Institute Professor Sheila Widnall, who serves on the Toyota Quality Advisory Panel, to share her experiences on the case with the class. In addition, Tom Baloga, BMW's vice president of product development and safety in the United States, gave a lecture outlining the automotive industry's strategy on safety (ISO 26262-Functional Safety Standard for Road Vehicles) and product recall. Overall, the context-based opportunity sets enabled students to practice what was taught in class on a real, up-to-date problem, which enhanced their learning experiences.

The third feature of this course was the term project, designed to give students an opportunity to apply the systems thinking and systems engineering framework to a project near and dear to their own careers or passions. The SDM students really took this assignment to heart and selected a wide variety of topics, including:
  • Alternative Energy Solution for the State of Hawaii, and Hawaii 2030 Energy Independence Goals Feasibility Analysis
  • Future Mobility Solution for Los Angeles Commuters
  • Systematic Development of Engine Control Software
  • Education System for Rural Secondary School Students
  • A System Failure Analysis of the BP Oil Spill
  • MIT Medical Flu Project
  • Emergency Room Operation System Improvements
All teams worked hard to collect data and analyze the systems they were designing. Many started with very broad topics, but with the guidance of the three faculty members in this class, students learned how to properly scope down a systems design problem and focus their resource on the most important questions. The term project exercise served as a sandbox to prepare SDM students to take real-world problems in stride.

Reflecting on this summer's teaching experience, I want to thank all of the SDM students for their effort and dedication. The active class discussions, the late night email questions, and project discussions inspired me to continue improving this course. I also want to thank my co-instructors SDM Director Pat Hale and the director of SDM's Certificate in Systems and Product Development, David Erickson, and my teaching assistants Ellen Czaika, SDM '08, and Ipshita Deepak, SDM '09. Their support, experiences, and knowledge helped me tremendously.

I look forward to teaching this class again next summer!

No comments:

Post a Comment