One Journey to Engineering Systems
The organizers of the 3rd International Symposium on Engineering Systems asked me to provide a brief story of my journey to engineering systems — how I came to be at this Symposium on this June evening in Delft. The idea is to stimulate your thinking and perhaps motivate you to share your stories during dinner this evening.
I have always been a planner, so the story of how I came to be at this meeting begins in 1961. I was 14 years old when I bought my first car. My intention was to teach myself to drive, but it turned out that I was also to teach myself some engineering. The object in Picture 1 prompted the need for this.
Picture 1. Carburetor
This picture shows the carburetor of the car I bought — the 1952 Plymouth in Picture 2. This is the same model and colors of the car I bought. However, for $35 the car did not exactly look like this one. In fact, as I repaired body dents and rust, red primer slowly replaced much of the blue and white.
Being too young to get a driver’s permit, I decided to learn by driving through the fields in this small, island town of Portsmouth, Rhode Island — the smallest state in the United States. Picture 3 shows my training track, which was captured during an aerial survey of the town prior to building a new road. The track started out very bumpy but over time the Plymouth smoothed it out.
Picture 2. 1952 Plymouth
Driving endlessly on this track, I became very good at left-hand turns, and right-hand turns, in second gear but could never get going fast enough to get into third gear. I also could make it to the center of town, with its 3-4 stores, without going on a public road. This meant driving through fields normally populated by cows or potatoes. Again, it was very bumpy. The carburetor did not like this. It failed often. Pieces bent or disappeared.
Picture 3. Training Track
I got to know the carburetor very well. I constantly had to figure out how to get it functioning again. I did not have the money to buy replacement parts. I resolved this quandary with my Erector Set, shown in Picture 4. I used the linkages and sliders from this set to restore the carburetor’s function, although not quite in the way originally intended. The result was a Rube Goldberg version of a carburetor.
Picture 4. Erector Set
The Erector Set could not compensate for my lack of knowledge that cars had oil filters. Eventually the Plymouth’s engine succumbed to a lack of oil — plenty in the reservoir but little making it to the engine. Next came a 1949 Chevrolet, again for $35. This car’s specialty was leaks. During heavy rains, the floor wells in the front seat would fill up. I learned that water would always have its way.
Each subsequent car provided its own lessons. There were many lessons. We managed to successfully implant a Ford V-8 in a Jeep. However, a Renault’s frame could not support a Thunderbird V-8, providing a lesson in mechanics of materials. We also learned that people were happy to get rid of old cars in their backyards and we could try all sorts of things, while selling parts salvaged from these cars.
By the time I was 20, I landed a job as an assistant engineer at Raytheon, a large defense contractor in Portsmouth. I was involved with the AN/BQS-13 sonar system, which looked much like Picture 5. During my two years at Raytheon, before heading to graduate school at MIT, I worked in mechanical, electrical, and systems engineering. It was a wonderful immersive experience.
My ultimate task was to figure out how many spare parts to bring on a submarine, given the different reliability and maintainability characteristics of the subsystems, assemblies and components and, of course, the limited space for spare parts. I also was able to participate in some of the discussions of the human operators of these systems in terms of the information and interface needed to support their tasks.
Picture 5. Sonar Room
Picture 6 shows my current involvement. In some ways, the hospital operating room is like the sonar control room. However, there are critical distinctions. The information systems and the incentive systems in healthcare are much more poorly aligned with human behavior and performance requirements. Inadequate information and poorly aligned incentives are issues that must be addressed in a broader context than depicted by Picture 6.
Picture 6. Operating Room
This observation reflects a life-long penchant to question external constraints — to wonder why “givens” are given. The “best” solution to a problem is almost always only best within a given set of assumptions. These assumptions often reflect constraints that, when viewed broadly, can be seen as totally arbitrary. In this way, problem solutions are often limited by problem definitions. I am always energized by the question of what is the real problem.
So, what have I learned? Three lessons seem like a good number. First, no matter at what level you address a complex system, there is always a broader context that impacts the system. Simply externalizing the context is not a good idea; this assumption will plague you later by hindering successful operation and maintenance of the system. For example, better design of hospital operating rooms will not, in itself, remediate the information and incentives problems of healthcare delivery.
Second, complex systems cannot be addressed successfully by a single, traditional discipline. Appropriate consideration of interactions at different levels of abstraction and aggregation require a mix of knowledge and skills. Consequently, one professor and one PhD student cannot single-handedly address complexity successfully. Interdisciplinary approaches are needed to transform complex systems.
Third, the engineering of complex systems requires a professional community that embraces the first two lessons. For me, the Council of Engineering Systems Universities and the International Symposium on Engineering Systems is that community. I am indeed fortunate that the 1952 Plymouth carburetor headed me in this direction five decades ago. I expect CESUN and the Symposium to be central to transforming complex systems for many years to come.