How to Think Like an Engineer

Most engineers spend a lot of time thinking about how to solve design problems. However, most of us don't spend very much time thinking about how we think. Fortunately, two cognitive psychologists at the University of Pittsburgh have done just that. Joel Chan and Christian Shunn are the authors of a recent paper in the journal Cognitive Science focusing on how engineers use analogies to develop innovative concepts. Their paper is based on observations of real-world engineers "in the wild," as they put it. (In this case, "in the wild" means in a conference room, not a tropical rainforest.)

Analogies are comparisons between two things that shed light on their similarities. Engineers sometimes make analogies between thermal systems and electrical circuits. For instance, when it comes to the flow of heat through systems, thermal resistance and thermal capacitance (heat capacity) follow essentially the same rules as electrical resistance and capacitance do for the flow of electric current. We also sometimes make analogies between electrical circuits and mechanical systems. For example, in analyzing RLC circuits, we can think of capacitors as masses and inductors as springs. In the design process, we often come up with many other analogies. These help us think about complicated or unfamiliar problems in terms of simpler or more familiar ones.

Chan and Shunn identified four different roles that analogies can play in the design process. Analogies can be used for problem identification. For example, if a certain type of problem is common in a design similar to yours, you should make sure your design doesn't have the same problem. Analogies can also be used for concept generation. A famous example of this is George de Mestral's invention of Velcro hook and loop fasteners after struggling to remove spiky plant seeds from his dog's fur. Another use of analogies is to explain concepts. If you have a new idea, it might be easier to explain it to your co-workers in terms of something they're already familiar with. Finally, analogies can also be used for function finding. An example from a similar design might inspire you to incorporate additional functionality into your own design.

Analogies can be divided into two types. In "near analogies" the two things being compared are mostly similar, and in "far analogies" they are mostly different, but have one or more important similarities. For example, comparing a crankshaft used in an engine to one used in an air compressor would be a near analogy. Comparing a crankshaft to a piece of spaghetti would be a far analogy. (So would comparing an orthopedic implant to a banana peel).

The two psychologists were trying to test the hypothesis that the use of far analogies would lead to more radical, innovative design concepts. This sounds reasonable enough. However -- as most of us engineers have learned when testing our designs in the real world -- just because something sounds reasonable is no guarantee that it's true.

In order to test their theory, the psychologists recorded a series of design meetings among a team of 10 engineers. The engineers were tasked with developing a thermal printing pen as a toy for children. The design meetings were structured as brainstorming sessions focused on specific aspects of the product: how to keep the print head level in a child's hand, and how to protect the print head from damage. The brainstorming sessions were videotaped.

When reviewing the videos, the researchers counted each time an analogy was used or a new concept was introduced. They classified these analogies as either "near" (for example, comparing the thermal printing pen to another kind of pen) or "far" (for example, comparing it to a garage door opener). They also compared the conceptual distance between each new idea and the ideas the team had already discussed.

They found that far analogies didn't necessarily lead to far-out concepts. In fact, they found that concepts resulting from far analogies were more likely to be incremental iterations of previous concepts. However, far analogies increased the overall number of concepts generated.

What does this mean for engineers? "Out of the box" analogies may not always lead to radical departures from existing designs, but that's not necessarily a bad thing. They do help keep the design conversation moving along. In this way, they can lead to more, and possibly better, ideas. After all, the goal of design is not to come up with novel concepts. It's to come up with ideas that work. Using analogies, especially from outside of the problem space, can help us to do that.

Related posts: