The elegance of simple design

Simple is beautiful.

Amy Smith of the MIT D-lab has come up with a concept for a corn-shelling device for certain communities in developing countries. You can read all about it, including how to make it, here. I think this work is brilliant, and that the characteristics that make this design so good are key characteristics that should exist in every design, regardless of context.

Smith's device has distinct advantages over both automated corn-shelling machines and the de-facto standard, manual shelling. It is clearly in that sweet spot that covers usability, cost, sustainability, functionality, and manufacturability. It is, in other words, a very well balanced product.

There is an essential lesson here for all designers: simplicity is the root of designing for humans. And while not all design directly affects humans, the human impact of design is of growing concern, and rightly so.

The obvious, "Western" solution for shelling corn is a machine, made of refined metal and plastic, using either electricity or some fossil fuel to operate, and that can produce in one hour as much shelled corn as might be produced in a day by manual methods.

But the conventional solution is not well-balanced: it cannot be made or maintained locally; it is likely beyond the technical capabilities of many of its potential users; it is too costly compared to the money available to its users; it is over-productive for its context; and it has too great an environmental impact. So even if such machines are reasonable from the point of view of the "developed world," it is not a good fit for its actual, intended context. This makes such machines bad designs.

I think the key to Smith's design is that simplicity was, consciously or otherwise, given privileged status in her process.

I also think Smith's concept is an analogy to the human hand and how the thumb pries kernels from the cob, but this is a common feature of most corn-shelling solutions. What makes Smith's solution special is not its dependence on analogy, but rather its use of simplicity.

John Maeda wrote an interesting book, The Laws of Simplicity, that explains what Maeda thinks are fundamental premises of applying simplicity to design. You can read about those laws online here. A number of those laws appear to apply, at least a bit, in the case of Smith's corn-sheller. Indeed, I reckon that seven of Maeda's ten laws (excluding laws 2, 3, and 9) seem to apply.

The problem with Maeda's laws is that, on the whole, they're neither operational nor generative; that is, they don't tell you how to apply them and what to expect from their application. Indeed, if you think about it, it's hard to imagine how one might have taken those seven laws and applied them to the point of getting Smith's design.

Don't get me wrong; "laws" are good - they give us guidance when in doubt, and also, perhaps more importantly, they tell us when we're veering too close to a figurative cliff. (One might argue that Maeda's laws might be more scientifically described as "rules." But we'll leave that to another post.) Still, there's a huge gap between a rule like one of Maeda's laws (or any of the classic rules of engineering design - like "minimize the number of parts") and a method to actually apply that rule in a specific case. And in practice, rules are only as good as the methods that let you apply them.

I don't know exactly how Smith's design came to be, but I would like to suggest that, if one looks at the product variations (you can see the image here) shown in the MadeGood web page, one notices common features that are at the heart of the concept, and that are suggestive of method.

1. The product's operation is self-evident. That is, it has an affordance that relates directly to how it should be used. This requires understanding the user. One must observe how corn is shelled, and understand the context in which one does the shelling. The context is essential because it explains why the corn is shelled as it is. (In opposition, consider how confounding the machine version is on this point.)


2. It empowers, rather than replaces, a human. It is not at all evident that techniques of mass production are warranted in some settings, such as in developing countries. To detach the human from the act of even something as simple as shelling corn may also detach that human from their role in the community. There is honour in any honest work, even shelling corn. To replace a human with a machine in communities unaquainted to mechanization and its social impact essentially tells them they are not even worth as much as a machine. That can be a hard, disillusioning, and society-fracturing lesson. Far better is simply to identify those things that humans do well in a given context and give them simple tools to help them, where "simple" and "help" are defined by the users, not the designers.
   (It may well be that the mechanization of a society can happen at only some maximum rate, and that this relates to a society's ability to absorb and accept change in general rather than being a feature of technology. For instance, one could argue that the Apple Lisa was not successful because it expected too much of a behavioural change in the user community. A non-technological example could be the social acceptance of same-sex marriage, which, I think, will one day be accepted universally in modern cultures, but which cannot be foisted quickly upon the general population without causing possibly extreme resistance.)


3. It is derived from local materials and manufacturing techniques. Keeping things local is a good way to improve effectiveness. Systems (i.e. a community) that require only local skills and materials tend to be far more flexible to contextual changes than systems that are at the mercy of distant (and possibly gargantuan) services that are not necessarily always acting in the local community's best interest. The distant service provider may decide that it is in its own best interest to act counter to the local community's best interests. A local system, on the other hand, is inherently tied to the community in a more positive way. The local system works in its own best interests, which is far more likely to benefit the community. Also, parachuting technology into a community can have a debilitating effect on the community's sense of independence. So local materials and manufacturing techniques should be constraints imposed on all designs.
   This also relates to Don Norman's notion that technology precedes the identification of needs. If technology can drive the identification of need - and I believe it does - then identifying the real need of developing communities is based on the technologies that they have available, not the technologies to which we are accustomed.


4. It requires a minimal change in behaviour. Large, automated corn-shelling systems may be more productive, but they also require substantive changes to the community's structure and operation. What do those who shelled corn do now that the machine outproduces them all? Who tends to the machine? What impact will increased shelling rates have on the availability and raising of the raw corn? Of course, there are answers to all these questions - even simple answers, from our point of view. But ours is not the point of view that counts. The answers to these questions may cause a huge shift in the nature of the community, which the community itself may not be willing to endure. Smith's corn-sheller, on the other hand, requires no dramatic community changes, but still improves the speed with which corn is shelled (compared to manual techniques) and addresses key human factors issues with regards to hand, wrist, arm, and back injury that can occur using only manual techniques. So, a good design is one that is sensitive to what people do, and that is least "disruptive," where the limit of acceptable disruption is defined by the users, not by the designers.

It's important to note that the characteristics that (I think) set Smith's design apart from the alternatives don't apply only to products for developing countries. These are, I think, characteristics of any good design.

I want to emphasize this: we should design everything as if we were designing for a so-called "developing country" because in doing so we lose all the flotsam and jetsam, all the complicated rhetoric and narrative, all the ideological posing and artifice of tradition.  And all that's left is design's most basic essence: to improve the world.

So here's what I think: we should all aspire to imbue our designs with the characteristics I've noted in Smith's corn sheller, we should use these characteristics as measures of balance of our designs, and we should assess student design work with respect to them too.