In the past I have written a few things about Occam's Razor and how, to paraphrase, it is stupid. Recently Rebound has been diving head first into our work in Mozambique. While we have been there a very striking example of Occam's Razor has manifested itself. In short, what has struck me is the way people developing and funding technology for the developing world tend to subconsciously oversimplify the issues the developing world face. In lieu of this oversimplification, we default to an Occam-style approach to tech development: make it simple, make it stand alone, and make it clever. Ironically, this approach often leads to technology that is inappropriate, expensive, hard to upkeep, and just down-right silly.
I will get to Rebound's experience in a second, but I want to highlight one example that has received a bunch of press in the developed world. Enter, the Soccket: a soccer ball that charges a small internal battery that can then be used to provide light at night. This solution defines the classic problem with developing world technology: it pre-supposes that the developing world is a blank slate where a technology must be developed from the ground up and handed out from the back of a truck.
The developers then are faced with a series of questions about how their design should be brought to fruition. Occam's bias comes in and tells them it must: operate in a technical vacuum, require no learning on the part of the user, and never break. These are, of course, impossible, but the designers of countless developing world technologies have striven to meet them.
The result is a technology like the Soccket. At first glance it seems extremely clever. But then you think about its application and you realize that it is going to be significantly more expensive than many other ways of solving the same problem and will, be its nature, be a gift to the developing world.
I only have experience in one part of the developing world, but I know that in Mozambique one could go to a local market at a smaller town and buy the components to build a solar powered, battery equipped light and put it together themselves in a day. This light would cost less than the Soccket, charge itself during the day, and could be fixed at any point by another trip to the local market.
So you may ask, if there is a simpler way of doing it why is nobody doing it this way. The only answer I have for you is that, simply put, nobody is doing it this way. This gets right down to the crux of the argument. Engineers in the developed world have an incorrect view of the developing world where they believe their solutions must operate in a stand alone, simple, Occamly defined way. But the fact is, there is a LOT going on in these regions of the world. There are VERY smart people already living there. And they are HIGHLY motivated to make their lives better. Why not make technology with the assumption that it will rely on the equipment already in the developing world using the smart, if less educated, people that live there. Don't make your technology simple to use, make it simple to fix, build, and learn from. And don't make it solve as many problems as it can. Make it solve one problem and rely on other technologies for the rest.
So what is Rebound's deal? The only reason I feel strongly about this is because, 1 year ago, I was that silly engineer designing a technology to work in a vacuum. I was making a refrigeration system that would solve all the refrigeration problems of the developing world. After being in the rural Inhambane province of Mozambique for about 10 minutes it occurred to me how stupid that was.
The problem about refrigeration in this rural setting was not that there was no technology available to provide the cooling. It was that nobody was using it to provide the cooling. There was no vacuum in which to develop a technology. Everywhere we went we saw houses with small air-conditioners. We saw PV panels for sale in even the smallest road-side markets. And we met people who knew how to keep both running using only parts available locally.
It took us just one night of very little sleep under a thatched roof and mosquito net with a laptop to model a new system. This system would come to be called SunChill. It is not a complete solution to horticultural refrigeration. In-fact, without some kind of cold storage on its back end, it won't really help anybody. The that is just the point! It does one thing: enable the use of the technology that already exists on the ground. It does so using parts that we can buy in this hardware store in Inhambane:
SunChill is not a simple technology. It has many parts, a complex set of operation modes, and many technical risks that still stand in its way. But fundamentally it is a technology that assumes it will be operating in the real world. It is designed around the supposition that its users are intelligent and that other technology can be used to solve the rest of the problems with horticultural cooling. Will we pull it off? Who knows.
In the end, SunChill is just another example. There are too many factors that lead to things failing or succeeding to say it will definitely work. But the point is that, at the very least, SunChill is being designing knowing that there are people and products already in Mozambique that we can rely on. There is no need to design for a vacuum that does not exist.
No comments:
Post a Comment