Nanotech, Black Holes and Chocolate

Sept. 27, 2007
Key to our understanding of nanotech is the ratio of surface area to volume.

By Dick Morley, control columnist

COFES (the Congress on the Future of Engineering Software) invited me to give a “disruptive talk” at its April 2007 meeting.

The result of this disruption? “Speculations on Thinking Nano—a Meditation on Black Holes, Nanotech, Chocolate and the Internet.”

The aim of my talk was to show that the contents of most processes are irrelevant. What’s important to the user is the surface or the interface.

Key to our understanding of nanotech is the ratio of surface area to volume. This ratio is also key to chemical reactions and social interaction, and it is the determinant of information in a black hole.

Nanotech is providing us with orders of magnitude of growth in surface-to-volume ratio. Somewhat surprisingly, this surface-to-volume issue may also apply to software, companies and to chocolate. Entropy and information theory are both applicable to making better software products.

Nano is a bottom-up technology. The interface to stars, computers and companies are all related. Examining the broad spectrum of interfaces is also key to software design. The physics of the black hole is key to understanding our changing world, especially in the burgeoning international engineering environment.

All information concerning black holes is at the surface (the event horizon). The volume of the hole or its contents are poorly understood and irrelevant. Since the surface is two dimensional, and the “volume” has a cubed function, the ratio of information to volume is a key consideration. All we need to know is at the surface.

Nanotech has similar characteristics. Carbon tubes have great strength. Why? At the molecular level, all characteristics are surface, not content-related. The strength of beams is concentrated at the surface. Current flow in copper wire occurs at the surface because electrons like to travel alone. The new A123 MIT batteries for portable equipment get awesome performance by removing most of the support structure to increase the surface-to-volume ratio.

Chocolate is another example. Most off-the-shelf chocolate has to have some encapsulation mechanism to increase shelf life; otherwise, it would deteriorate rapidly. Edible wax is one method of encapsulation. Adding it to the recipe or physically surrounding the candy helps. The wax is tasteless, but the absorption into the taste buds occurs only after the wax seal is dissolved. Another example is coffee. We make coffee not with a single bean in hot water, but with special grinds. The point is to make everything as small as possible with a maximum surface area.

What has this to do with software? Early designs had a single function per “machine.” As software design matured, it took a page from the hardware guys and used separate autonomous designs for each application. Defining the interface for each application is defining an event horizon. I’m writing this column using OSX on my Mac. Contamination does affect individual components, but the machine keeps on trucking. The new alphabet soup of software techniques embrace independent surface interaction while keeping the internals “irrelevant.”

How does this view change our thinking? Process engineers should think about the black-box approach to automation design. Game designers should only worry about the user interface. How and what generates the surface is irrelevant; only the output is of concern. A process plant is a black box with inputs and outputs. All information about this process is from measurements that are at the “surface” of the box. Each of the sub-processes should be treated in the same manner—boxes upon boxes, each treated as an autonomous black hole with its own event horizon.

I came across a wonderful book on this subject; The Ideal Enterprise, Managing by the Law of the Sphere by Hans D. Baumann. He has data on the enterprise relationship to its surface. The surface is the customer/user. The implication is that profit is related to company size, and profit per employee is predictably based upon the ratio of volume to surface. All this implies that performance is a given and NOT dependent upon executive talent or a charismatic CEO. As in all things, exceptions are infrequently present. Think about what systems DO, not what they are.

Tread carefully. Small wins; top-down does not.