Exploratory Engineering:
Applying the predictive power of science
to future technologies

by Eric Drexler on June 26, 2009

While I’m on the subject of foundational concepts in the relationship between science and engineering, here’s the outline of a methodology for applying current science to assess lower bounds on the capabilities of a select subset of future technologies. (As many of you know, some of those lower bounds are startlingly high.)

A subset of the potential capabilities of future levels of technology can be understood by means of a design process that can be described as exploratory engineering. This process resembles the first phase of standard design engineering (termed conceptual engineering, or conceptual design), but it serves a different purpose:

  • In standard engineering, design leads to the manufacturing of a product.
  • In exploratory engineering, design leads to understanding of what a future manufacturing process could produce.

Because of this difference in objective (products vs. knowledge) conceptual engineering and exploratory engineering have both similarities and differences:

 

Conceptual engineering

Exploratory engineering

Objective

Create and evaluate system-level designs (typically parameterized)

— Same —

Background technology

Known capabilities delivered by current fabrication technologies

Conservative, physics-based estimates of capabilities delivered by future fabrication technologies

Process of analysis

Identify required subsystems and components, and their required performance parameters

— Same —

Basis of calculations

Typical performance of available materials and components
(Design may exploit poorly understood phenomena)

Conservative, physics-based estimates of lower bounds on the performance of feasible materials and components.
(Design must avoid poorly understood phenomena)

Results of calculations

Estimates of the performance of fully refined, system-level designs

Lower-bound estimates of the performance of fully refined, system-level designs

Production and performance

Product must be manufacturable and competitive today

Product is not manufacturable today, and need not be competitive in the future

Implication for design criteria

Seek efficient configurations to maximize performance and minimize cost of production

Seek simple designs and assume large design margins to maximize confidence and minimize cost of design analysis

Results of the design process

Choice of system-level design concepts for refinement and possible production

Estimates of the capabilities of future levels of fabrication technology for evaluation of research and policy options

In the early 20th century, a missing fabrication technology was the combination of engineering expertise and metalworking techniques (among others) that were required to build large aerospace vehicles. The physics of rocket propulsion, however, were well understood, and the strength and weight of large, well-made aluminum structures could be estimated with reasonable accuracy.

On the basis of exploratory engineering applied to this kind of knowledge, engineers who studied the matter were confident that orbital flight could be achieved by means of multistage chemically fueled rockets. By the 1940s, a study by the British Interplanetary Society had filled in considerable detail and given a good estimate of the size of a vehicle that could reach the Moon.

Those who hadn’t studied the matter were sometimes confident of the opposite.

Astronomers who knew the daunting speed of orbiting objects (but didn’t trouble to examine the exploratory engineering analysis) sometimes called spaceflight absurd. In a characteristic move, a Professor of Physics and Chemistry, one A. W. Bickerton, demonstrated the impossibility of something irrelevant, then generalized it: He did a silly calculation that implicitly assumed that a rocket must put its fuel into orbit (and a bad fuel, at that), then announced that spaceflight was simply impossible.

Reasoning from the unworkability of a bad idea to the impossibility of a well-researched engineering concept is a repeated theme of interactions at the interface between science and engineering. (Engineers ignoring a crucial scientific fact is, of course, another.)

Here’s an entertaining list of quotations from the history and prehistory of flight, and spaceflight.


See also:


   StumbleUpon button

{ 14 trackbacks }

Predictive Power « Just Thinking . . .
06.27.09 at 4:50 am UTC
the Foresight Institute » Super-dense magnetic memory
06.30.09 at 1:01 am UTC
Everything News Portal! » Blog Archive » Super-dense magnetic memory
06.30.09 at 2:36 am UTC
What is simple? Polyethylene, molecular modeling, and molecular machines
07.08.09 at 1:50 pm UTC
How to Learn About Everything
07.10.09 at 9:02 am UTC
How to Understand Everything (and Why)
07.10.09 at 9:07 am UTC
A Map of Science
07.10.09 at 9:09 am UTC
Evolutionary Capacity: Why organisms cannot be like machines
08.03.09 at 2:53 pm UTC
Asia and the elements of innovation
08.09.09 at 12:27 pm UTC
Knowledge about Knowledge: The most popular posts in the first year
10.29.09 at 5:29 pm UTC
Quantum-coupled single-electron thermal to electric conversion
12.04.09 at 9:06 am UTC
How many minds produce knowledge (and how they don’t)
12.11.09 at 12:39 am UTC
The promise that launched the field of nanotechnology
12.26.09 at 5:52 am UTC
Molecular Manufacturing: The NRC study and its recommendations
01.08.10 at 9:07 pm UTC

{ 2 comments… read them below or add one }

Jeffrey Soreff 08.29.09 at 5:33 pm UTC

Would you consider “Heteroaromatic Rings of the Future” (William R. Pitt*, David M. Parry†, Benjamin G. Perry‡ and Colin R. Groom, J. Med. Chem., 2009, 52 (9), pp 2952–2963
DOI: 10.1021/jm801513z) to be exploratory engineering? They aren’t specifying the fabrication technology in detail, but do select a large subset of the skeletons as synthetically tractable. In the application to medicinal chemistry, simply increasing the number of viable alternatives is a help, so in
that sense they are improving a lower bound on performance.

Eric Drexler 10.31.09 at 8:35 pm UTC

Yes, of a sort — but a rather unusual sort.

Leave a Comment

You can use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>

Older post: The Antiparallel Structures
of Science and Engineering

Newer post: A Renaissance Weekend