The III rd Generation of
The INDUCTION CITIES / INDUCTION DESIGN
WEB FRAME / Program Generated Architecture 2000
applied to SUBWAY STATION / IIDABASHI 2000
The significance of Induction Cities lies in the search for better
solutions to given conditions.
What were the conditions that Web Frame had to solve ?
There were three issues:
1. Restrictions on space.
2. Conditions imposed by each component.
3. The extension of the given space.
The first of these was an absolute condition allowing no margin
for improvisation, just as one cannot choose the site when designing a building above ground.
As for 2., any variety of forms and quantities is possible under a
Computer Graphic simulation, but in reality, conditions are imposed
by the kind of installation that can be carried out. For example, it is difficult to achieve an intersection at the same point of five frame tubes with an angular variation of one degree each. Individual parameters were established to allow for automated clearance of such specific conditions.
This is essentially the same kind of task as designing structural
frames for conventional architectural work.
The third condition -- spatial extension -- became another
parameter. By specifying the approximate position and volume of component parts, the desired space is generated.
This is a flexible specification.
It is a lot of work to develop a program that will satisfy just
these three conditions. Several attempts were needed to get it right.
Even an automated program for designing a free frame "closed"
in three dimensions turns out to be difficult. There are restrictions on the solid angles that can be employed, and all points must be joined together.
The issues here are different from those of conventional space
frames assembled in regular fashion from materials with fixed angles.
Simply because the degree of freedom is great, divergences can occur and lead in unpredictable directions.
Freedom can, of course, readily slip over into chaos.
But an important element of this concept is to give the
appearance of chaos while in fact obeying certain regularities.
While the result may appear to be arbitrary and willful, the
necessary conditions are rigorously met.
The same can be said of chaos and of all forms of complex phenomena.
The coexistence of freedom and harmony !
This sounds like a catchphrase put forth at some kind of meeting by people fully aware that such a thing will never come about in reality. But this is not an empty slogan.
We are (just) beginning to see signs that it can be realized.
Introducing Arbitrariness / Returning from "Design-less" to Design
With the Web Frame project, we have moved forward from the
first phase of Induction Cities into the field of 'esthetic' evaluation.
That is our fourth objective.
In the first phase, we selected as the basis for our criteria of
evaluation such quantifiable variables as exposure to sunlight,
distance, gradients, wind speed and resistance, etc.
In the case of the City of Generative Neighborhoods project, we defined "interesting" by means of certain formulae and thus introduced a factor of sensibility, but we were not evaluating whether the resulting plans were after all interesting or not.
With the Web Frame project, however, we tried to go beyond
the principle of randomness (to which we have thus far adhered) and bring into play some measure of arbitrariness.
By arbitrary, I do not mean that we are inputting directly
specifications for factors such as space or forms.
What we intend rather is a program to satisfy "fuzzy" criteria such as "enjoyable" or "dynamic."
The designer's hands, tied up until now, will begin to move, just
But the hands in question are not human -- they are artificial.
At this point, however, we ran into a surprising (though not
entirely unforeseeable) difficulty.
The method used for the programs for the City of Generative
Neighborhoods allowed for the definition of "enjoyable" on the basis of specific attributes, but the results did not meet our expectations.
One reason for this was the complex three-dimensional spaces
and forms in which the Web Frame had to unfold.
Another factor was apparently the rigid spatial limitations of the available site.
It seems that methods based on complexity theory cannot
become really effective without ample space for implementation. The practical results of natural selection, for example, can only begin to appear in wide savannas or large oceans where numerous species of life have room to live and compete.
Also, by contrast with evaluation criteria based on clearly
definable indices, viable indices for the matters of sensibility or
feeling are difficult to pin down.
And of course implementable designs are more difficult to
achieve than research results. We are not playing a game like SimCity.
After any number of initial efforts, the program did not seem
able to deliver the kinds of good solutions we were hoping for. We were stuck.
Living Creatures / Self-evolving Programs
At this point, we have to return to our earlier question, what is a
"good" thing ?
In City of the Sun Goddess, we chose as an index for evaluation exposure to sunlight, and in On Demand City, our index was distance.
Except for the requirement of meeting these conditions,
everything was randomized.
We made a point of not manipulating the output of the programs.
The program for City of the Sun Goddess generated an
aggregate that looked like a natural colony.
In the case of On Demand City, the resulting plan for location of facilities was similar to that of a naturally occurring town.
In both cases, that is, some aspect of "naturalness" showed up.
Naturalness is something that everyone can understand.
By excluding the intentions of a designer and letting criteria of
the form of physical laws determine the outcome, a plan with all the "persuasiveness" of a natural phenomenon was generated.
Would it be unfair to call that persuasiveness "beauty" ?
Let me put it this way: It is the physical laws underlying their
regularity that cause us to feel that snowflakes or the waves on the surface of a river are beautiful.
The basic principles outlined above (1, 2 and 3) are at work in
the Web Frame project, as well. But the results vary widely depending on how the parameters are established.
There is a large margin for instability. In short, it is close to chaos. "Naturalness" does not emerge.
If a large number of parameters are combined and finely tuned
in the pursuit of naturalness, an enormous amount of trial and error is required. In practice, such an approach is not feasible. We end up making the best of what comes out and giving up.
At this rate, there is no much difference from designing the whole plan by hand.
One means of avoiding this kind of impasse is to incorporate
laws of "nature" in the program.
Why not introduce some principle from nature -- for example,
the laws of motion governing the movement of waves -- that gives such a sense of pleasure?
If dynamic force can create for us rational and beautiful patterns,
then it should be enough to add to our program a simulation of
There is a history to this line of thinking. When Gaudi
suspended weights from inverted models to make decisions about designing, he was in effect performing an analog computer simulation.
The same can be said of the use of soap bubbles, in the 1960s,
as the basis for designing the structure of membranes.
Today, we don't have to use either weights or soap bubbles.
We can use Navier-Stokes equations and deploy supercomputers to simulate fluid dynamics. But the literal application of natural laws looks tediously like mere imitation of nature.
Induction Cities is not seeking to reproduce natural phenomena.
It would be more to the point to incorporate principles which
don't bear directly on the requirements.
We began searching for effective code that would be both more
specific and simpler.
At the same time, we began examining the possibilities for
This other approach was to have the program search for its own
The program is run and then its output evaluated by human
The results are scored -- are they satisfactory, or not quite good
When this process is repeated often enough, the program,
instead of simply outputting more plans, begins to generate plans which are likely to receive higher scores.
If you praise the program, it learns… "AIBO" was a first step in this direction.
If the process continues long enough, the solutions output by
the program should improve markedly -- in theory at least.
The idea is to create a program which is based on this
What is interesting about this is that the question of what is
"good" is never given a clear answer.
(It is true, of course, that if the results obtained by this process
were analyzed, it would be possible to get a clear picture of the values involved. What you are seeking to do is just this… A table of evaluative criteria is drawn up.
For the Induction Cities project, the mechanisms for devising a program are in principle also the means for analysis.)
As if by magic, good plans are generated, even while the
criteria for evaluation are not clarified. This is our trump card for
escaping the impasse of making value judgements.
Learning functions for software in simple form are built into
word processors, today.
If we pursue this idea further, to the point that the program
learns to modify itself, there should be no objection to calling this an "evolutionary function.
" What we do call it should depend on how advanced (smart) the program really is.
For these purposes, inheritance algorithms are also useful.
This program is still undergoing development.
Structure-generating Program / Wing
The Architectural Seed germinates deep in the ground, 35
meters below the city. (Seeking more water, more light…)
After a time, its underground stem reaches the surface and there, a flower blooms.
This is called "Wing."
Wing is a ventilation tower. It houses the ventilation and air-
conditioning equipment for the entire subway station: a respirator for the space below the ground.
Wing is the respiratory organ put forth above ground by an
invisible, subterranean stem.
For its structure, we sought a mechanism of auto-generation
using a computer program. The program is not yet completed,
(→ completed : KeiRIki-1program 2005 / ShinMinamata MON 2006)
What we show here is a model of what the structure will perhaps look like once the program is operative.
We attempted here to incorporate the structural dynamics --
something which was not a condition to be solved by the program for Web Frame -- and made this a primary condition.
To design a conventional structural frame, a simple grid-work
is devised, weight is applied, and the effects are calculated. Proper materials are selected to meet the load requirements of those portions under greatest stress. The same materials are then used throughout the frame. This is true for both rigid frames and tubes, and regardless of whether the shape is a box or is curved.
But if materials are selected not by this uniform rule but varied
from section to section as actually necessary, a different form of frame will appear.
And if we substitute the word "design" for "necessary," still
other forms will emerge. This is what I have attempted to do with
The framework is thick and large which forces are greatest, and
thin where forces are weak.
Materials are fused at the joints to better withstand transmitted
Instead of joining pillars to beams, the material extends,
separates, rejoins and forms a single overall frame without distinctions between verticals and horizontals.
Moreover, the arrangement of structure and material is
optimized so nothing is superfluous.
Its structure is that which has already been achieved by living
Wing is a model of one form that frames will take when such
structures can be generated automatically.