|
channelbeta - canale d'informazione sull'architettura
contemporanea |
|
Ice_FLOC [Iceland Fuzzy
Logic Operations Center] Somewhere
between true and false, black and white, yes and no, exists a third region where
outcomes emerge on a sliding scale. In 1965, Lotfi Zadeh articulated fuzzy logic
in his paper, "fuzzy sets," where he questioned the degree to which something
is true or false. The aim of fuzzy logic is to provide a very precise approach
for dealing with uncertainty. Accordingly,
the degrees of uncertainty in predicting weather lend themselves to the methods
of the fuzzy inference system. The [FIS]
identifies trends of past performance and reinvests this information to generate
more precise forecasts. The more the model is applied, the more accurate the forecast.
When patterns re-emerge after long dormancies, we find this information already
invested in the system. Triggers surrounding the event indicate when the setting
is ripe for a repeated incident. The cyclical flow takes into account not only
quantitative data, but qualitative information that consults everything from the
farmer's almanacs to local fisherman's lore. Located
in the fishing village of Grindavik, home of the blue lagoon geothermal spa, Ice_FLOC
is accessible from both the Keflavik international airport and the city of Reykjavik.
This location is an ideal spot for monitoring coastal weather patterns as well
as accessing the pool of native research scientists and attracting transatlantic
corporations and governments interested in forecasting trends. The premise of
the center is to cross-pollinate various forecasting teams within a structure
where they can optimally learn from each other's fuzzy logic investigations. Applying
these methods to design, I followed the FIS model to create my own fuzzy inference
system. By deploying a Maya physics model, distinct bodies, linked together, reacted
to one another over time. By assigning speed to program, the tendencies in program
aligned themselves with tendencies of bodies. The sections generated over time
provided an abstract model which could then be re-sequenced according to overall
flow in the building. Gavin
Bardes Columbia
University Graduate School of Architecture
[GSAPP] Fall 2000 studio: Iceland
Weather Monitoring Station Studio
critic: Winka Dubbledam Phase
1 input Phase
2 fuzzification Phase
3 Inference procedure Phase
4 fuzzy rule base Phase
5 defuzzification Phase
6 output |
|
channelbeta COPYFREE |
|
[03-2003] |
|
Pictures provided by
the author |
