Friday, March 23, 2007


"the AA autrement"... is a set of abstracts rendered while figuring out the explicit and encoded rules set for LightHive "diffuseur grammaire": one could argue that each drawing is acting as some sort of map of the AA school, where the representation devices are a set of unique geomtries -carefully labelled- showing directions of obstacles, accidents and other folds within the terrain of every room...

Simplified as sets of closed polysurfaces, the 3D model allows the code to check for every room which light sources -selected as a point cloud- are enclosed within...

One of many shape... each arrows is growing along the direction of a vector pointing toward any corner of the room boundaries, or furniture piece...

Labels: , , , ,

Tuesday, March 20, 2007


Latest "abstracts" from the development of "LIGHTHIVE"

Lighthive is Alex Haw (Atmos) interactive exhibition at the AA, London (opening early May) for which Marc Fornes & theverymany have been invited to collaborate and write an explicit generative process abble to automatate the shape grammair of most of the 1500 components required to map the exisitng light condition within the AA building.

Each of those "3D star" component is an assembly of two 2D star which can be unrolled and laser cut flat onto thick sheet of acrylic.

INPUT (rhinoscript)
-get the 3D model of the AA school as a set of closed polysurface (each polysurface is one room including the boolean of the furnitures)
-get all the light source modelled at their scaled position as a point (color according to light type)

PSEUDO CODE: at the scale of the phenotype, here the 3D model of the school:
- for every pt (lightsource) do
- if pt is within closed polysurface do
- create two planes through each light (orthogonal to each other)
- for each plane do
- intersect plane and polysurface
- for each intersection curve do
- if pt is within curve do

COMPONENT: at the scale of the component "genotype", the challenge is to allow replication, proliferation while developping fondamentales genetic modifications such as custom amount of "spikes", each of those "spikes" having custom lenght, some "spikes" being able to bundle according to neighbourhood condition, etc, etc...

Those components are not displaying any swarm behaviour or collective intelligence, neither is based on dynamic feedback... BUT the generation of each "spike" is following a so long list of rules (possible target tests, few distance tests, many angle tests, etc, etc...) and possible branching transformations according to which there are able to develope and mutate... sometimes up to the point of displaying extreme behaviours impossible to expect... once more "very many" allows or describe a blurry boundary between "automatisation" and Emergence... between Repetition and Intelligence? blurry...

Labels: , , ,

Tuesday, March 13, 2007

070312_HoneyComb (re-scripting)

A plane-filling arrangement of plane figures or its generalization to higher dimensions. Formally, a tiling is a collection of disjoint open sets, the closures of which cover the plane. Given a single tile, the so-called first corona is the set of all tiles that have a common boundary point with the tile (including the original tile itself).

Wang's conjecture (1961) stated that if a set of tiles tiled the plane, then they could always be arranged to do so periodically. A periodic tiling of the plane by polygons or space by polyhedra is called a tessellation. The conjecture was refuted in 1966 when R. Berger showed that an aperiodic set of 20426 tiles exists. By 1971, R. Robinson had reduced the number to six and, in 1974, R. Penrose discovered an aperiodic set (when color-matching rules are included) of two tiles: the so-called Penrose tiles. It is not known if there is a single aperiodic tile.

As part of a re-visiting codes series -most often following students requirments- I went back to one of my very first rhinoscript code generating Honeycomb cells onto a host nurbs surface and rewrote it... mainly shorter, for sure much faster and somehow within a concern of a certain "elegance" -to quote the latest trend of fitness criteria accoding to Ali Rahim's latest issue of AD magasine!- as an exercice within the exercice...

Tips and tricks:
The generation of the honeycomb cells within rhinoscript is now defined by one unique conditional statment using many booleans operations.
If (i>1 And j>1) And ( ( ((i-1)Mod 2) And (((j-2)Mod 4)= 0) ) Or ( (((i-1)Mod 2)=0) And ((j Mod 4)=0) ) Or ( (((i-1)Mod 2)=0) And (((j-1)Mod 4)=0) ) Or ( (((i-1)Mod 2)) And (((j+1)Mod 4)=0) ) )

The axes of honeycomb cells are always quasi-horizontal, and the non-angled rows of honeycomb cells are always horizontally (not vertically) aligned. Thus, each cell has two vertical walls, with "floors" and "ceilings" composed of two angled walls. The cells slope slightly upwards, between 9 and 14 degrees, towards the open ends.

There are two possible explanations for the reason that honeycomb is composed of hexagons, rather than any other shape. One, given by Jan Brożek, is that the hexagon tiles the plane with minimal surface area. Thus a hexagonal structure uses the least material to create a lattice of cells with a given volume. Another, given by D'Arcy Wentworth Thompson, is that the shape simply results from the process of individual bees putting cells together: somewhat analogous to the boundary shapes created in a field of soap bubbles. In support of this he notes that queen cells, which are constructed singly, are irregular and lumpy with no apparent attempt at efficiency.

It is likely that the honey bee constructs the honeycomb based on instinct, and the prevailing theory of biology is that the appearance of such efficient shapes in nature is a result of natural selection.

The closed ends of the honeycomb cells are also an example of geometric efficiency, albeit three-dimensional and little-noticed. The ends are trihedral (i.e., composed of three planes) pyramidal in shape, with the dihedral angles of all adjacent surfaces measuring 120°, the angle that minimizes surface area for a given volume. (The angle formed by the edges at the pyramidal apex is approximately 109° 28' 16" (= 180° - arccos(1/3)).)

The shape of the cells is such that two opposing honeycomb layers nest into each other, with each facet of the closed ends being shared by opposing cells.

Individual cells do not, of course, show this geometrical perfection: in a regular comb, there are deviations of a few percent from the "perfect" hexagonal shape. In transition zones between the larger cells of drone comb and the smaller cells of worker comb, or when the bees encounter obstacles, the shapes are often distorted.

In 1965, László Fejes Tóth discovered that the trihedral pyramidal shape (which is composed of three rhombi) used by the honey bee is not the theoretically optimal three-dimensional geometry. A cell end composed of two hexagons and two smaller rhombuses would actually be .035% (or approximately 1 part per 2850) more efficient. This difference is too minute to measure on an actual honeycomb, and irrelevant to the hive economy in terms of efficient use of wax, considering that wild comb varies considerably from any human notion of "ideal" geometry.

Labels: , ,