sponsored links
TEDGlobal 2012

Michael Hansmeyer: Building unimaginable shapes

June 27, 2012

Inspired by cell division, Michael Hansmeyer writes algorithms that design outrageously fascinating shapes and forms with millions of facets. No person could draft them by hand, but they're buildable -- and they could revolutionize the way we think of architectural form.

Michael Hansmeyer - Computational architect
Michael Hansmeyer is an architect and programmer who explores the use of algorithms and computation to generate architectural form. Full bio

sponsored links
Double-click the English subtitles below to play the video.
As an architect, I often ask myself,
00:15
what is the origin of the forms that we design?
00:18
What kind of forms could we design
00:21
if we wouldn't work with references anymore?
00:23
If we had no bias, if we had no preconceptions,
00:26
what kind of forms could we design
00:30
if we could free ourselves from
00:32
our experience?
00:34
If we could free ourselves from our education?
00:36
What would these unseen forms look like?
00:40
Would they surprise us? Would they intrigue us?
00:43
Would they delight us?
00:47
If so, then how can we go about creating something that is truly new?
00:50
I propose we look to nature.
00:53
Nature has been called the greatest architect of forms.
00:56
And I'm not saying that we should copy nature,
01:00
I'm not saying we should mimic biology,
01:04
instead I propose that we can borrow nature's processes.
01:06
We can abstract them and to create something that is new.
01:10
Nature's main process of creation, morphogenesis,
01:14
is the splitting of one cell into two cells.
01:18
And these cells can either be identical,
01:22
or they can be distinct from each other
01:24
through asymmetric cell division.
01:26
If we abstract this process, and simplify it as much as possible,
01:28
then we could start with a single sheet of paper,
01:32
one surface, and we could make a fold
01:34
and divide the surface into two surfaces.
01:37
We're free to choose where we make the fold.
01:39
And by doing so, we can differentiate the surfaces.
01:42
Through this very simple process,
01:47
we can create an astounding variety of forms.
01:49
Now, we can take this form and use the same process
01:52
to generate three-dimensional structures,
01:55
but rather than folding things by hand,
01:57
we'll bring the structure into the computer,
01:59
and code it as an algorithm.
02:02
And in doing so, we can suddenly fold anything.
02:05
We can fold a million times faster,
02:08
we can fold in hundreds and hundreds of variations.
02:10
And as we're seeking to make something three-dimensional,
02:13
we start not with a single surface, but with a volume.
02:16
A simple volume, the cube.
02:19
If we take its surfaces and fold them
02:20
again and again and again and again,
02:22
then after 16 iterations, 16 steps,
02:24
we end up with 400,000 surfaces and a shape that looks,
02:27
for instance, like this.
02:31
And if we change where we make the folds,
02:33
if we change the folding ratio,
02:36
then this cube turns into this one.
02:38
We can change the folding ratio again to produce this shape,
02:41
or this shape.
02:45
So we exert control over the form
02:47
by specifying the position of where we're making the fold,
02:49
but essentially you're looking at a folded cube.
02:52
And we can play with this.
02:57
We can apply different folding ratios to different parts
02:58
of the form to create local conditions.
03:00
We can begin to sculpt the form.
03:03
And because we're doing the folding on the computer,
03:05
we are completely free of any physical constraints.
03:08
So that means that surfaces can intersect themselves,
03:12
they can become impossibly small.
03:15
We can make folds that we otherwise could not make.
03:16
Surfaces can become porous.
03:20
They can stretch. They can tear.
03:22
And all of this expounds the scope of forms that we can produce.
03:24
But in each case, I didn't design the form.
03:29
I designed the process that generated the form.
03:32
In general, if we make a small change to the folding ratio,
03:36
which is what you're seeing here,
03:41
then the form changes correspondingly.
03:42
But that's only half of the story --
03:46
99.9 percent of the folding ratios produce not this,
03:48
but this, the geometric equivalent of noise.
03:53
The forms that I showed before were made actually
03:58
through very long trial and error.
04:00
A far more effective way to create forms, I have found,
04:01
is to use information that is already contained in forms.
04:05
A very simple form such as this one actually contains
04:09
a lot of information that may not be visible to the human eye.
04:11
So, for instance, we can plot the length of the edges.
04:15
White surfaces have long edges, black ones have short ones.
04:17
We can plot the planarity of the surfaces, their curvature,
04:21
how radial they are -- all information that may not be
04:24
instantly visible to you,
04:28
but that we can bring out, that we can articulate,
04:30
and that we can use to control the folding.
04:33
So now I'm not specifying a single
04:36
ratio anymore to fold it,
04:38
but instead I'm establishing a rule,
04:40
I'm establishing a link between a property of a surface
04:43
and how that surface is folded.
04:45
And because I've designed the process and not the form,
04:48
I can run the process again and again and again
04:51
to produce a whole family of forms.
04:53
These forms look elaborate, but the process is a very minimal one.
05:08
There is a simple input,
05:13
it's always a cube that I start with,
05:14
and it's a very simple operation -- it's making a fold,
05:16
and doing this over and over again.
05:19
So let's bring this process to architecture.
05:23
How? And at what scale?
05:26
I chose to design a column.
05:27
Columns are architectural archetypes.
05:29
They've been used throughout history to express ideals
05:32
about beauty, about technology.
05:35
A challenge to me was how we could express
05:40
this new algorithmic order in a column.
05:42
I started using four cylinders.
05:46
Through a lot of experimentation, these cylinders
05:49
eventually evolved into this.
05:52
And these columns, they have information at very many scales.
05:56
We can begin to zoom into them.
06:00
The closer one gets, the more new features one discovers.
06:03
Some formations are almost at the threshold of human visibility.
06:06
And unlike traditional architecture,
06:10
it's a single process that creates both the overall form
06:12
and the microscopic surface detail.
06:15
These forms are undrawable.
06:20
An architect who's drawing them with a pen and a paper
06:23
would probably take months,
06:26
or it would take even a year to draw all the sections,
06:28
all of the elevations, you can only create something like this
06:30
through an algorithm.
06:32
The more interesting question, perhaps, is,
06:35
are these forms imaginable?
06:37
Usually, an architect can somehow envision the end state
06:39
of what he is designing.
06:42
In this case, the process is deterministic.
06:44
There's no randomness involved at all,
06:47
but it's not entirely predictable.
06:49
There's too many surfaces,
06:51
there's too much detail, one can't see the end state.
06:53
So this leads to a new role for the architect.
06:56
One needs a new method to explore all of the possibilities
06:59
that are out there.
07:03
For one thing, one can design many variants of a form,
07:05
in parallel, and one can cultivate them.
07:08
And to go back to the analogy with nature,
07:10
one can begin to think in terms of populations,
07:12
one can talk about permutations, about generations,
07:15
about crossing and breeding to come up with a design.
07:18
And the architect is really, he moves into the position
07:23
of being an orchestrator of all of these processes.
07:26
But enough of the theory.
07:29
At one point I simply wanted to jump inside
07:32
this image, so to say, I bought these red and blue
07:34
3D glasses, got up very close to the screen,
07:38
but still that wasn't the same as being able to
07:41
walk around and touch things.
07:43
So there was only one possibility --
07:45
to bring the column out of the computer.
07:47
There's been a lot of talk now about 3D printing.
07:50
For me, or for my purpose at this moment,
07:53
there's still too much of an unfavorable tradeoff
07:56
between scale, on the one hand, and resolution and speed, on the other.
07:59
So instead, we decided to take the column,
08:06
and we decided to build it as a layered model,
08:08
made out of very many slices, thinly stacked over each other.
08:11
What you're looking at here is an X-ray
08:15
of the column that you just saw, viewed from the top.
08:17
Unbeknownst to me at the time,
08:19
because we had only seen the outside,
08:21
the surfaces were continuing to fold themselves,
08:24
to grow on the inside of the column,
08:26
which was quite a surprising discovery.
08:28
From this shape, we calculated a cutting line,
08:31
and then we gave this cutting line to a laser cutter
08:34
to produce -- and you're seeing a segment of it here --
08:37
very many thin slices, individually cut, on top of each other.
08:41
And this is a photo now, it's not a rendering,
08:48
and the column that we ended up with
08:51
after a lot of work, ended up looking remarkably like the one
08:53
that we had designed in the computer.
08:56
Almost all of the details, almost all of the
09:00
surface intricacies were preserved.
09:02
But it was very labor intensive.
09:07
There's a huge disconnect at the moment still
09:10
between the virtual and the physical.
09:12
It took me several months to design the column,
09:15
but ultimately it takes the computer about 30 seconds
09:17
to calculate all of the 16 million faces.
09:20
The physical model, on the other hand,
09:23
is 2,700 layers, one millimeter thick,
09:25
it weighs 700 kilos, it's made of sheet that can cover
09:29
this entire auditorium.
09:33
And the cutting path that the laser followed
09:35
goes from here to the airport and back again.
09:37
But it is increasingly possible.
09:42
Machines are getting faster, it's getting less expensive,
09:44
and there's some promising technological developments
09:47
just on the horizon.
09:49
These are images from the Gwangju Biennale.
09:51
And in this case, I used ABS plastic to produce the columns,
09:54
we used the bigger, faster machine,
09:58
and they have a steel core inside, so they're structural,
10:00
they can bear loads for once.
10:03
Each column is effectively a hybrid of two columns.
10:06
You can see a different column in the mirror,
10:08
if there's a mirror behind the column
10:11
that creates a sort of an optical illusion.
10:13
So where does this leave us?
10:16
I think this project gives us a glimpse of the unseen objects that await us
10:18
if we as architects begin to think about designing not the object,
10:23
but a process to generate objects.
10:26
I've shown one simple process that was inspired by nature;
10:30
there's countless other ones.
10:33
In short, we have no constraints.
10:36
Instead, we have processes in our hands right now
10:39
that allow us to create structures at all scales
10:43
that we couldn't even have dreamt up.
10:48
And, if I may add, at one point we will build them.
10:51
Thank you. (Applause)
10:55
Translator:Morton Bast
Reviewer:Thu-Huong Ha

sponsored links

Michael Hansmeyer - Computational architect
Michael Hansmeyer is an architect and programmer who explores the use of algorithms and computation to generate architectural form.

Why you should listen

Classical architecture is defined by "orders" -- ways to connect a column to a building, to articulate the joining of materials and structural forces. Colloquially, these orders are based on elemental forms: the tree trunk, the plank, the scroll, the leaf. Michael Hansmeyer is adding a new elemental form: the subdivision algorithm. He turns his math and programming skills to making ornate, organic, hyperdetailed columns generated from lines of code and then comped up in cross-sections of cardboard, almost as if they're being 3D printed.

His recent work with cupolas and domes is even more mesmerizing, like looking deep inside an organic form of near-unbearable complexity. See images at digital-grotesque.com >>

The original video is available on TED.com
sponsored links

If you need translations, you can install "Google Translate" extension into your Chrome Browser.
Furthermore, you can change playback rate by installing "Video Speed Controller" extension.

Data provided by TED.

This website is owned and operated by Tokyo English Network.
The developer's blog is here.