Data you can hold: Physical model of a cell’s movement on a flat plane, produced on a 3-D printer.


Jan|Feb 09 Contents
Gazette Home


“We are approaching an era of these way-beyond-gigabyte quantities of data from individual patients or cells,” Peter Jones says. “How the heck can you possibly use traditional methods to number-crunch and display those data? You probably can’t. So can we do it by looking at form and patterns?”

What design students bring to LabStudio’s table are the skills to winnow and sift through the vast mountains of information to bring relevant patterns to the surface. That may sound hopelessly abstract, but in a sense, it is an architect’s most basic job description. “Design is an act of filtering,” Sabin says. A building begins as a list of demands and constraints that essentially takes the form of raw data: the lot size, the number of square feet needed for a variety of activities, the budget, the cost of materials, zoning restrictions on height and shape, fluctuating plumbing and lighting and energy needs … and on and on. “You’re constantly navigating all of these different constraints and parameters that make up a project, because in the end, the design is a synthesis of all of them.”

As the list of parameters has grown to encompass an increasingly complex set of requirements—everything from minimizing construction waste to maximizing a building exterior’s capacity to channel air flow toward roof-mounted wind turbines—digital design tools have changed radically.

“You may think that designing with a computer involves using a mouse,” Detlef Mertins says. “And it does, if you’re using AutoCAD and other conventional software. But it’s also possible to design by writing script or code—that is to say, you do it numerically through functions and algorithms.”

The most recognizable product of this approach is probably Beijing’s National Aquatics Center, where Michael Phelps hauled in his eight gold medals at the 2008 Olympic Games. The Watercube derives its iconic façade from a set of mathematical equations describing the structure of soap bubbles. Each of the building’s four walls is nearly as long as two football fields, yet they alone support the gigantic, 7-acre roof. “Over such a wide span of column-free space, the need to minimize the self-weight of the structure is paramount, as most of the structural work involves ensuring the roof can hold itself up,” architect and writer Michael Weinstock observed in the journal Architecture Design. The laws of gravity and torsion, in other words, become tricky parameters indeed. So the shapes and locations and thickness of the 4,000 weight-bearing “bubbles” that make up the Watercube’s structure were determined by the meticulous refinement of digital scripts and algorithms—which can be converted seamlessly into factory instructions. “At the scale of very large architectural projects,” Weinstock continues, this process “becomes not only the significant design strategy, but also the only economic means of reducing design data for manufacturing.”

In an analogous way, Savig applies scripts—albeit “simple” ones, she stresses—to her data in Rhino. Only rather than deriving a structure, she is creating novel graphic representations that make relevant patterns legible.

What Peter Jones wanted, and Jenny Sabin hoped to provide along with her students, was a new way of seeing. They are starting to get it.

“This isn’t immediately fundable,” Jones says. “But it’s starting to move toward some sort of acceptance. If we’d put this in two years ago to the NIH, it would have been a joke to them. Now they’re crying out for new modes of visualization. And so is the NSF. So there’s a paradigm-turn occurring, I think.”

In some ways, it is a paradigm shift that the art world has anticipated. Over the summer Sabin exhibited a different LabStudio product at SIGGRAPH, an annual exhibition of computer graphics and interactive design that features the work of world-renowned architects like Frank Gehry and Zaha Hadid. The piece, “Branching Morphogenesis,” is an abstract sculptural rendering of lung endothelial cells interacting within a three-dimensional matrix environment. Made from some 75,000 color-coded cable zip ties, it comprises five interwoven curtains 12 feet high, 15 feet wide, and eight feet in depth. The curator of the Ars Electronica Museum in Linz, Austria, recently selected it for an exhibition celebrating that city’s stint as the 2009 European Capital of Culture.

“It was interesting when Peter Davies walked in,” Jones recalls of his colleague at the Institute for Medicine and Engineering. “I think we’d put up three sheets that day, and he immediately identified each component as a data point.”

“The thing that struck me,” Davies says, “is you had rows and layers, and each of those tags of the tapestry was an interpreted unit of information coming from the cells … And while we were there, one of the art students projected light through it, so that it threw an image onto the wall behind the curtain—and that’s another derivative, because you are turning a three-dimensional model into a two-dimensional image!” The passage of time has not blunted Davies’ intellectual high. “And I was getting ahead of myself, saying, so now you can gain further information by manipulating that three-dimensional model … It starts stimulating you to think about different stratagems for taking the derivative of huge amounts of biological information.”

Or as Jones puts it with a chuckle, “The data become something very different when you’re walking amidst it.”

An Architect Walks Into the Lab By Trey Popp

  page 1 | 2 | 3 | 4 | 5 | 6    
  ©2009 The Pennsylvania Gazette
Last modified 12/29/08