Time is on Skylar Tibbits’s side. And it’s not just because at an age when many students are finishing M.Arch. programs or experimenting with career paths, Tibbits, 28, is directing his own research group, the Self-Assembly Lab, at the Massachusetts Institute of Technology (MIT). If you’ve heard about 4D printing, it’s likely due to Tibbits, who coined and popularized it with numerous TED Talks. The fourth dimension? Time, of course.
Tibbits has built the practice he founded, SJET, around generative design, digital fabrication, and robotics. Though he interned at several prominent firms, including Asymptote Architecture and Point B Design, it was during a 2006 stint at the office of Zaha Hadid, Hon. FAIA, that he found his calling. There, he worked with project manager Marc Fornes, founder of TheVeryMany. “That was the first time I saw someone writing code to generate architecture,” Tibbits says. “I was just like, ‘Whoa, OK,’ and then I started teaching myself how to write code.”
As a student at Philadelphia University, where he earned his B.Arch., Tibbits talked his way into getting access to a laser cutter in another department. By chance, the father of one of Tibbits's classmates, Jared Laucks, who is now a fellow MIT staff member, owned a sign shop with a CNC router. This was a still several years before CNC routers became ubiquitous in architectural schools and firms, but at a time when “generative design was booming,” Tibbits says. He and Fornes parlayed their experiments with the router into Scriptedbypurpose, a 2007 exhibition of computer-scripted art in Philadelphia, which led to invitations from around the world to erect installations in venues such as the Guggenheim Museum.
In 2010, Tibbits earned dual M.S. degrees in architecture and computer science at MIT. That same year, NADAAA principal Nader Tehrani became head of the school’s architecture department. A proponent of emerging digital and fabrication technologies, Tehrani offered Tibbits a faculty position, a rare opportunity for a newly minted graduate.
Tibbits belongs to the new class of designers whose projects span multiple disciplines and include more lab work than site work. His frustration with the laborious construction process is what led him to pursue the fourth dimension of printing: programming objects with the ability to self-transform over time. “Architects should be … using their skill sets and the ability to think radically, push the boundaries of what’s possible, and articulate those ideas, digitally, physically, visually,” he says. “Architecture is changing radically—I hope—and it should change.”
Designers and students should seize the opportunity to collaborate in areas beyond the conventional realm of architecture, Tibbits says. “It’s a really vital time because in almost every field—synthetic biology, chemistry, physics, material science—there’s a boom around design tools and software.”
In a chance meeting at a coffee shop near MIT, Tibbits asked a representative from 3D-printer company Stratasys why the structures he printed couldn’t fold on their own. And the idea for 4D printing was born. Strands are a common theme in Tibbits’s work. 1D printing results in linear strands, a shape commonly found in building components such as cables and beams. Transitioning to 2D printing is straightforward, he says: “It’s either grids of 1D elements, or it’s filled-in surfaces” such as a sheet. 3D printing begins to enter the realm of self-assembly, where components with fixed properties and shapes can come together to form new objects if given an energy input.
Objects that are 4D printed have the capability to transform themselves over time. These smart materials are embedded with programmed behavior and combined in a multi-material structure. For instance, the multi-material strand on the right comprises a rigid material that provides shape, and a polymer that expands 150 percent in water, providing the transformation capability. In 15 or 20 minutes, the linear strand placed in hot water self-folds into the acronym “MIT.” Tibbits, who is collaborating with Stratasys and Autodesk on the project, envisions that such materials can tackle inefficiencies in construction and design. He is currently working with Geosyntec Consultants to design flexible pipes that can expand, contract, or undulate to vary flow rate as needed.
3D printing is becoming cheaper and more precise, but to become a serious contender in manufacturing, it must be able to print large, human-scale objects. Tibbits sees three potential solutions. The first solution is to make bigger printers, which he dismisses. “If you want to build a skyscraper, you don’t want to have to build a skyscraper-size machine.” The second is to employ multiple printers. But “not everyone has access to lots of printers,” he says, and assembly remains an issue.
The third solution is to displace density, or strategically compress a large-scale object to fit into the small volume of current print beds. In Hyperform, Tibbits and his collaborators Marcelo Coelho and Formlabs investigate ways to maximize what can be printed through digital folding. First, a given object is broken down into a linear strand of individual units connected by universal joints notched at preset angles, which will later guide assembly. The chain is then folded into a super-dense 3D Hilbert curve to fit within the bed size and printed. Then, it is unraveled and assembled. Through this technique, Tibbits has printed a 50-foot-long strand and a chandelier fixture with a final volume eight times that of the 5-inch-by-5-inch-by-6-inch print bed. Admittedly, the assembly process could borrow some lessons from the 4D-printing realm, Tibbits says. “Although there is a lot of clicking, it has the coding [embedded in the printed joints] so there’s no decision-making in the assembly process,” he says. “It’s like if you couldn’t fail Ikea.”
Self-Assembly Line (images above) builds on an earlier project in which a 3D-printed model of a virus capsid could be shaken apart or together in a laboratory flask (bottom left), and enlarges the scale to that of furniture. “The vision here is … to program things to come together in nonrandom ways, even with random energy,” Tibbits says. (His collaborators include Seed Media Group, TED Conferences, Autodesk, and Arthur Olson, director of the Molecular Graphics Laboratory at the Scripps Research Institute.) The disorderly parts would come together “to build an ordered, precise structure using noisy, cheap energy.”
Self-Assembly Line combines a discrete set of module components with a universal geometry and is outfitted with attraction mechanisms, or magnets, inside a larger container. With the addition of an energy input via a stochastic, or random, rotation—in this case, rolling the container along the plaza—the units contact each other and self-align into predetermined configurations. By varying external conditions and forces, unit geometries and quantities, and the attraction forces—here, the polarization of the magnets—Tibbits sees many potential applications for programmable self-assembly in architecture, such as the construction of multistory structures activated by wave undercurrent energy. It can also apply at the other end of the life cycle, where products can self-disassemble for recycling.