With each swipe of the 3-D printer, more of the car emerges. At first, there are only four thin lines on the bottom of the airtight, clear plastic box. But each time the printer head passes, vapor jets out of it and mixes with a fine powder to make a solid stratum. The lines grow fatter and taller until they’re tires. A few hundred passes later, Larry Sass, SM ‘94, PhD ‘00, opens the printer’s cover and pulls out a plaster model of a four-door car that’s just slightly bigger than a matchbox.
The car is a quick example of the printer’s ability to turn a computer-aided-design (CAD) drawing into a tangible scale model. The technology is not new; it’s been used for more than a decade by manufacturers of everything from automobiles to hearing aids. But Sass sees a bigger application for it–much bigger. He wants to use it to revolutionize how buildings are constructed.
Sass, an assistant professor of architecture, is a specialist in a process that has come to be known as rapid prototyping. The idea is to translate computer designs into three-dimensional, fully formed or easily assembled scale models using any one of a half-dozen machines like those that cram his lab. Alongside the 3-D printer are another machine that precisely directs streams of liquid plastic and a set of computer-directed cutters that can shape paper, metal, plastic, and wood into pieces that have only to be fit together to form a miniature skyscraper or dollhouse-sized furniture.
A CAD design, no matter how intricately detailed, is still only a conceptual model. Machines like Sass’s faithfully incarnate the ideas of designers as models that can yield aesthetic and structural insights or reveal flaws, helping the designers perfect their products. Rapid prototyping bypasses the abstractness of blueprints and allows hands-on experience of a design as a whole.
Many product manufacturers today use this method to refine CAD designs that will, in turn, direct an almost entirely automated construction process. From cars to computer chips, human hands are involved in manufacturing only during design, inspection, and assembly. Sass’s vision is to bring this type of automation to architecture. At first, he believes, rapid prototyping will be implemented only in classrooms and drafting studios as a learning and design tool. But in the future the technology could enable safe and rapid mechanized building construction.
The architectural traditionalists who still dominate the field, however, have resisted this change. With their rulers, pencils, and blueprints, they are wary of departures from the minimalist practices honed since the Middle Ages. “The people that really control architecture today have an old-school mentality,” Sass says. “They’re not familiar with a lot of planning tools that are available to them, so they tend to stick to what they’re familiar with.”
During his first three years on the faculty, Sass pieced together his lab in a department that at first seemed reluctant to lend space to such a new type of study. Through grants, he raised nearly $250,000 for machines. With each device, he received a little more square footage of space. The lab in Building 7 has now been active for a year.
Much of Sass’s time these days is spent learning how the technology can be used to teach and investigate architecture as an art. Students who take one of his special design courses are given projects such as creating innovative replacements for Boston’s tallest structure, the John Hancock Tower. The buildings they concoct often twist and turn in curvy shapes that don’t lend themselves to the Lego mentality that seems to pervade modern architecture. But rapid-prototyping technology can also teach of the past. Sass works with graduate students to make models from centuries-old sketches by some of the world’s finest designers, such as the Renaissance architect Andrea Palladio. The models turn out to be object lessons in both the brilliance and the weaknesses of the early designers.
In September, some of this work will become part of an exhibition at the MIT Museum devoted to an architectural drawing attributed to Baldassare Peruzzi, a Sienese contemporary of Michelangelo. The exhibition will feature, among other things, three-dimensional models based on the 16th-century perspective rendering. The models will help Gary Van Zante, curator of the museum’s architecture and design collection, and other historians determine if Peruzzi really was the architect and whether the drawing is of an unknown building or a play’s set design, or whether it’s simply a study piece.
Sass envisions rapid prototyping’s slow evolution from a valuable historical tool to an eventual replacement of the blueprint-based relationship between designers and contractors. At first, he believes, the technique will merely be a design aid for architects. But as robotic machinery appears on more construction sites, computer models will directly drive the building process. In a few decades, Sass predicts, homes could be entirely manufactured by robots taking instructions from a central computing source, relegating contractors to a supervisory role. When that happens, Sass says, a traditional ranch-style home, which takes at least four months to build now, will take one month.
The idea may seem far fetched, especially given the unpredictability of construction projects. But robots are already working on sites in crowded areas of Japan, where manual labor is in short supply and very expensive. Though still requiring some human guidance, these robots perform tasks such as installing windows and help make the construction process safer, cheaper, and faster. With each passing year, building becomes more automated, as better computer algorithms take more construction- site variables into account. “Crowded cities in Japan aren’t places where slow and dangerous construction can take place,” says Han Hoang, one of Sass’s graduate students. “So these types of situations are where ideas like Larry’s will come into play best. But that doesn’t mean they won’t spread.” Hoang’s own desk is covered with components of the type of robotic arms you can buy as toy-store science experiments. He’s tinkering with ideas for wheeled robots that would construct a house bit by bit, although that possibility, he believes, is a few decades away.
Closer to implementation may be a technology developed by Behrokh Khoshnevis, a professor of industrial and systems engineering at the University of Southern California. Khoshnevis calls his method of constructing houses “contour crafting.” In this process, a large robotic arm deposits layer after layer of concrete to build walls in much the same way that Sass’s 3-D printer deposits powder. Khoshnevis and Sass are exploring questions such as how the roofs of contour-crafted homes could be built. Contour crafting has generated considerable interest in Japan; Khoshnevis says that NASA is also interested in the technique as a possible way to build the first domiciles on other planets.
But not all of Sass’s efforts are aimed at such pie-in-the-sky applications. He realizes that it will be years before his ideas catch on among architectural firms, and decades before automation moves full force onto the construction site. Until then, he says, he is happy educating others and finding smaller, more practical targets for the technology. For example, he is developing a type of emergency shelter that can be assembled from plywood pieces shaped using automated techniques. He and graduate student Nicolas Rader are working out the kinks in a design that would allow the pieces to be cut out by a computer-controlled router and shipped in a couch-sized box. During assembly, the pieces should fit so snugly that no glue or nails are required–just a rubber mallet to bang them into place. Sass is working with Georgia Institute of Technology professor of architecture Chuck Eastman to plan a way to use Georgia Tech’s technologically advanced woodworking lab to produce the parts. The final product will be less than quaint–but nonetheless an important and elegant demonstration of a technique that could someday change the world’s architectural landscape.
The dark secret behind those cute AI-generated animal images
Google Brain has revealed its own image-making AI, called Imagen. But don't expect to see anything that isn't wholesome.
Inside Charm Industrial’s big bet on corn stalks for carbon removal
The startup used plant matter and bio-oil to sequester thousands of tons of carbon. The question now is how reliable, scalable, and economical this approach will prove.
The hype around DeepMind’s new AI model misses what’s actually cool about it
Some worry that the chatter about these tools is doing the whole field a disservice.
How Charm Industrial hopes to use crops to cut steel emissions
The startup believes its bio-oil, once converted into syngas, could help clean up the dirtiest industrial sector.
Get the latest updates from
MIT Technology Review
Discover special offers, top stories, upcoming events, and more.