Some Assembly Required

IF YOU WERE TO SCOUR NEW YORK CITY for the future of architecture, the last place you might look is the rump’s end of the Barclays Center. As first impressions go, 461 Dean, the mixed-income residential tower that rises precipitously at the corner of Flatbush Avenue and Dean Street, checks all the clichés of contemporary large-scale construction. Artless setbacks do little to alleviate the problem of boxiness that blight so many of the new glass high-rises on Flatbush Avenue. Alternating composite metal panels meant to “evoke” the neighborhood read like a bad joke. (In the project description, the red panels are said to signify the area’s beloved brownstones, while their blue counterparts denote those high-rises.) But the discordant façade conceals a feat worthy of your attention: 461 Dean was entirely constructed out of prefabricated steel modules—930 in total—that were assembled at a nearby factory in the Brooklyn Navy Yard. Upon its completion in 2016, it was the tallest modular building in the world and to this day, it remains the tallest in the United States by far.

In architecture circles, modular construction is viewed almost as a cyclical fad, emerging every other decade promising to completely upend building practices—only to fade a few years later. In rare instances, exponents of this fad actually achieved their aim, albeit in a very limited or arguably counterproductive manner. For reasons beyond height, 461 Dean features in this canon, despite having neither the sculptural appeal of Moshe Safdie’s ingenious Habitat 67 housing complex in Montreal nor the suburban sublime of a Levittown. Its significance instead arises from both its prolonged construction and the numerous audits that followed. From this imbroglio emerged a pair of new modular companies that promise to succeed where all previous attempts have failed—to bring about a sea change in modular construction not just in New York but across the country.

In 2012, developer Bruce Ratner announced that his company, Forest City Ratner, had “cracked the code” on high-rise modular building. Atlantic Yards, which was soon rechristened Pacific Park, was supposed to be its model development. The Dean Street project, then called B2 (short for Building 2), was conceived as the second high-rise of sixteen—several of which, including B1, have yet to be built. But the modular construction that Forest City was hoping to, but never did, deploy across Atlantic Yards took, at 461 Dean, over twice as long as anticipated and was mired in problems: As modules were stacked on top of one another, slight differences in their dimensions, a by-product of rudimentary off-site manufacturing, multiplied—the tight tolerances demanded by thirty-two stories of modules seem to have been ignored. According to a Freedom of Information Law request by watchdog journalist Norman Oder, the growing gaps in between modules led to “significant water damage,” torn gaskets, and loose façade panels. The project was also beset by internal strife. In 2014, the building’s general contractor, Skanska USA, shuttered the Brooklyn Navy Yard plant that produced the modules and furloughed its workers, citing Forest City’s unwillingness to “resolve the significant commercial and design fees.” Following a countersuit, Forest City eventually bought out Skanska’s share for an undisclosed sum and built the rest of the modules completely on its own.

By the time the building opened, mismanagement—including what investors described as a “tangled web of nepotism and self-dealing”—had pushed the developer to start restructuring. In January 2016, Forest City converted from a traditional real estate company to a shareholder-friendly real estate investment trust (or REIT) and started shedding assets, including its module-building business. It spun out the 100,000-square-foot module-producing factory it was sitting on into a separate, smaller company called FullStack Modular, led by departing Forest City senior vice president Roger Krulak. (In 2023, FullStack moved to Connecticut, where it had found a “more efficient factory with integrated logistics,” and just this January, it opened a plant in Los Angeles, citing California’s “huge housing deficits.” It also announced that it was building the largest modular student housing project in the country, for Cal Poly in San Luis Obispo. The project is expected to cost more than a billion dollars.)

Eventually, 461 Dean Street would be sold for a steep loss to an Iowan investment firm. Assembly OSM, the other modular enterprise that sprang from Forest City’s failure, was spearheaded by the development’s lead, ShoP Architects. Its involvement in Dean Street began with the Barclays Center (2012), where it had worked directly with a fabricator—using the software CATIA to deliver models—to make the building’s signature rust-covered megapanels, some of which are forty-five feet long, ten feet deep, and ten feet wide. SHoP was looking to employ this kind of technology on a new project when Forest City approached Christopher and William Sharples, twin brothers and partners at the architecture firm, about doing something with the infill south of Barclays.

“Things didn’t go the way we wanted them to,” Christopher told me over a video call. SHoP’s website continues to describe 461 Dean as a “landmark technological achievement,” but Christopher now downplays his firm’s contributions: “We developed a very comprehensive Revit model for that project, but the contractor decided not to hire us to develop the actual instruction model for that building.” After Skanska’s departure, SHoP was brought back to help finish designing the modules.

According to Christopher, bad press around Dean Street “impacted [SHoP’s] brand” and caused the brothers and their coprincipals to reexamine their design methods. Taking a hiatus, they embarked to Turin to meet with the Italian car designer Pininfarina. There, they learned about car assembly, which the Sharpleses described to me as both simpler and more technologically advanced than building construction. Pininfarina followed the architects back to New York and performed an audit of 461 Dean at SHoP’s request. Tesla, a company that Assembly OSM would later compare itself with, conducted a separate audit and Turner Construction a third.

The audits reaffirmed that incorrectly managed tolerances had played a fundamental role in accentuating misalignment and construction problems. This heap of documentation also highlighted the “critical elements” necessary for handling tolerances in the future. There was the module’s steel chassis, which the architects needed not only to be the framework for the enclosure of the module but also its structure and its alignment system so that all the modules could fit together. Then there was the need to be able to develop a system to customize façades—because, as Christopher told me, “no developer wants their building to look like their neighbor’s.” From 2021 to 2024, SHoP filed a series of patents with titles like “mounting system for building panels”; “connection node for modular structures”; and “chassis guidance system and method for modular buildings.” “There’s a lot of IP in … [these] systems,” Christopher said.

Assembly OSM’s founding in 2019 provided the context for this windfall in patents. Despite the shortcomings of 461 Dean, financing poured in, including a $38 million Series A stamp of approval from “proptech” (property technology) venture capital giant Fifth Wall in 2021. Like many modular firms before it, Assembly had ambitious plans from the get-go; unlike those other firms, it still exists. “Our modular peers have died,” Andrew Staniforth, Assembly’s CEO, told me over Zoom. Most infamous was Katerra, a modular startup that went under in 2021 after burning through more than $2 billion in investors’ funds. (Masayoshi Son, the iconoclastic head of Softbank, called it one of his biggest mistakes ever.) Another startup, Veev, went under in 2023; it managed to build just two prefabricated homes and torch $600 million in the process. Nabr, cofounded by Bjarke Ingels alongside a former WeWork executive and a former Google Sidewalk Labs project lead, raised $48 million in venture capital after it announced plans to build an apartment complex in San Jose. The company pulled out during pre-sales—despite reporting a waitlist 4,000 people long—to “pivot” and look “at markets that are growing more quickly.” After being named CEO, Nabeel Al-Kady wrote on LinkedIn that Nabr was carrying the “torch” of a “global movement of residential real estate development” to the Middle East. (Nabr did not respond to requests for an interview.)

“WHAT KIND OF ARCHITECT are you?” The question leads off SHoP’s sole monograph, published in 2012, and sounds like a test of faith: Are you a craven capitalist or are you an artist? But as the ensuing text makes clear, there is only one possible answer: An architect must be neither a businessperson nor an artist—“an insidious divide,” to quote from later in the same passage—but some unholy and necessary meld of the two.

Inasmuch as twentieth-century prosperity in the West was a product of hard-won political arrangements between labor and capital—the New Deal being the preeminent case study—many of the great buildings and plans that followed the Second World War were a result of negotiations between architects and developer-clients. (In a few, like that between Louis Kahn and Jonas Salk, the architect might even be said to have come out on top.) This complex relationship—codified in copyright law, building codes, and licensing requirements—professionalized the architect, who could make a living wage designing buildings

It is precisely due to the sanctity of this arrangement that Kenneth Frampton accused SHoP of selling out to developers in 2007. It’s also why Christopher Sharples, who—in keeping with the double-entendre of the monograph’s title, SHoP: Out of Practice—responded by almost copping to Frampton’s accusation: “It’s not about [selling out],” he said in a 2008 interview with Perspecta. “It’s about understanding what the developer’s operating model is. Really, it’s about understanding the financial models.

The impetus for this back-and-forth was that SHoP had started taking out equity in the buildings it designed, introducing a relationship to design that would be continued by Assembly. While SHoP has dressed this move up in a myriad of ways—by likening architects to butterflies, developers to sharks, and hedge funds to elephants, for example—it’s simple to understand why the firm made the decision it did: Today, the sale values of buildings so exceed production costs that the price of property is largely a function of its size and location, not its quality. In order to recoup some of this difference, the architect can pick and choose where to take out equity in projects in lieu of a fee or part of a fee. But if a design firm takes out equity in a project, its incentives are further skewed away from responsibilities of good design and toward maximizing profit. The result of this, in turn, is further downward pressures on construction budgets and even on the budget of the firm itself. Not to mention worse buildings.

According to Christoper Sharples, SHoP continues to engage in this practice. “We are still involved with working with developers beyond just the traditional design process,” he said. “It’s part of our whole attitude.” Assembly further embodies this attitude. While the company and SHoP are separate legal entities, they are largely intertwined, sharing clients, employees, and even an office address. This corporate coziness illustrates, for one, how SHoP leadership has positioned itself as a firm that incubates projects with more revenue-making potential than could be realized through the traditional practice of architecture—the sort that attract elusive venture capital money. “Our whole process is all about how we manage risk,” Christopher told me. “And the first thing is making sure that the financial model is working before you go too far down the line.” (It’s a logic that seems to have motivated what was alleged to be a union-busting campaign against SHoP employees at the end of 2021.)

But how does a company like Assembly differentiate itself from its failed peers?

I POSED THIS QUESTION to Staniforth, who worked, before Assembly, at both Forest City Ratner during the construction of 461 Dean as well as Alphabet’s canceled Sidewalk Labs project in Toronto. It was a layup for the Wharton-trained millennial: What separates Assembly, he made clear, was its focus on cities, yes (“the surface-level differentiator is that we build in urban environments”), but more importantly, “the real difference comes down to how we think about the business model.” The fantastic implosion of other modular companies, he explained, was due to their inability to adjust to the demand of real estate development, “the most cyclical business in the world.” The economic lives of Katerra and Veev, both born into the low-interest-rate environment of the 2010s, testify to this. (Rising interest rates were even cited as a cause for Nabr stalling out in San Jose.) A successful modular company, in contrast, would have lower investment in “fixed capital.” It would have to “be able to dial up and dial down the productive capacity of the factory with demand.” Yet this presents its own problem: How exactly is a company that manufactures buildings supposed to dial up and dial down its massive operations in accordance with demand? Staniforth’s answer is, essentially, to outsource. “Instead of us doing everything, we are going to focus on developing the physical systems … and then leverage third parties to actually manufacture self-assemblies.” These third parties—“partners,” as he put it—would build kitchens, bathrooms, structures, walls, floors, ceilings, and more. “There is an ecosystem of bathroom partners in the US,” he added. “So what we do is just assembly.”

The levels of this optimization are manifold. Generally speaking, in off-site building, there are two kinds of work: panelized construction, where you build component parts like walls, and volumetric modular construction, where you build larger, sometimes apartment-sized, modules. Staniforth wants to synthesize these somewhat conventional methods into something he dubs “postmodular.” “Rather than having a single optimization function for volumetric, which is typically like, ‘How far can you move the facility away from high-cost-of-labor centers while mitigating logistics costs?’” his firm, he said, “is able to add in an optimization function for every single self-assembly.” These self-assemblies include those walls, floors, ceilings, and bathrooms sourced from third parties. “Then you optimize each of those and then combine that into a volumetric module.” (The OSM in the firm’s name stands for “off-site manufacturing.”)

Staniforth’s fetish for optimization extends past the construction itself. On the podcast Bridging the Gap: Insights & Innovations in Construction, the CEO highlighted how his firm would “optimize” its labor force. Whereas a more traditional style of building might have sequestered teams—general contractors, architects, subcontractors—kept apart by various regulations and contractual structures, Staniforth told the host that he wanted to change “how the contractual structure works” by bringing all the various players “beneath the Assembly umbrella.” Assembly would effectively be in charge of the entire building process. (If you were wondering whether Assembly’s factory—where it puts all the self-assemblies together—is unionized, like most construction sites in New York City and even the old FullStack Modular factory in the Navy Yards, the answer is no.) Combining design-build contracts and uber-outsourced construction, Assembly OSM looks less like a modular construction company and more like a logistics company. As an interviewer for Freethink, a pop-technology YouTube channel, told Staniforth, “You guys are the AWS of housing,” referring to Amazon Web Services. “Exactly,” Staniforth responded.

Yet for all this talk, Assembly has only built one building: an innocuous three-story infill mixed-use project at 147 Felix Street in Fort Greene’s historic district, less than a ten-minute walk from 461 Dean. The property, Staniforth explained to me, was bought and developed by Assembly—a prototype, with two residential units and a commercial space on the ground floor; one of the residential units, a three-bedroom, was rented for $8,500 per month in March, according to StreetEasy. (Another unit will be occupied by Staniforth himself.) As our call was wrapping up, I asked him where he planned to find demand and revenue to build more, especially as the real estate industry faces high interest rates and various exogenous shocks, like the postpandemic move away from the office. Staniforth said Assembly was zeroing in on for-profit developers whose specialty was affordable housing; their access to government tax credits and subsidies meant that they have had steady capital—even now—and they are less bureaucratic than government builders.

(Staniforth and I spoke shortly before Trump’s inauguration, so we may now have entered a different “now.” He did not respond to follow-up interview requests, and it is unclear whether or not such projects would be affected by HUD cuts. Federal funding notwithstanding, New York state’s latest budget, passed this year, allocates a modest $50 million for incentivizing the building of more “modular and starter homes.” In mid-July, the still unresponsive Staniforth announced on LinkedIn that Assembly OSM was being acquired by an unnamed company. The tone of his missive was searching: “Is the VC model fundamentally broken for capital-intensive startups like ours?”)

In any event, there’s reason to believe that Assembly’s goals go beyond affordable housing. Over the course of my interview with the Sharpleses, the question of where demand for Assembly might come from, sooner rather than later, became clear. Christopher told me about a few current nonresidential projects, including modular AI data centers: “several” for the State Department, as well as one for a start-up that’s working on liquid cooling technology for said centers. There was also a project in the works to develop a day care for the navy.

Christopher also laid out his vision of where Assembly would grow its business. “It’s about creating relationships with people who operate in the market at scale,” he said, before referring to SHoP’s work with the federal government. “These issues are issues that are critical to defense.” Then he jumped into a story about how Russian soldiers who had died at the beginning of the invasion of Ukraine had relied upon a fragile supply chain: “When you talk about what happens in defense, you don’t want a long supply chain. You want to be able to deploy an expeditionary force to a place and they can produce their own water, their own energy, and food.” Buildings, he said, would be at the core of this problem—buildings that would be able to be delivered and erected instantly, “not exposing people in the open.”

“Everything we got from the military, like GPS or fly-by-wire, that all came out of DARPA [Defense Advanced Research Projects Agency],” he went on. “Now it’s all being commercialized. I think there’s a lot of opportunity with the federal government, whether it’s with State—or with DOD—so that’s how we are approaching it. We are having these conversations at many different levels with these different groups.”

Assembly’s approach isn’t atypical in corporate America. Former employees from companies like Tesla, SpaceX, and Boeing are reported to have offered Assembly “backing and technical support”—whatever that means—and the company has embraced comparisons to such manufacturing giants. “Assembly OSM makes pre-fab high rises like Boeing makes airplanes,” reads a CNBC headline from 2022, and references to Tesla and SpaceX dot various press releases and articles promoting the firm. These sorts of companies depend on various forms of public investment. Boeing receives over a third of its revenue from the federal government. SpaceX, now intertwined with NASA, was awarded nearly $4 billion in state contracts in 2024. Tesla, famously, was saved in 2010—when it was much smaller—through a $465 million loan from the Department of Energy, and nearly half of its profits come from selling government-issued carbon credits to automakers that can’t meet emission standards. (With the cancellation of federal tax credits for electric vehicles, that may be changing soon.)

But the necessary injection of funding, governmental or not, hasn’t yet arrived. Assembly’s $60 million in venture capital pales in comparison to Katerra’s or even Veev’s. It’s also dwarfed by investments in other nascent technologies, like AI. “The biggest problem we face right now is a lot of money is going into AI; not a lot of money is going into AEC [Architecture, Engineering, and Construction],” Christopher said, with William—the less quotable of the twins—nodding along. And as much as developing modular data centers might close that gap, the Sharples brothers made it clear that a significant amount of action would be required to take Assembly and the “postmodular” industry into the next stage.

“Construction is just a nasty business,” Christoper told me. “It’s noise, it’s dust, it’s street closures. It compromises businesses and residences. So, if we can make something happen sooner and keep most of that activity outside of that city … that’s a huge benefit to our living cities.” It’s also a huge benefit for Assembly (whose factory is in New Jersey). Moving that activity outside of urban centers would both figuratively and literally mean “leveraging”—that is to say, generating risk in—the supply chains. It would potentially result in relocating construction away from organized labor strongholds and toward newer nonunion factories. And, of course, it would mean moving risk out of Assembly. Risk would be further mitigated on the labor front because Assembly’s structure necessitates striking down the walls that have kept architects, engineers, and developers “operat[ing] in a siloed mentality,” which, as Christopher explained, “has everybody trying to manage their own risk.” Not the company’s.

The Sharpleses know they’re selling the shovels to a potential gold rush. “Whoever gets there first is going to really set the market.… This is about the long game.” But Assembly’s acquisition, in particular its split with venture capital, casts some doubt on whether these feelings are shared by return-hungry investors.

ALREADY THIS YEAR, wildfires in Los Angeles have presented a proving ground for prefab companies looking to provide quick and cheap housing solutions not just for poorer Angelenos but even for the wealthy. “Prefab home designers can avoid a tight LA construction market,” claimed arecent article in Bloomberg Wealth. “Fire victims could become incredibly significant business.” A partner from one of the firms profiled, Resolution: 4 Architecture, told the outlet that while Resolution “had never done a disaster rebuild … that will likely change after this fire, especially in neighborhoods like the Pacific Palisades and Malibu.”

Modular building stands to become a Band-Aid applied to a growing list of global crises. The Paris-based architecture studio Cutwork, for example, started a modular housing scheme in war-torn Ukraine called ReHome, which it hopes will be a “low-cost, rapid development housing solution to rebuild the country not only after the war but during the conflict.” The government of Lebanon, meanwhile, was purportedly developing long-term prefab villages for Palestinian refugees in the fall of 2024. (This past spring, the Lebanese newspaper L’Orient-Le Jour reported that more than fifty of the houses had been destroyed by Israeli strikes.)

The website for a Chinese company called GS Housing Group spotlights a few large-scale modular housing projects scattered around the globe. One of these domiciles is for laborers at a Chinese cobalt mine in the Democratic Republic of the Congo (DRC). Another one accommodates workers on NEOM, the Saudi megaproject that includes the Line. (GS Housing’s project there, essentially a construction camp, looks strikingly similar to the town in the background of the viral “day in the life” vlog posted by a South African influencer living in the development.)

“When you start saying words like modular, people think trailer parks and jails,” Christopher Sharples admitted. He has loftier ambitions for Assembly than building company towns or even upmarket prefab homes in fire-torn Malibu. But throughout our conversation he made it evident that he understood that building off site would come with its own set of potential ethical problems, especially given the opacity of present-day supply chains. He assured me that his team was looking into the matter and had partnered with nonprofits like Design for Freedom that are dedicated to removing forced labor from the AEC industry.

His other solution to this thorny ethical issue was new to me—and a little shocking. Sharples told me about a company that SHoP works with called Fair Supply, which describes itself as “a cloud-based ESG risk intelligence and compliance solution.” It provides users with algorithmically produced odds that suppliers are using forced labor or violating ESG regulations. Sharples also mentioned FRDM, a company producing “supply chain risk management software” that does the same. “What’s exciting is that now you can actually type in a material, let’s say gypsum, sourced from Vietnam, from a certain fabricator. What are the chances of ethical sourcing challenges? And AI will actually start giving you information like that.”

Listening to Christopher, I had my doubts—commissioning AI to determine how ethically fraught a supply chain is? Who would certify the soundness of such a system? SHoP? Or FRDM (which, after all, is getting paid by SHoP)? I thought of the recent criminal charges brought against Apple by the DRC, which accuses the company of “using deceptive practices to assure customers that the tech giant’s supply chains are clean.” It appeared that Sharples, like many chief executives and firms before him, was asking me to trust that companies would do the right thing.

THIS PAST DECEMBER, I spoke with Reinier de Graaf, a principal of the Office of Metropolitan Architecture (OMA). In 2022, de Graaf was nominated for a prestigious EUmies award for perhaps his most notable built project of the past decade, Stockholm’s modular high-rise complex Norra Tornen (“North Towers”), whose precast ribbed concrete façade gives the deliberate appearance of a béton brut building of yesteryear. (De Graaf even described it, perhaps jokingly, as a “plattenbau for the rich.”) Norra Tornen’s constituent units were built off site not merely to emulate a modernism that clearly captures de Graaf’s imagination but also for more practical reasons. Stockholm’s extreme cold meant that the concrete required for a project of such size—at 410 and 360 feet, the two towers are the city’s tallest by some measure—would have to be poured in a conditioned setting.

De Graaf and I discussed his writing on the architectural profession and its future—particularly what he says in his books Four Walls and a Roof: The Complex Nature of a Simple Profession (2017), with its chapter on Soviet modular construction, and the more recent Architect, Verb: The New Language of Building (2023), which includes a chapter titled “Architecture Without Architects: Innovation” that highlights some of Silicon Valley’s recent forays into the profession (loosely defined). This includes the work of Nabr, Sidewalk Labs, and Katerra.

“There is this kamikaze inclination in architecture,” de Graaf told me, in reference to an argument made in Four Walls and a Roof: namely, that the embrace by modernist architects of industrial progress brought about the demise of their profession—or, as de Graaf puts it there, how “the culmination of modern architecture … [was] the disappearance of the architect as the author of buildings.” But the author is neither accusatory nor sentimental. In fact, in our chat he speculated that the “kamikaze inclination” might actually help his profession “reinvent itself in a more contemporary iteration.”

To this end, de Graaf hailed the transformative possibilities of artificial intelligence. “You could delegate a lot of frivolous tasks to the machine … a lot of things related to taste,” he said, before relaying an anecdote about a “very spoiled” client of OMA’s who requested a design somewhere “between Danish furniture and a Saudi sandstorm.” Using the text-to-image AI generator Midjourney, de Graaf told me, he discovered a solution to satisfy “even the most whimsical of clients.”

One such paymaster: The developers of the Line, a key element of Saudi Vision 2030, which according to a British television documentary has resulted in the deaths of 21,000 laborers. Still, de Graaf makes an interesting argument. He finds great explanatory power in the thesis of Thomas Piketty’s Capital in the Twenty-First Century (2013)—that wealth is now more and more the result of appreciating assets, rather than wages (the enterprising architects of SHoP would surely agree). As de Graaf wrote in a 2015 article for The Architectural Review, buildings have become more “a means to generate financial returns” instead of “a means to provide shelter.” In a flamboyantly rhetorical gesture, he disabuses his peers of the style wars by which they’ve come to define their professional identities: “There may ultimately be no such thing as Modern or Postmodern architecture, but simply architecture before and after its annexation by capital.”

But to what end does this financialization lead? It isn’t the case that financialization, or the annexation of architecture by capital, was one singular event. Financialization is instead a program, with different actors—capitalists—negotiating its expansion.

In the context of de Graaf’s sweeping narrative of twentieth-century architecture, the modernists detonated the vest, so to speak, in order to “deploy the … maelstrom of industrial development for the greater good.” They saw the road to progress and stepped aside. The Sharpleses have dwelled on this history and come away with a different conclusion. Christopher told me that “the challenge” facing architecture today requires “getting back to the basics … to the point that we were master builders, where we understood every element of the [building] process.” He has invoked that image—the heroic master builder of yore—in interviews going back to 2008. Seen through this lens, Assembly’s postmodular architecture is less a culmination of the so-called kamikaze inclination than a last stand against it. But such romantic nostalgia for the master builder rings hollow. What was described to me is not a physical builder but a financial risk manager: a generator of risks in the supply chain and a mitigator of risks in business.