Consolidation year: what Sunbelt's 18 plants, MODUS–ROC, Boxabl's listing and Harbinger's asset sale say about who survives.

In a single week in September 2026, the Modular Building Institute's news feed carried two consolidation announcements. On September 10, MODUS Structures of Crossfield, Alberta acquired ROC Modular of Bow Island, creating a group of more than 460 employees across three Alberta plants that describes itself as one of Canada's largest modular builders.² Four days later Sunbelt Modular, backed by the private-equity firm Littlejohn & Co., announced it had absorbed Titan Modular Systems' four Georgia plants along with Apollo Modular Systems, Spartan Cargo Trailers and a lumber-supply business, taking the platform to eighteen plants — fourteen months after it bought Whitley Manufacturing.¹

Two months earlier, Boxabl, the most publicised modular company in North America, had begun trading on Nasdaq as BXBL after a SPAC merger with FG Merger II.³ And in the spring, Harbinger Production — the renamed Factory_OS, the Vallejo plant that Google, Autodesk, Citi and JPMorgan had backed, that built some four thousand units of mostly affordable housing — had issued a WARN notice for roughly 290 employees and then, on April 30, assigned its assets to a creditors' representative for an accelerated sale with bids due June 12.^5,6

The trade press covered each as its own story. They are one story.

Four deals

Three tiers

Read together, the year's deals describe an industry sorting into tiers defined not by factory quality but by capital structure and backlog.

Platforms. Sunbelt is the clearest example: a private-equity owner buying regional relocatable and commercial modular builders — segments with repeat institutional customers (school districts, health systems, energy companies), short sales cycles and lease income — and assembling them into a national footprint. WillScot and McGrath RentCorp did this a decade earlier in the leasing segment; ATCO's acquisition of NRB Modular Solutions, Canada's largest modular housing builder, put a listed parent behind a permanent-modular business.⁷ Volumetric Building Companies' strategy of buying distressed plants — Katerra's Tracy factory, then Berwick — and keeping each unit independently profitable belongs here too.⁸ These businesses have balance sheets that can carry a factory through a slow quarter. That, more than any technology, is the moat.

Regional specialists merging. MODUS–ROC is the template. Two Alberta builders with overlapping segments and complementary geographies combine to reach the scale that a federal buyer — Build Canada Homes issued its request for information on modern methods of construction in February, and a prequalified list is presumably coming¹⁰ — will want to see before awarding programme-scale contracts. Expect more of these in Canada over the next eighteen months, and expect the ones that happen before the federal pipeline is defined to be better positioned than the ones that wait.

Venture-era factories running out of runway. Katerra in 2021, Veev in 2023,⁹ Harbinger in 2026. The pattern holds across very different companies: a factory's fixed costs carried on a balance sheet that depends on the next raise or the next contract, in a market where both arrive late. The mission does not save you. The customer list does not save you. Only backlog and a patient balance sheet do, and venture capital is neither.

The question for any modular business in 2026 is not "how good is your line?" It is "who carries your fixed costs through a bad year?" If the answer is "our next contract," you are in the third tier whether you know it or not.Volumetric Building editorial position

Harbinger: the third strike

Harbinger deserves its own section because it is the hardest case for the industry to explain away. Factory_OS was not a hype company. Founded in 2017 by Rick Holliday and Larry Pace, it built real buildings — around four thousand units, ninety percent of them affordable — for real institutional clients, with Google as its first customer. It was acquired by private equity in 2024, renamed, and reported in December 2025 to be ramping production. Eight weeks later it told Vallejo officials it might close, citing a gap in its project pipeline and a loss of capital funding, and gave itself a seventy-five-day window to land contracts.⁵ It did not. By the end of April its assets were with an assignee for creditors; an industrial-asset firm marketed a "complete facility closure."⁶

Management cited cheaper competition from out-of-state and overseas factories. That is the location problem in its purest form: a Bay Area plant paying Bay Area wages and Bay Area rent, competing for California affordable-housing work against Idaho and beyond, where the wage bill is lower and the transport cost, it turns out, is not prohibitive. It is also the argument for the prevailing-wage extensions Oregon and New York attempted, seen from the other side — and a warning to the California legislature, which has spent years and a $10 million pilot trying to grow in-state factory capacity, that the economics are working against it.

We could not confirm, as of this writing, whether the Vallejo plant found a buyer. If it did, the buyer will be in the first tier. If it did not, one of the most-cited case studies in North American modular is now an empty building, and every "factory-built housing will save California" deck should be updated to say so.

Boxabl: public, and still unproven

Boxabl's listing is the inverse case: a company that has never lacked for attention converting it into a public currency before proving it can manufacture at volume. The first-half 2026 filing shows a net loss of $13.6 million against a small profit a year earlier, driven by a $9.7 million fair-value loss on a forward purchase agreement and $4.8 million in merger costs; revenue recognition under the new structure began only after the July 17 close, so meaningful sales figures are not yet public.³ An analyst critical of the company's crowdfunding history put 2024 revenue at $3.4 million against a $50.9 million loss and cumulative shipments near six hundred units.⁴ Treat those figures with the source in mind; they come from filings, but the framing is adversarial.

The public listing does one useful thing for the rest of the industry: it forces quarterly disclosure of output, revenue and cost from at least one volumetric factory. When Boxabl reports its first full quarter, we will finally have an audited number for units shipped per period from a named plant — the figure our Factory Ledger lists as "none published" for nearly every other entry. Whatever one thinks of the company, that is a contribution.

Alberta consolidates

The MODUS–ROC combination is the most strategically interesting deal of the year and received the least analysis. ROC delivered North America's first cross-laminated-timber modular project in 2021 and specialises in social, affordable and Indigenous housing — precisely the typologies a federal housing agency buys. MODUS brings twenty years of off-site delivery and the larger balance sheet. The combined group's 260,000 square feet across three plants is modest by US platform standards but large for a Canadian permanent-modular builder, and Alberta now has, between this group and ATCO–NRB, two of the country's most credible bidders for whatever Build Canada Homes eventually procures.^2,7

The question for Ottawa is whether it will structure its demand so that these consolidated builders can invest against it — multi-year, standardised, with published volumes — or whether it will run project-by-project procurement that resets the learning curve each time, as Toronto's Modular Housing Initiative did. The consolidation is the industry preparing for the first scenario. The policy has not yet chosen.

What to do with this

If you run a factory, know your tier and act accordingly: a third-tier business should be seeking a first-tier owner while it still has a backlog to sell, not after. If you invest, the platforms are pricing in the roll-up thesis already; the value is in regional specialists in jurisdictions where a public pipeline is about to be defined, which today means Canada. If you buy for the public, understand that every consolidation announcement is the industry betting on your volume, and that the bet fails — and the capacity disappears, as it did in Vallejo — if the volume does not arrive. And if you read the trade press, count the plants, then ask who carries them through a bad year.