An MIT spinout has already supplied recycled-composite trusses to the U.S. Army Corps of Engineers for a 40-foot bridge in a Massachusetts wetland, and it says the same waterless process can scale to thousands of robotic production lines worldwide.

Atlas Building Composites, which emerged from MIT HAUS — Home Architecture for Universal Sustainability — in the university's Department of Mechanical Engineering, converts discarded single-use plastics into structural components: decking, foundation elements, and truss pieces for walls, floors, and roofs. The company runs an AI-powered robotic manufacturing platform, according to MIT News.

The claim that matters for the recycling sector is narrower than the branding suggests. Atlas says it skips the wash phase that makes conventional plastic recycling water-intensive, and it accepts lower-grade, contaminated and mixed plastics that many material recovery facilities turn away — material that typically flows to landfill or risks leaking into waterways.

What does the process actually change?

Conventional mechanical recycling of rigid plastics leans heavily on wash lines, which impose water demand, effluent handling and permitting costs. Atlas's waterless route removes that dependency. Founder A.J. Perez argues the implications sit at the siting level, not just the process level.

"This is key to democratizing recycling," Perez, chair, co-founder and MIT research scientist, told MIT News. "Now, every country around the world, regardless of their water access, will be able to do something about their plastic."

The positioning is deliberate: rather than downcycling sorted plastic into low-value pellets, the company manufactures durable structural products with an end market in construction. Perez frames the two waste streams as inseparable: "You can't divorce these things from each other," he said, referring to construction's dependence on timber, mining and refining, and the parallel overflow of plastic waste.

The scale of the ambition is stated bluntly. Perez told MIT News the company wants to "build 1 billion homes." His phrasing for the operation: "We're turning bottles into buildings."

Built capacity versus announced vision

What exists today is limited but real. Atlas materials are in service in sheds, barns, decks and docks, per the company and MIT News. The Army Corps bridge project in a Massachusetts wetland — built with American-made recycled composite trusses — is the flagship verified installation cited by MIT.

What remains announced rather than built is the end-state: Perez and co-founder Matt Pouliot envision thousands of Atlas robotic production systems deployed globally, allowing regions to process their own plastic waste into usable products on-site instead of exporting it or landfilling it. Neither a plant count, tonnage capacity nor deployment timeline appears in the company's public statements to date.

That gap is the story for anyone tracking this segment. The company has disclosed no throughput figures, no named offtake agreements beyond the Corps project, and no commercial-scale facility. The 5% U.S. plastic recycling rate in 2021, reported by Greenpeace, defines the feedstock opportunity — and the volume of contaminated, mixed material available at low cost to any processor that can tolerate it.

Why contaminated bales are the opening

Atlas's feedstock strategy targets the weakest link in the current plastics chain. Mixed and contaminated plastics routinely fail sortation specs at MRFs, driving disposal costs and export bans rather than reprocessing revenue. A process that structurally embeds that material into foundation elements or decking creates a demand pull that pellet markets, depressed by low virgin resin prices, have not provided.

The waterless design also changes the siting calculus. Where water is scarce, or where discharge permits and infrastructure are expensive and slow to secure, a wash-free line removes a regulatory chokepoint. Perez's argument is that this opens regions — and countries — that conventional recycling economics have never reached.

The near-term test is whether the Army Corps bridge and the barn-and-dock installations convert into repeat structural orders with published volumes. Until Atlas discloses plant capacity, tonnage intake or a deployment schedule for the envisioned thousands of robotic lines, the company sits in the announced-project column. The first commercial facility announcement with hard throughput numbers is the milestone that will move it into built capacity — and give the sector a real benchmark for what unwashed, mixed plastic is worth per tonne as a structural feedstock.