The East River waterfront of Queens, New York, once was a busy manufacturing hub. Pepsi had a bottling plant there, Swingline produced staplers, and Eagle Electric made circuits and switches.

Much of this industry has moved on to Mexico and China, leaving behind nostalgic signage and some prime real estate. But at one 25,000-square-foot warehouse, a startup called Shapeways is aiming to resurrect manufacturing in New York City.

Shapeways’ “factory of the future” has no assembly line or plastic injection mold tuned for cheap mass production, and it doesn’t produce any specific product or part. Instead, the company fabricates customer-ordered designs with top-of-the-line industrial 3-D printing machines. Turnaround times range from a few days to a few weeks. Through its website, the company also helps customers share 3-D design files and directly sell wares in an Etsy-like marketplace while they outsource production, fulfillment, and shipping to Shapeways.

3-D printing technology, also called additive manufacturing, is decades old, but what Shapeways and a small number of similar startups are trying to accomplish is a first: making 3-D printing production profitable at a high volume while expanding its use. “We’re trying to design an automated process for infinite variation,” says Ben Wilkinson-Raemer, an industrial engineer at the company.

Today, everyone from President Obama to DIY garage tinkerers has hailed 3-D printing as a potentially revolutionary new way in which goods might be made. General Electric and Boeing are looking to print engine and aircraft parts, while startups like MakerBot and office chains like Staples aim to put cheap printers in homes and offices. New York City Mayor Michael Bloomberg attended Shapeways’ ribbon-cutting last October.

But while the value of 3-D printing for one-off prototyping and producing high-value specialized parts is already clear, the technology’s broader potential is harder to predict (see “The Difference Between Makers and Manufacturers”). “Compared to over $2 trillion of conventional manufacturing in the U.S., [3-D printing] is still 1,000 times smaller,” says Cornell University professor Hod Lipson, who recently wrote a book chronicling the technology’s evolution. “It will take quite a long time before toothbrushes are made this way.”

Shapeways began construction on the Queens factory, its second such facility, in October and raised $30 million in April to keep growing. The factory now houses 10 refrigerator-sized printers, in which, layer by layer, a laser sinters vats of white powder into a jumble of nylon objects, ranging from functional jewelry and household goods to strange art and unidentifiable parts. When the factory is fully up and running, it will contain as many as 50 machines. Some will likely print in ceramics or stainless steel, as Shapeways does in a European factory, or in Shapeways’ most recent experimental material, an Elasto Plastic being tested in its new materials-development program. The Queens factory also has a post-processing room for dyeing the white nylon different colors.

Regardless of all of these manufacturing activities, Shapeways defines itself first and foremost as a software company, Wilkinson-Raemer says. It has started making apps to simplify the 3-D design process, which remains none too simple (see “Wanted: A Print Button for 3-D Objects”). Shapeways also sees software as vital to speeding up production and improving margins.

Because materials are the most expensive part of Shapeways’ operation, and because each batch can take a whole day to print and another day to cool, Shapeways has developed software to help triage orders and fit as many objects as possible into each tray. In a room off the main floor, engineers use the software to solve the 3-D puzzle of cramming hundreds of random objects into a production run. Wilkinson-Raemer says the company hopes eventually to use visual recognition and sorting tools to automate the matching of objects in each batch to their order, a process that is currently done by hand. (Such a program could also help Shapeways catch orders that violate its terms of service, such as weapons or gun look-alikes.)

But what will really help Shapeways do 3-D printing efficiently and quickly is a steady stream of orders, and that means Shapeways must keep growing its customer base. A few entrepreneurs formed the company GothamSmith to sell unusual metal cufflinks and jewelry, and threw in their lot with Shapeways. Says GothamSmith cofounder Daniel Stillman, “That’s a real tension in their business model, between serving us brands who use them as our manufacturing partner and trying to grow their base of customers who just want to print something.”

Shapeways does all of GothamSmith’s production and some shipping, depending on whether a piece requires finishing. Mostly, then, GothamSmith is a design and marketing operation. Though it sells in low volumes, it does have wholesale retail customers, and it can do runs of hundreds, says Stillman. He notes that GothamSmith will never achieve economies of scale this way, because with 3-D printing, the fifth and 500th bicycle-chain cufflink cost the same. But he says the company can still grow this way, and it benefits from having a “lean” operation and paying only to produce on demand.

Shapeways CEO Peter Weijmarshausen, who founded the company in 2007, wants to build similar production facilities around the world and make Shapeways the Amazon of custom product design. To date, Shapeways says, it has printed a million products and it hosts 10,000 “stores.” However, even if Shapeways achieves the success Weijmarshausen envisions, it won’t provide New York with the same number of jobs that manufacturing once did: employment at the Queens facility will top out at about 50 people.