A boundary line of manufacturing history cuts across the factory floor of Siemens Hearing Instruments in Piscataway, NJ. On one side, skilled technicians use casting techniques, precision tools, and years of experience to craft the acrylic shells of hearing aids modeled from silicone impressions of actual ear canals.

On the other side of the factory floor, two pizza-oven-sized machines create similar shells from nylon dust. Inside the machines, needles of laser light, guided by digital design files, robotically scan back and forth, cinching paper-thin layers of dust into tough strata of plastic. Four hours and several hundred laser sweeps later, a batch of 80 hearing-aid shells is completed (see "From Dust to Hearing Aids," bottom). The process saves hours of human labor and produces hearing aids that fit and sound better than traditional ones.

It works so well that Siemens, the world's largest maker of hearing aids, is completely switching to the technology at several factories. "This whole process allows us to be more accurate and eliminate human error. This is going to change the business," says William Lesiecki, director of software and e-business solutions for Siemens Hearing Instruments.

Boning Up

In some ways, direct manufacturing is a natural consequence of the relentless pressure to reduce the time it takes to move a product from concept, through design and development, to commercial reality. When computer-aided design and digitally controlled tools began infiltrating factories in the 1970s and 1980s, the stage was set for rapid prototyping, which uses printing technologies to create three-dimensional objects that serve as prototypes for, say, toys or car parts. With prototypes in hand in just hours-rather than the weeks or months hand-carving and casting once took-designers can more quickly refine products, and engineers can quickly detect and correct problems.

The first rapid-prototyping machines used lasers to bind successive layers of a liquid polymer-a process called stereolithography. Later versions used a broader range of raw materials, such as powders that would fuse together when hit by a laser beam. Another leap came in the 1990s, when the method expanded beyond lasers to include printheads that spewed binding liquids onto powders, adding speed and an even greater variety of materials (see "Players in Direct Manufacturing," bottom). At the same time, the push was on to develop these technologies to the point that they could make finished products, not just prototypes. "In the late 1980s, stereolithography had just come out, and it was very inspiring to see," says Emanuel Sachs, a mechanical engineer at MIT who developed the printhead method. "What I set out to do was to shift the focus from making prototypes to creating functional parts directly."

That goal has now been met. On a recent day at the Therics laboratory in Princeton, NJ, two employees in cleanroom suits watched as a car-sized printer made 300 two-centimeter-long chunks of substitute jaw bone. A linear array of eight printheads swept over successive layers of a powder called hydroxyapatite (the major mineral in natural bone), selectively dispensing tiny droplets of an organic binding liquid that would later be burned out during a furnace treatment. Under the relentless sequence of droplets-800 per second-the otherwise formless mass of powder began to take shape. The U.S. Food and Drug Administration approved Therics's bone substitute in late May, and while it hasn't yet been used in an implant in humans, it is already in the hands of surgeons who intend to test it soon. As a means of making replacement bone, direct manufacturing has some advantages. Say an accident victim has lost a fragment of arm bone. The piece can be digitally reconstructed using images of the same bone on the other arm. What's more, the printing technology is able to create pores just 50 micrometers wide, which allow the bone segment, once implanted, to host real cells that make real bone, strengthening and eventually supplanting the implant.