Future Now
The IFTF Blog
Future of manufacturing
I recently finished a revised version of the magazine article on the future of personal fabrication technologies. It's now off to the editors, but I thought I'd reproduce the latest version here, with an addendum.
Raising the Floor
For many people, the word "factory" conjures up images of William Blake's "dark Satanic mills" or Charlie Chaplin's Modern Times. They imagine landscapes of machinery, consuming men and raw materials, blackening skies and destroying lives. Whatever they produce, factories are inhuman and unnatural. Certainly such factories still exist; but companies that aren't trying to win the race to the bottom are taking different paths. The outsourcing movement, and more recently the attention given to product design, have eclipsed a quiet transformation of the factory from a vast machine into a more knowledge-intensive, even creative, space. In surprising ways, the factory is now following a path blazed by the design studio and modern office: it's becoming a more knowledge-intensive and flexible, even as it grows more tightly connected to markets and suppliers.
Almost since the Industrial Revolution began in the 1750s, engineers and managers have sought to make factories more efficient and productive. Industrial engineering and operations research developed in the mid-twentieth century to put factory design on a more scientific foundation. Total Quality Management and Six Sigma brought a new focus to these efforts: they made quality improvements the centerpiece of factory reform, and made quality a key consumer benefit. They also generated vast quantities of information about factory operations, and required large amounts of information to succeed. Likewise, robotics and supply chain management made manufacturing more information-intensive.
Industrial engineers are now looking beyond the production line: Georgia Tech dean William Rouse argues that industrial engineers will design supply chains and entire enterprises, not just factories. Meanwhile, new technologies are moving into the factory floor. Put most simply, they'll make products more intelligent; make manufacturing more information-intensive; and turn the factory floor into a center for a new kind of knowledge work.
Products will become more intelligent thanks to the emergence of pervasive computing. Ever-smaller and more-powerful processors, sensors, and memory are increasing the power of handheld devices like cell phones; soon, flexible and printable electronics and displays will let us put electronics on clothes and packaging. At the same time, the growth of wireless networks and IPv6 (a new Internet protocol) will give devices greater opportunities to communicate with users and each other, and to cooperate in ways we can only dimly imagine today. These capabilities will also give manufacturers the chance to learn more about how their products are used. In some cases, networked products might report this information back to manufacturers throughout their lives; in others, products will keep digital diaries that companies can recover in eco-friendly takeback programs. (At least one printer company is quietly gathering data from recycled printer cartridges, and breaking down used printers to look for consistent failure points, causes of breakage, and over-engineered areas.)
Manufacturing, meanwhile, will become more information-intensive thanks to rapid prototyping, which allows engineers to make precise working prototypes from CAD files. Two methods for rapid prototyping (or, alternately, freeform manufacturing or layered manufacturing) have become especially important in the last decade. Both are additive processes, which build up objects one layer at a time, like rows of bricks in a wall; neither requires any tooling, which virtually eliminates the set-up times and costs of conventional manufacturing processes. In inkjet manufacturing, an inkjet printer sprays fine beads of plastic or resin instead of ink, eventually building a free-standing structure. In laser sintering, a laser draws the shape of an object in a layer of powder. The laser fuses the powder into a solid; the object is then covered with another layer of powder, and the process is repeated.
Rapid prototyping has already had a significant impact on product design. It gives designers to work faster, and catch problems in products before they reach production. It also allows users to participate in the design process, something that appeals to industries with demanding customers and a taste for ethnography. Snowboard manufacturer Burton and white-water sports company Watermark give working prototypes to fans, incorporate user feedback into the CAD files, then generate new prototypes in a cycle lasting days rather than months. (More broadly, companies as different as Xerox, Proctor & Gamble, and Hermann Miller bring users closer to product and business development by sending anthropologists to watch users interact with their products. Danish toymaker Lego has taken a different approach: Lego users can share designs on the company's Web site, and purchase kits based on popular designs.)
Rapid prototyping is now morphing into rapid manufacturing. Hearing aid manufacturers Siemens and Phonak are laser sintering silicone earbuds encasing super-small hearing aids, and makers of artificial limbs and orthodontics are following suit. Aerospace companies are bringing rapid prototyping to the factory floor to make small runs of highly complex aircraft parts. Boeing even spun out an On Demand Manufacturing subsidiary in 2002. Experts predict that machines that can fabricate electronics and displays along mechanical structures will be available by the end of the decade.
Rapid manufacturing will let companies produce new goods more rapidly, and ethnography will bring them fresher and more detailed knowledge of what consumers want; but translating that knowledge into products is still a challenge. High tech consumer products in particular have a perverse genius for encompassing conflicting demands. The first Apple mouse developed in the early 1980s had to be more reliable, easier to use, and 98% cheaper than the $700 Xerox PARC prototype that inspired it. Ever since, users have demanded products that are faster, more powerful, and more functional-and smaller, cheaper, and easier to use. Add networked, energy efficient, environmentally sustainable, and recyclable, and you have a perfect storm of contradictory demands. Balancing these needs-or better yet, finding new technologies that can satisfy them all-requires something more than conventional problem-solving techniques drawing on textbook knowledge. Problem-solving techniques like TRIZ offer ways to balance conflicting needs, by analyzing problems in ways that reveal hidden continuities between them.
New fields in science and technology are also emerging as sources of solutions. Advances in materials science, nanotechnology, and computing are obvious places to look for things that will boost power, speed, or strength; but designers and scientists across a range of fields are discovering that biomimicry-reverse-engineering natural materials and processes-has a lot to offer. Nature's designs constantly balance competing demands. Nature self-assembles biodegradeable materials into objects stronger and faster than anything humans can engineer; creates composites and microstructures that exceed the capabilities of engineered materials; and does so without any pollution. Materials scientists and chemists are learning how to improve designs and products by copying natural structures and compounds. Scientists are now studying geckos to understand how they can adhere to smooth surfaces, analyzing the composites that make abalone shells hard, reverse engineering the adhesives that let mussels resist crashing waves, cross-sectioning bones to reveal their microstructures. Architects have long copied natural forms, but are now learning how to emulate natural processes: the prize-winning Eastgate Building in Harare, Zimbabwe, has a zero-energy ventilation system based on the climate controls in giant termite mounds. Industrial engineers and chemists are studying natural processes that spin marvels out of self-assembling, biodegradeable materials.
Taken together, the trends toward greater intelligence in things, in manufacturing, and in problem-solving point to a factory that is as different as Ford's assembly line as a modern office is from its 19th-century predecessor. The office of yesteryear was a giant information-processing machine, organized to produce standardized information products and services. But vast rows of desks or cubicle farms are unsuitable for modern, knowledge-intensive industries that need innovation and creativity, not better paper-pushing, to stay ahead of the competition. Call centers can have cubicles; R&D centers, in contrast, need space for collaboration, for a variety of working styles, and that gently encourage exploration and chance discoveries.
So what will the factory of the future be like? It will be aware of how users are reacting to both its latest products and still-under-NDA prototypes, feeding off streams of information coming in from prototypes, recycled units, market-watching software agents, and blogs and discussion boards. It will be able to shift production lines in a matter of days or hours, and will constantly incorporate the latest insights from the lab and the natural world. The combined effects of cascades of information and pressure for constant innovation will turn the factory floor from a space populated only by machine-tenders, into a space in which production and innovation happen simultaneously. The factory will follow a transformation similar to the recording studio. Until the 1950s, music studios were places where groups just made recordings: they were production lines. Then rock and roll musicians like Buddy Holly and the Beatles turned the studio into a place to write songs, improvise, and experiment with new sonic effects. As Brian Eno put it, the studio became an instrument, a space for creation and experimentation as well as production.
Some rapid manufacturing factories won't produce things, but living organisms. 3D printing's ability to generate very complex shapes, and to work at relatively low temperatures, makes it a promising tool for bringing biomimicry to the production line. Tissue engineers at Stanford University's Freeform Fabrication Laboratory and elsewhere are working on laser sintering biodegradeable "scaffolds" on which human cells can grow. Eventually these scaffolds may be the structures on which fully-functioning organs are manufactured.
Whatever it makes, this is an industrial order that will also need a smarter approach to people. A variety of factors in the traditional factory converged to discourage too much intelligence in workers. Assembly lines needed workers who were as reliable as machines; managers wanted workers who were interchangeable; and labor unions pushed for strict rules defining workers' responsibilities, in the name of workers' rights and job security. This strictly hierarchical, well-ordered shop won't survive. The shift from mass production or rapid manufacturing will create a demand for workers who are entrepreneurial, highly skilled, and able to collaborate with others-and a shop floor flexible enough to let that happen.
Where will these workers come from? Where will they have learned the skills necessary to do well in this world? The unexpected but most likely answer is online games. Part of the appeal of massive multiplayer games like Entropia and Second Life is that they allow players build all kinds of interesting virtual stuff, from bodies to buildings. Turns out they're fun-ride versions of computer-aided design and computer-aided architectural design systems. Thanks to these games, a generation of kids is becoming intimately familiar with design and manufacturing-skills that, thanks to 3D printing, can move straight from the living room to the factory floor. In other words, countries with the most advanced game cultures today may take the lead in rapid manufacturing tomorrow.
This leads to a radical conclusion. The industrial world of rapid prototyping won't be one that rewards cheap labor, but smart labor. Countries that compete on the basis of labor costs and nonexistent regulations may find that the game has changed. In a world in which factories print or grow their products and pollute far less, and need workers who are imaginative enough to redesign products on the fly, cheap wages and lax environmental regulations won't be attractive. They won't even be incentives. Countries with more expensive but better-educated workforces, with well-developed consumer and gaming cultures, will be much more attractive. As it has everywhere else, the triumph of mind over matter, of brains over brawn, is coming to the factory floor.
It appears that the future may be jumping the shark. Gizmodo reports on a new service that lets you create models of objects you create in Second Life, pulling the words of gaming and manufacturing ever closer together:
Those amongst you who spend all your waking time on Second Life: rejoice! Simon Spartalian and Mike Beradino of Recursive Instruments are launching a milling service for SL users on June 1, so you can have actual physical representations of your avatar, builds or favorite SL objects made out of anything from foam to wax to stainless steel, up to 9"x5"x5".
As 3pointd writes,
Part of the goal of the project is to bridge the virtual and the real “by developing a cultural authority in the virtual that till now has been reserved for the physical," Spartialian says. The service will allow residents to create physical objects that can take on personal importance or perhaps even come to have financial weight around the edges of SL’s in-world markets.
The Recursive Instruments blog has lots of geeky goodness.
More generally, it now strikes me that there's a subject here-- at the intersection of 3D gaming (or more generally, the creation and distribution of cheap 3D modeling tools) and rapid prototyping-- that deserves a TYF article.