Cheaper than printing it out: buy the paperback book.

Out of Control
Chapter 11: NETWORK ECONOMICS

I bought my first computer to crunch a database of names for a mail order company I owned. But within several months of getting my first Apple II running, I hooked the machine up to a telephone and had a religious experience.

On the other side of the phone jack, an embryonic web stirred -- the young Net. In that dawn I saw that the future of computers was not numbers but connections. Far more voltage crackled out of a million interconnected Apple IIs than within the most coddled million-dollar supercomputer standing alone. Roaming the Net I got a hit of network juice, and my head buzzed.

Computers, used as calculating machines, would, just as we all expected, whip up the next efficient edition of the world. But no one expected that once used as communication machines, networked computers would overturn the improved world onto an entirely different logic -- the logic of the Net.

In the Me-Decades, the liberation of personal computers was just right. Personal computers were personal slaves. Loyal, bonded silicon brains, hired for cheap and at your command, even if you were only 13. It was plain as daylight that personal computers and their eventual high-powered offspring would reconfigure the world to our specifications: personal newspapers, video on demand, customized widgets. The focus was on you the individual. But in one of those quirks reality is famous for, the real power of the silicon chip lay not in its amazing ability to flip digits to think for us, but in its uncanny ability to use flipped switches to connect us. We shouldn't call them computers; we really should call them connectors.

By 1992 the fastest-growing segment of the computer industry was network technology. This reflects the light-speed rate at which every sector of business is electronically netting itself into a new shape. By 1993, both Time and Newsweek featured cover stories on the fast-approaching data superhighway that would connect television, telephones, and the Sixpack family. In a few years -- no dream -- you would pick up a gadget and get a "video dialtone" which would enable you to send or receive a movie, a color photograph, an entire database, an album of music, some detailed blueprints, or a set of books -- instantly -- to or from anyone, anywhere, anytime.

Networking at that scale would truly revolutionize almost every business. It would alter:

  • What we make

  • How we make it

  • How we decide what to make

  • The nature of the economy we make it in.

There is hardly a single aspect of business not overhauled, either directly or indirectly, by the introduction of networking logic. Networks -- not merely computers alone -- enable companies to manufacture new kinds of innovative products, in faster and more flexible ways, in greater response to customers' needs, and all within a rapidly shifting environment where competitors can do the same. In response to these groundswell changes, laws and financing change, too, not to mention the incredible alterations in the economy due to global 24-hour networking of financial institutions. And not to mention the feverish cultural brew that will burst as "the Street" takes hold of this web and subverts it to its own uses.

Network logic has already shaped the products that are shaping business now. Instant cash, the product which is disgorged from ATM machines, could only be born in a network. Ditto for credit cards of any stripe. Fax machines, too. But also such things as the ubiquitous color printing in our lives. The high quality and low cost of modern four-color printing is made possible by a networked printing press which coordinates the hi-speed overlap of each color as it zips through the web of rollers. Biotech pharmaceuticals require networked intelligence to manage living soups as they flow by the barrelful from one vat to the next. Even processed snack foods are here to tempt us because the dispersed machines needed to cook them can be coordinated by a network.

Ordinary manufacturing becomes better when managed by netted intelligence. Networked equipment creates not only purer steel and glass, but its adaptive nature allows more varieties to be made with the same equipment. Small differences in composition can be maintained during manufacturing, in effect creating new kinds of precise materials where once there was only one fuzzy, imprecise material.

Networking will also inform the maintenance of products. Already, in 1993, some business equipment (Pitney Bowes's fax machines, Hewlett-Packard's minicomputers, General Electric's body scanners) can be diagnosed and repaired from a distance. By plugging a phone line into a machine, operators at the factory can peek inside its guts to see if it is working properly and often fix it if not. The technique of remote diagnostics was developed by satellite makers who had no choice but to do repairs at a distance. Now the methods are being used to fix a fax machine, to dissect a hard disk, or to speed repair of an X-ray machine thousands of miles away. Sometimes new software can be uploaded into the machine to create a fix; at the very least, the repairman can learn beforehand what parts and tools he'll need if he visits and thus speed up the on-site repair. In essence, these networked devices become nodes of a larger distributed machine. In time all machines will be wired into a net so that they warn repairmen when they are flaking out, and so that they can receive updated intelligence and thus improve while on the job.

The Japanese perfected the technique of combining well-educated human beings and networked computer intelligence into one seamless companywide network to ensure uncompromised quality. Intense coordination of critical information in Japanese manufacturing corporations gave the world palm-size camcorders and durable cars. While the rest of the industrialized sector frantically installs network-driven manufacturing machinery, the Japanese have moved on to the next frontier in network logic: flexible manufacturing and mass customization. For instance the National Bicycle Industrial Company in Kokubu, Japan, builds custom bicycles on an assembly line. You can order any one of 11 million variations of its models to suit your taste, at prices only 10 percent higher than mass-produced noncustomized models.

The challenge is simply stated: Extend the company's internal network outward to include all those with whom the company interacts in the marketplace. Spin a grand web to include employees, suppliers, regulators, and customers; they all become part of your company's collective being. They are the company.

Cases in both Japan and America where corporations have started building an extended distributed company demonstrate the immense power it releases. For example, Levi Strauss, makers of jeans for the whole world, has networked a large portion of its being. Continuous data flows from it headquarters, its 39 production plants, and its thousands of retailers into a economic superorganism. As stone-washed jeans are bought at the mall in, say, Buffalo, a message announcing those sales flies that night from the mall's cash register into Levi's net. The net consolidates the transaction with transactions from 3,500 other retail stores and within hours triggers the order for more stone-washed jeans from a factory in Belgium, or more dye from Germany, or more denim cloth from the cotton mills in North Carolina.

The same signal spurs the networked factory into action. Here bundles of cloth arrive from the mills decked in bar codes. As the stacks of cloth become pants, their bar-coded identity will be followed with hand-held laser readers, from fabric to trucker to store shelf. A reply is sent back to the mall store saying the restocking pants are on their way. And they will be, in a matter of days.

So tight is this loop of customer purchase/order materials/make, that other highly networked clothiers such as Benetton boast that they don't dye their sweaters until they are on their way out the door. When customers at the local chains start ringing up turquoise jumpers, in a few days Benetton's network will begin dyeing more jumpsuits in that color. Thus, the cash registers, not fashion mavens, choose the hues of the season. In this way, hip Benetton stays abreast of the unpredictable storms of fashion.

If you link computer-assisted design tools, and computer-assisted manufacturing, then not only can colors be nimbly manipulated but entire designs as well. A new outfit is quickly drawn up, made in low volume, distributed to stores, and then rapidly modified or multiplied if successful. The whole cycle is measured in days. Up until recently, the cycle of a far more limited choice was measured in seasons and years. Kao, a detergent and toiletry manufacturer in Japan, has developed a distribution system so tightly networked that it delivers even the smallest order within 24 hours.

Why not make cars or plastics this way? In fact, you can. A truly adaptable factory must be modular. Its tools and workflow can be quickly modified and reassembled to manufacture a different version of car or a different formula plastic. One day the assembly line is grinding out station wagons or Styrofoam, the next day jeeps or Plexiglas. Technicians call it flexible manufacturing. The assembly line adapts to fit the products needed. It's a hot field of research with immense potential. If you can alter the manufacturing process on the fly without stopping the flow, you then have the means to make stuff in batches of one.

But this flexibility demands tiptoe agility from multi-ton machines that are presently bolted to the floor. To get them to dance requires substituting a lot of mass with a lot of networked intelligence. Flexibility has to sink deep into the system to make flexible manufacturing work. The machine tools must themselves be adjustable, the schedules of material delivery must turn on a dime, the labor force must coordinate as a unit, the suppliers of packaging must be fluid, the trucking lines must be adaptable, the marketing must be in sync. That's all done with networks.

Today my factory needs 21 flatbed trucks, 73 tons of acetate resin, 2,000 kilowatts, and 576 man hours. The next day I may not need any of those. So if you are the acetate or electric company, you'll need to be as nimble as I am if we are to work together. We'll coordinate as a network, sharing information and control, decentralizing functions between us. It will be hard at times to tell who is working for whom.

Federal Express used to deliver key parts for IBM computers. Now they warehouse them too. By means of networks, Federal Express locates the just-finished part recently arrived in a FedEx warehouse from some remote overseas IBM supplier. When you order an item from an IBM catalog, FedEx brings it to you via their worldwide delivery service. An IBM employee may never touch the piece. So when the Federal Express man delivers the part to your door, who sent it, IBM or Federal Express? Schneider National, the first national trucking company to have all its trucks fully networked in real time by satellite, has some major customers who deposit their orders directly into Schneider's dispatching computers and who are billed by the same method. Who is in charge? Where does the company end and the supplier start?

Customers are being roped into the distributed company just as fast. Ubiquitous 800-numbers just about ring on the factory floor, as the feedback of users shape how and what the assembly line makes.

continue...