25 Cool Things you wish you had...and will.
SOME OF THE FOLKS AT MIT'S MEDIA LAB in Cambridge, Massachusetts, seem a bit strange. One team of geniuses adds features to Mort, a stuffed penguin who can tell time and order sushi in Japanese. Another fellow walks around wearing a multimedia computer on his face. The things people do for science. But however ridiculous they may seem, they are heroes of sorts. They represent the vanguard of the embedded-microchip revolution.
That revolution, so far, has been more or less invisible. A computer chip makes your car window go up, but who knows? Who cares? "People just want to hear about PC chips," says James Turley, senior editor of the Microprocessor Report. "Embedded chips account for almost 100% of the microchips out there, but they have 0% of the sex appeal."
That's about to change -- or so we heard when we asked 30 analysts, engineers, and futurists to show us where the embedded-chip revolution will take us in the next 5, 10, and (why not?) 20 years.
To hear some seemingly rational people tell it, embedded chips are right up there with the Second Coming. They told us about chip-powered products that will help the blind see (#17), the deaf hear (#9), and the lame walk (#25), as well as chips that will diagnose and cure disease (#3, #13, #18).
We heard all about chips that will drive your car for you (#21) and chip-controlled garments that will keep you warm in winter (#14). Chips that will improve your gardening (#22). Even chips that will let you be incredibly lazy. "Why do I have to get out of bed to answer the phone?"
groused Silicon Valley veteran Todd Mozer, who is working on voice-
recognition chips to free humanity from that task (#10).
Like many revolutions, this one is powered by economics -- in this case, the fact that chips keep getting faster, better, and cheaper. The 64-bit chips that run today's Nintendo 64 video games go for around $20;
similar processors cost more than $1,000 in 1993 when they were used to run high-end servers. No wonder the new James Bond game is much more fun than the games on the original Nintendo Entertainment System, with its 8-bit chip. "The same processor that powers the game that keeps your kids quiet on Christmas morning was running the most complex CAD/CAM programs five years ago," says Tony Massimini, chief of technology at Semico Research, a semiconductor industry research group in Phoenix.
None of this would happen if we had to wait for Intel to lower its prices. Intel's 85% to 90% market share in the PC chip market gives it such enormous pricing power that even the oldest Pentiums fetch at least $60. Fortunately, the embedded-chip business is a free-for-all. It includes roughly 40 contestants, from heavy hitters like Motorola to no-
names like SandCraft. That makes it easier for embedded-chip buyers like Ford or Sears to drive a hard bargain, which in turn is why some chips in your car or your microwave cost less than $1.
Fact is, some of today's embedded chips are already nearly as powerful as the much more expensive chips from Intel. "The power and capacity difference between embedded chips and chips that go into PCs is already small, and it's shrinking," says Turley. "Dozens of chips in the $25 price range are just as advanced as the Pentium with MMX, which came out last year."
Five years from now? Turley figures $25 will get you an embedded chip that's at least 10 times more powerful than a Pentium with MMX. Today's most advanced embedded chips will cost around $1 and will power toys. In five years, at least a few products that would seem almost miraculous today will be routine. Ten or 20 years out, we start to get the real miracles.
That said, even the smallest, fastest, cheapest chips can't accomplish those miracles alone. If they could, you'd already own some of the futuristic products profiled on the following pages. "With many potential products, the chip technology is there and it's affordable, but we're waiting for other technologies to get up to speed," says Turley.
What's missing then? Designers must find ways to reduce chips' power consumption and heat generation. For some applications, such as smart labels (#20), that's a hurdle that could take 10 years or more to clear.
Many of the coolest products will also require sensors that experience the world around them. Those chip-run devices will hear through better voice-recognition technology so that you will be able to tell your front door to unlock (#10)...see with cameras, radar, and other imaging technology to rescue earthquake survivors (#7) or avoid car collisions (#11)...and feel through smart materials such as ceramics that respond to physical stimulation -- for example, skis that stiffen or relax based on snow conditions (#1).
"The improvement of sensors that connect chips to their applications is going to drive all of this," says Paul Saffo, a director at the Institute for the Future in Menlo Park, California. "We're at the beginning of a sensor revolution."
Infrastructure is another crucial and so far largely missing ingredient. The parking space finder in your smart car (#11) will do its job only if someone has installed chips in the mall parking lot. The chips that run your wearable computer (#5) or a personal shopper-on-a-card (#8) will require ubiquitous wireless networks. Sean Baenen, a principal at Global Business Network, a future-planning consultancy, uses the example of a 300-employee firm with 100 chip-powered devices per office. "You'd need a nifty low-power, short-range transceiver system with a bandwidth potential of 7.5 gigabits per second," he points out. "The wireless 'pipes' will have to be stupendous."
Some of the more fantastic applications will require advances in different types of chip technology. Case in point: neural networks, which mimic the structure of the human brain to learn things, such as recognizing patterns or interpreting data intuitively. Work on neural networks could some day yield an implanted chip that replaces lost neurological functions in paralysis victims (#25).
It's harder to guess when the market will be ready for these miracles. Consumer economics will play a decisive role. What people want will ultimately determine whether the future reminds us of Star Trek (lots of cool handheld and wireless devices like #2); the Jetsons (smart appliances like #4); Dick Tracy (wrist-borne gadgets like #23);or the Six Million Dollar Man (bionic body parts like #9 and #17).
Our guess? All of the above -- and you probably won't have to get out of bed to answer the phone.
Smart Skis Granted, the new parabolic skis are really cool --
but smart skis will be even better. Within 10 years, an embedded microprocessor will allow you to adjust your skis' flexibility. When you feel like Austrian skiing
uuml;bermensch Hermann Maier, press your skis' "racing" button with your pole: They will instantly stiffen. As the day wears on and your legs turn to Jell-O, you can switch to the "recreational" setting for more flexible skis.
This fabulous excuse to buy new skis will emerge from an ongoing revolution in sensor technology, which allows microchips to interact with the world in increasingly complex ways. In this case the sensor is a piezo material -- typically a ceramic that gives out a charge when bent and can be manipulated with electrical impulses. A microprocessor will control a current to relax or stiffen a piezo layer in the skis. According to Institute for the Future's Paul Saffo, the chips to do the job already exist and are relatively inexpensive. And piezo materials are already cheap enough to use in cigarette lighters (they make the spark). But no one has yet figured out how to provide sufficient power to keep you skiing all day.
Meanwhile, K2 already sells skis that use electrical charges generated by a piezo layer to dampen vibrations. Result: The skis chatter less on icy slopes.
Powerful PDA Analysts such as Tom Starnes at Dataquest figure that many chip-run products due to surface during the next five years will combine the functions of today's gadgets -- and perform those functions more effectively. Personal digital assistants are a perfect example of that trend.
Think the hugely popular PalmPilot (more than 1 million units sold since it was introduced in 1996) is cool? In five years, you might pay around $400 to $500 for a vastly souped-up PDA. Everyone has his or her own idea of what this marvel will do. But results of our poll suggest that it might combine the PalmPilot's functions (Rolodex, date book, organizer) with a wireless Internet connection for Web access and email; spreadsheets and word processing software; access to your office PC; and a pager and cell phone.A processor comparable to a midlevel PC chip can handle all of these features. To be sure, the processor must consume less power than it does in a PC so it can run on batteries -- and it must generate less heat. (The joke is that if you put a Pentium II in a PalmPilot, you'd burn your hand.) Chipmakers such as Intel maintain they have the technology to address that problem.
Perhaps a bigger hurdle: The devices require a widespread wireless network, and that's three to five years down the road. "Mobility in personal computing will be driven by the emergence of a wireless network," concludes Intel's general manager Tom Franz.
PHOTOS: (TOP TO BOTTOM) GREGORY A. DALE; BOB MICHAELS PHOTOGRAPHY
In five years, you'll be able to buy a digital camcorder that connects to your TV and the Net so you can email your home movies to grandmother's house.
|
]]>
Injectable Health Monitor What's going on in your body right now? If you have diabetes, you really need to know—in particular, you need to know your glucose levels. But to get that information, you must prick your finger to get a blood sample as often as six times a day. That helps explain why some diabetics don't monitor their condition as closely as they should, even though neglecting to do so can lead to fatal complications.
A Germantown, Maryland, firm called Sensors for Medicine and Science (S4MS) is developing a tiny implanted or embedded glucose monitor for people with diabetes; it hopes to begin selling the monitor in five years. S4MS CEO M. James Barrett figures a monitor that lasts one year will cost somewhere around $1,200—about what a patient now pays annually for finger-prick test strips and monitors.
The implant includes a light-emitting device, a sensor that detects changes in the fluorescence of a molecule—which is sensitive to changes in glucose—and a chip that translates information into radio signals. Those signals then go to a computer in a special watchlike device worn on the wrist that gives you the glucose readout.
The five-year wait Barrett predicts for the product has little to do with silicon technology; he says the processor in the watch will be "standard, off-the-shelf stuff." But the firm still needs to make the implant smaller. Right now it's the size of a shirt button; in two years or so it will be small enough to fit inside a syringe. There's also considerable work to be done on the telemetry, the power (the implant will draw it from radio waves), and the chemical reaction sensor.
Eventually, expect chip-run implants to monitor things such as metabolic waste products for kidney patients or oxygen levels for high-altitude mountaineers. "The limits are related to what can be usefully measured clinically," says Barrett.
All-Knowing Appliances We don't just mean appliances that talk and obey voice commands, although these may do that. We mean ovens that know just how long to cook a casserole or bake a cake without being told; dryers that run only until the clothes are dry; washing machines that know how much hot water to use with a load of cotton sheets. We mean robotic vacuum cleaners guided by micropower-impulse radar that can clean your house while you watch Oprah.
These kinds of appliances will save lots of energy. They won't be drying dry clothes or burning casseroles, and you can set them to operate when electric rates are lowest. And get this: When a smart appliance is broken, it will call the repair shop on the Internet and describe the problem.
Meanwhile, your machines can keep up with the Joneses' machines by downloading software upgrades (via the Internet, of course). That will be handy when new features come out, such as software that lets your washing machine communicate with your smart clothes (see #14). You'll pay a monthly fee for some bells and whistles, just as you subscribe to some phone services now.None of this requires much processing power, perhaps a chip with 50 to 100 mips (millions of instructions per second), which costs less than $40 today.
The real job is convincing GE that consumers will be more likely to buy their appliances if they have more fancy features attached. Once that happens, you'll be able to buy a smart oven or washing machine for not much more than the price of a dumb one.
Wearable Computer With chips turning up everywhere, computers will be everywhere as well. That's the idea behind Ubiquitous Computing, a.k.a. Ubicomp. The Ubicomp paradigm includes the wearable computer—which encompasses CPU, hard drive, and power supply; input devices such as cameras, microphones, and other sensors; and special keyboards or handwriting tablets. The output is a tiny monitor mounted on a pair of glasses. Folks at MIT and elsewhere have models that run on Intel 486 or Pentium 133 chips; the next generation will use a Pentium 166 processor.
Unlike, say, a laptop, a wearable computer will operate all the time, even when you're walking around. It will let you know (without being asked) when you've got email. It will allow you to carry an entire reference library with you and provide easy access to the information there. You'll be able to carry your notes with you and pull them up for quick perusal. It will integrate CD player, fax machine, pager, and audio journal. It will sit and recognize the acquaintance you bump into at the grocery store and tell you his or her name. It will tell you how to fix the sink while you're fixing the sink.
Who will wear these things? For now, several researchers at MIT spend up to 16 hours a day wearing their prototypes. Eventually, perhaps within a decade, they will weigh a lot less and become affordable.
When that happens, wearable computers might become standard issue for stockbrokers, doctors, real estate agents, traveling salespeople, emergency service workers, lawyers—anyone who needs information on demand but can't be tied to a desk or even a laptop.
Meanwhile, for about the same cost as a medium- to high-end laptop, you can build one like the MIT guys wear. Full specs are on the Media Lab Web page at media.mit.edu/projects/wearables.
Surfing Camera If you didn't get a digital camcorder last Christmas, you might want to hold off for a couple of years. Already they make editing a snap; they even let you insert graphics and other material to make custom productions. But within five years, you'll be able to buy a digital camcorder that connects to your TV and the Internet—perhaps through a high-speed cable modem—so you can email your home movies to grandmother's house.
Today's digital videocams, which cost around $2,500, run on chips (including the latest 32-bit numbers) that together cost some $227. But during the next several years, we'll see chips at or below this price that can run 400 gops (giga operations per second), which in turn will allow higher-resolution digital images and networking.
In a future earthquake, rescuers will use small devices with micropower impulse radar to pick up people's heartbeats under tons of debris.
|
]]>
Earthquake Survivor Detector This should appeal to our California readers. Invariably, television reports of earthquakes and other disasters show rescuers searching through mountains of rubble trying to find survivors. Steve Azevedo, a researcher at Lawrence Livermore National Laboratories in Livermore, California, figures that in three to five years those rescuers will be armed with devices, $1,000 or less, capable of quickly guiding them to survivors. These detectors will use micropower impulse radar to pick up people's heartbeats or lung contractions under or through tons of debris.
Azevedo figures his team will be able to place the necessary radar on a single $1 to $3 chip in a year or two. Meanwhile, someone will have to come up with the rest of the gizmo's parts—including an antenna specifically designed to locate heartbeats. It could take a few years for all of this to come together; let's hope the San Andreas fault can wait that long.
Personal Shopper-on-a-Card If you make your living as a personal shopper, think about going back to trade school. Mark Weiser at Xerox PARC predicts that within three or four years your clients will replace you with a $35 shopping aid, essentially a credit-card-size wireless display. The shopping aid will track your shopping habits and record your retail wish list, then alert you when an item or service that's likely to appeal to you goes on sale. Say you want a red leather jacket, size extra large, but don't want to pay more than $300. Your shopping aid will tell you when a jacket goes on sale at or below that price.
Much of the infrastructure is in place. Most big retailers already have their inventories on their own company networks, and wireless networks are growing rapidly. But Weiser figures the devices will need energy-efficient chips that can deliver 1,000 mips for every watt of energy they use.
Bionic Ear Too much loud music in your youth? The bionic ear may solve your problem. Traditional hearing aids act like amps to make things louder for people whose ability to detect sound waves has merely faded. But Advanced Bionics, a firm in Sylmar, California, sells devices called cochlear implants, which restore hearing in people who are profoundly deaf. The implants, which cost around $18,000 to $20,000 today, require an embedded chip as well as an array of electrodes to stimulate hearing nerve fibers, a microphone to pick up sound, and a rechargeable battery to power the whole scheme.
Pretty cool—but the computer part is so big you have to wear it on your belt. To be truly bionic, we figure the whole thing should be small enough to implant in your ear or even in your skull. Right now, the cochlear implants rely on Texas Instruments's 320 line of chips; at $30 to $50 apiece, they process around 80 to 100 mips. The implants use only 3 to 10 mips—that keeps power requirements low—but limits the implants to speech processing.
Advanced Bionics's human factor engineer Michael Faltys figures that in five years we'll see inexpensive chips that can run more mips off lower power (chips that run off body heat will have to wait much longer—maybe forever). That will reduce software costs—right now, the biggest expense. Meanwhile, smaller microphones are on the drawing board. Five-year target: an implantable device that can accurately process speech.
Intelligent House Sensors will identify you and unlock the front door as you walk up the porch steps. The living room will turn on your online home entertainment center when you enter; you can then ask it to play your favorite music or film. You'll program your home to maintain a certain temperature in the dining room at suppertime and another at bedtime.
You'll be able to flip on lights and answer the phone by issuing a simple voice command. After your alarm clock wakes you, it will ask if you want coffee—then tell the kitchen to brew it. Anchored displays throughout your home will provide information about everything from news headlines to which room the kids have invaded this morning.
Tom Franz of Intel figures that today's home PCs could easily run one of these setups. So why don't we all have smart homes already? Many functions will require an expensive array of sensors, cameras, and microphones. And while voice recognition already works with today's $29 chips (about 200 mips), researchers are still working on software that can identify your voice in a noisy environment. Once that's licked, the big obstacle to the smart house may be irate consumers: Most existing houses will need a $5,000 rewiring job.
Thinking Car So far, chips in cars control stuff that is either invisible (antilock brakes, fuel-injection systems, transmissions) or boring (power windows). But that will change when chips in cars start interacting with one another, with drivers, and with the rest of the world.Sensor chips known as MIRs (micropower impulse radar chips) will hide in your car bumpers and warn you when you're too close to other cars or objects. The chips already exist and cost less than $10; with any luck, you might pay as little as $25 for radar on your next car. Other sensor chips (5 to 10 mips, about a buck apiece) in seats will monitor passengers' weights and positions; that way, your smart air bag (powered by the same kind of chip) won't deploy if there's a child in the seat. The system is already in BMWs in Europe; in 5 to 10 years it will be cheap enough for most new cars.
Motorola, Texas Instruments, Hitachi, Samsung, and NEC sell 20- to 75-mips chips for $30 or $60 to run onboard global positioning systems in Cadillacs and Mercedes; the option sells for around $1,000 and tells you the best way to get from point A to point B on a map.
In five years, those systems will be a lot cheaper; meanwhile, we'll begin to see more advanced navigation aids powered by similar chips. They'll use better wireless networks to tell you, say, where the nearest Mobil station is and how to get there.
E-Book Book design hasn't changed much since Johannes Gutenberg invented movable type in 1450. Enter Joe Jacobson, assistant professor of media arts and sciences at MIT.
His plan: Embed a central processing unit in a book's spine. Make smart paper by coating paper with electronic "ink"; the ink includes embedded electrodes, and the ink changes color when you apply an electric field. Power up with a couple of AA batteries. Then plug the thing into a network. Result? Within two or three years, you'll be able to access thousands of volumes with a single book.
Say you've just finished reading Crime and Punishment and you want to read another book by the same author. Your e-book lets you go online to view Fyodor Dostoyevsky's other titles at a library site. Erase Crime and Punishment, download The Idiot, and read on.
Jacobson expects e-books to hit the market within the next two years, with an initial price of around $400; Sean Baenen of Global Business Network figures the manufacturers will give them away and then charge subscription fees for services.
Either way, a device with a PC interface and the ability to download one book will probably require almost no onboard processing power. An e-book that can store multiple books, pictures, and animation might need a chip with 10 to 300 mips, maybe something like the StrongARM chip that Intel just licensed from Digital; it sells for around $40 today.
If you suffer from diabetes or kidney disease, the ultimate smart toilet will analyze your urine and then forward the information to your doctor's office.
|
]]>
Dr. Toilet Concerned about your health? A mite obsessive about your diet? The smart toilet is just the ticket.
Already, rumor has it, a prototypical version tells patrons of swanky Japanese nightclubs their blood alcohol content (assuming they can hit the bowl). The ultimate smart toilet will act as a lifestyle, health, and diet monitor. It will look like...a toilet, with maybe a modem-sized box attached. The bowl will contain a sensor pad to analyze your urine (labs use a similar apparatus today). The pad will send data to a garden-variety embedded microprocessor, and software will forward that information to a screen for you to read or to your doctor's office via the Internet or a wireless system. It will tell you if you're ingesting, say, too much salt. It will diagnose and monitor chronic conditions such as diabetes or kidney disease.
Why can't you buy one yet? Largely because no one has gotten around to designing the thing. The technology's already here.
Smart Clothes This is another one of those products of the future (see also #1, #10, photo opposite) that will require sensor technology built into a chip or some other material. The chip in your parka or jacket will take note of rising or falling temperatures (your own or the environment's); the chip will then send a message to a reactive fabric built into your coat to adjust for comfort.
For example, the fabric might be wool interwoven with fine electric heating elements that use a thin battery or even a chemical reaction to change the temperature. Presto: virtual layering! If that doesn't excite you, you must live in southern Florida.
The rest of you will want to know where and when you can buy such a marvelous garment. The technology required to make affordable smart clothes includes a very low-end embedded chip—the kind we use in thermostats today; it produces less than 1 mips and costs maybe 50 cents to $1.
Someone will also need to design the materials that become more or less insulating when you run a charge through them; that might cost a manufacturer $20 to $50 to do now and could add $40 to $100 to the price of the garment. At that price, consumers may take awhile to warm to the admittedly strange concept. Until then, hang onto your pile jacket.
Common Cold Detector Thanks to Saddam Hussein and his ilk, the Pentagon is spending heavily to develop chip-run devices to detect anthrax and other biological weapons. Within 10 or 15 years, we'll be able to buy our own scanners (handheld) or badges (pin them to your clothing) to tell us if the babysitter's lingering cold is contagious or the subway car we're about to board is full of flu germs.The devices don't require tons of electronic processing speed; 8 to 10 mips (equivalent to the power of today's $15 Motorola 68300) will be sufficient. But they will also require a microchemical processor—a so-called "lab on a chip"—sophisticated enough to do thousands of chemical analyses each day.
Those will be available in a few years and will cost hundreds or even thousands of dollars apiece; it will take 10 to 15 years to bring their price down to $100 or so. Like handheld food testers (#24), contagion detectors also will require a method for combining a microchemical processor with an electro-optic processor; that's 5 to 10 years off.Are you really going to scan the babysitter? Maybe not. But if enough people use these gadgets, colds might become less common. In any case, the things will be worth their $200 or so price tag (we're guessing here), if only because they'll provide the perfect excuse to skip work when you're even a little under the weather. "Sorry, boss, but my scanner lit up like a Christmas tree..."
Ask the future fridge what you can make for dinner, and it will offer recipes based on your current food inventory.
|
]]>
Wise Refrigerator Take a regular refrigerator, add a display screen, a bar-code scanner, an Internet connection, and a high-end PC chip such as a Pentium or comparable embedded chip. Result: the most important advance in refrigeratoronics since the automatic ice maker.
Ask your very smart fridge what you can make for dinner: It offers recipes based on your current food inventory (it knows because you scan your groceries when you bring them home and again when you take them out to eat). Heart set on lasagna tonight? Email your fridge from the office to find out what you'll need to pick up on the way home. If you're too busy to shop, the fridge will order the ingredients from an online grocery store.
You can create this genius yourself today—if you are willing to buy the components and write the software. But why bother? The major manufacturers are likely to start making them for us in five to seven years. Expect to pay maybe $300 to $500 more than you'd pay for a standard one.
Bionic Eye The Six Million Dollar Man had metal replacement eyeballs that gave him telescopic vision. The bionic eye that's coming in five years will be built with silicon, and it will merely restore sight to the blind.
The target market includes the 700,000 Americans afflicted each year by macular degeneration (the Western world's leading cause of blindness), as well as sufferers of retinitis pigmentosa (1.6 million victims worldwide). Both diseases involve the degeneration of cells that convert light to neural messages at the back of the retina. Solution: Place a chip in the back of the eye to mimic the lost cells.MIT and the Massachusetts Eye and Ear Infirmary are developing a product that will work like this: Glasses housing a video camera and a signal processing chip obtain an image and translate it into electronic code. A laser beam carries the code to a device implanted on the retina (the laser also provides the power to operate the implant). A chip in the implant decodes the picture information and sends electric pulses to appropriate cells in the retina. The pulses go to the brain, which interprets them as images.
After nine years, researchers are ready to try temporary implants on human volunteers to see if the brain can interpret signals from the device. Longer term prostheses will require a biocompatible implant that can survive the salty environment of the eye. All existing encapsulating materials eventually leak sodium ions, destroying the electronics. That, not more powerful or cheaper chips, is the major obstacle to the bionic eye: "Speed isn't an issue because the nervous system is pretty slow compared to today's chips," says Dr. Joseph Rizzo, who is developing the implant with MIT's John Wyatt. "The fundamental problem is in the anatomy, not the electronics."
Miracle Implant Beginning last August, people with Parkinson's disease could receive implants that stop their tremors by sending high-frequency electrical impulses to their brains. The implant requires an external device to provide the electric current and processing power. But Jeffrey Greiner, president of neural stimulation-device company Advanced Bionics, expects a more elegant chip-driven implant to hit the market within five years.
The chips to do the job, which require less than 1 mips each, will be a trivial part of the device's cost. The key is finding ways to power the chip with an implantable battery that will last for long periods. There's also considerable research to be done to figure out where to place the electrodes that deliver electricity to the body parts in question. With luck, that work will also lead to possible implants to treat other neural disorders such as urinary incontinence and sexual dysfunction.
Drug Screener Whatever disease you get in the future, there's probably a drug somewhere that would help. Trouble is, the drug might not be produced. "Only 10% of drugs that go to clinical trials ever make it to market," says Dale Pfost, CEO of Orchid Biocomputer in Princeton, New Jersey. "Many of the rest are scrapped because a few people react badly to them."
Handheld drug testers might salvage those drugs for people who can tolerate them. Before prescribing a drug that occasionally has bad side effects, your doctor will screen you to make sure you'll react well to it.
The testers will require electronic processors that can handle 8 to 10 mips; that would be something like a Motorola 68300 (current price: around $15). But they will also need chips that can process chemical and biological reactions to do a quick blood workup and DNA analysis; that's 5 to 10 years away for a large version, longer for a handheld one.
Label Alert! Isn't life a pain? All those things to remember: when to take your pills, how long ago you bought that ham, where the hell you put the garage door opener.
Smart labels, due in a decade or so, won't restore your vision or drive your car. As Xerox PARC's Mark Weiser notes, "These things aren't going to end world hunger, but they will take care of those nagging annoyances."
A label on your bottle of pills will take temperature and humidity into account to give you a more accurate expiration date. It also will communicate with your personal digital assistant to remind you to take your medicine. Food labels will talk to your refrigerator to tell you when it's time to throw out that ham. And your garage door opener's label will give you a call on your PDA to let you know where it is.
Then again, a label can be too smart. Retailers will use smart labels to figure out who picked up a product or stood in front of a shop window—making it easier to target and pester potential customers.
The chips to power these labels already exist, and they're pretty cheap at around 35 cents apiece, down from $10 to $20 several years ago. In another year or two, the price will come down to a few pennies at most. "Talk about these like they're free because they practically will be," says Sean Baenen of Global Business Network.
Trouble is, no one's figured out how to make batteries small enough to power the chips. Thus, smart labels must wait 5 or 10 years for researchers to figure out a way to come up with a solution. Solar energy?
Why do your bleeding hibiscus plants croak every summer? One of these yearssay, by 2008gardening chips will have the answer.
|
]]>
Automotive Autopilot A smart car (#11) tells you where the nearest gas station is or warns you when you're too close to the guy in front of you. A very smart car drives itself.
Already a prototype truck in Germany uses sensors to track lane position, traffic signs, and other vehicles, sending data to onboard computers that brake, steer, and accelerate. Meanwhile, MIT's Media Lab has built a computer that can predict your next move from shifts in your posture and then help initiate the action—a left turn, for example.
Processing power is not a major obstacle. Affordable radar, video, and other technology for chips that run the cars probably will take longer. Whatever happens, you may have to give up your steering wheel. "I can see the steering wheel playing less and less of a role in the actual steering of the car, until it becomes a sort of vestigial leftover," says editor James Turley of the Microprocessor Report. "It'll be a hard sell, but I think it will happen."
Green-Thumb Chip Why do your bleeding hibiscus plants croak every summer? One of these years—say, by 2008—gardening chips will answer questions like that on your PC. In fact, it's not difficult to imagine a time when microchips will communicate wirelessly with a server in a farmer's office. Such messages will tell farmers when the lower 40 is too wet or the alfalfa has fungus. Sean Baenen of Global Business Network says the farming chips will be ready for market in 5 to 10 years. The chips need to be cheaper, a few pennies apiece at most—but that will happen in two to three years. Better sensors also are needed: Some already can pick up humidity levels, but they can't yet detect more complex conditions such as fungus. And someone must figure out a way to power the chips (you can't use solar energy to run something that's buried). One possibility: Use energy in the radio waves that relay information from your PC.
Dick Tracy Watch "It's no secret," confides Texas Instruments Senior Fellow (and leader of the team that invented Speak & Spell) Gene Frantz. "The goal is the Dick Tracy watch."
Like the iron-jawed detective's timepiece, this watch will enable wireless voice and video communication (it will also carry television). First, however, guys like Frantz have to solve a few problems.The audiovisual equipment to create the watch is already here. So is the processing power: Videoconferencing, which currently requires about 1 bips (billion instructions per second), is already being done on a Texas Instruments TMS320C6X. The chip costs only around $100 now, and that should fall to tens of dollars after 2000.
Power is a bigger hurdle. Frantz thinks the gadget will need to run on microwatts or even picowatts of power; that compares to 0.6 watts for cell phones. Researchers also must find a way to fit radio signals onto the chip along with devices that can handle analog and digital signals.
All of that work could take 15 or 20 years. Frantz figures that early adopters of the technology will be able to buy the watch for around $1,000. But the ingredients will be cheap enough to allow competition that will quickly bring the price down to $200 or less.
Toxin Tester If you've ever eaten a bad oyster, you'll buy one of these as soon as they hit the market. The handheld food tester of the future will rely on a microchemical processor dedicated to searching for E. coli, salmonella, and other pathogens and toxins in meat, poultry, and seafood.
Microchemical processors, which borrow microfabrication and photo- lithography from the semiconductor industry, already can detect biological or chemical compounds. Unfortunately, the processors will cost several hundred to several thousand dollars each.
Moreover, researchers must figure out how to combine a microchemical processor with a traditional electronic processor to create a single chip. Orchid Biocomputer CEO Dale Pfost, whose firm is developing these processors, figures that all but the simplest testers will take 5 to 10 years—and it will be at least 15 years before the chips are cheap enough to power a food tester in the $10 to $25 range. Until then, be wary of oysters in August.
Digital Spine Cochlear implants (see #9) and other devices already use chips to send electronic messages to various nerves in the human body. Eventually—though no one's saying when—
that could lead to a cure of sorts for paralysis.
Such a "walking" device will probably require sensors in the feet or on the legs to record information about your gait and to feed data to a chip in your spine. The chip will translate the data into electrical impulses and send them to the brain via nerves. We'll also need a neural network chip to send messages back to the legs and feet. Neural network chips try to mimic the biological network of the brain to recognize patterns and process data in quasi-intuitive ways.
How long? Probably at least 10 or 20 years away, depending on how much money there is to spend on the problem. "That's what these technologies need," says Jeffrey Greiner, president of Advanced Bionics, which is developing neural stimulation devices. "Lots and lots of money."
Sean Donahue and Nate Hardcastle also contributed to this story.