Monthly Archives: November 2005

Esperanto for Toasters

In the not-too-distant future, your cell phone might become the key to your home. By transmitting a signal to a sensor, your phone will announce your arrival and the front door will unlock.

And that’s just the first step. Transmitters in the door will send signals elsewhere in your house, switching on the lights, turning up the heat, warming up the hot tub, queuing up your favorite MP3s on the home theater system, and telling your home computer to power on and download the latest e-mail.

Such fantasies have been a staple of the home automation market for years, of course. They’re already being tested in Japan, and they’re a bit closer to reality now in the United States, with an emerging home networking standard called ZigBee and some close competitors.

“We’re rapidly approaching a world where the most important devices in our lives are ones we don’t even realize exist,” says forecaster and strategist Paul Saffo, a director of the Institute for the Future, in Palo Alto, CA. Just as cheap microprocessors spawned the PC revolution of the 1980s and inexpensive lasers enabled the telecommunications and Internet revolution of the 1990s, Saffo argues that cheap sensors are ushering in a revolution in intelligent, interconnected devices, many of which will operate quietly in the background, without drawing any attention to themselves.

But before a swarm of sensors can turn into an intelligent network, though, they need a way to communicate with each other. Enter ZigBee. Based on an IEEE radio standard called 812.15.4, it allows digital transmissions of up to 1Mbps in one of two frequency ranges, 2.4GHz or 915MHz (in the Americas).

Overseeing what radio engineers call the upper “layers” of the ZigBee specification, the ZigBee Alliance governs such issues as how packets of electronic information are routed between ZigBee transmitters and receivers, and how these devices interface with various software programs. The alliance also certifies compatible devices and promotes the standard — in the same way that the Wi-Fi Alliance promotes, certifies, and helps develop the now-universal IEEE 802.11 set of wireless networking standards.

More than 150 member companies already belong to the ZigBee Alliance, including such electronics heavyweights as Honeywell, Motorola, Philips, and Samsung. Alliance chairman Bob Heile claims that ZigBee will enable any compatible device — regardless of the manufacturer — to communicate with any other ZigBee device, right out of the box. What’s more, the specification allows ZigBee devices to form mesh or cluster networks spontaneously, without any intervention from end users, installers, or (gulp) system administrators.

“When they’re being put together by people who don’t know beans about networking, these devices have to be intelligent enough that they organize themselves into a network and maintain the network if something breaks,” says Heile.

ZigBee doesn’t require high power consumption and makes it easy for devices to go in and out of a low-power sleep mode, so a ZigBee device should be able to run for years on an AA alkaline battery, Heile says.

That may not be an issue for appliances, such as lights, which are plugged into the power grid. But for battery-powered devices, such as remote controls and smoke detectors, power consumption is a key consideration, says Heile. This is a primary reason why existing wireless standards, such as Wi-Fi and Bluetooth, aren’t appropriate for home automation.

ZigBee’s success, however, is far from assured. Because the specification is just one year old, there are very few ZigBee devices currently available to consumers. One exception is a home entertainment and control system from Control4. Heile also points to a test network of 2,000 ZigBee nodes created by South Korea’s SK Telecom, which is investigating the possibility of including ZigBee radios in its cell phones.

Meanwhile, several competing home automation standards are also on the market or under development. A leading contender is Zensys’ proprietary Z-Wave standard, which has been out for several years. Z-Wave is a low-cost, somewhat less capable alternative to ZigBee — but more than 75 compatible products are already available for purchase, which may give Z-Wave a leg up.

Smarthome’s Insteon offers many of the same advantages as ZigBee, say analysts, but — also like ZigBee — it’s unproven in the market. Finally, there are communications systems that transmit information over home powerlines, such as X10. So far, these devices have very limited capabilities, although they have the advantage of tapping into a pre-existing infrastructure.

“We don’t expect ZigBee to have a major impact in this space for at least the next two to three years,” says George West, a senior analyst at West Technology Research Solutions LLC, in Mountain View, CA. That’s primarily because ZigBee devices cost more than alternatives like Z-Wave, and the standard is more complex than most home automation products currently require, says West. For this reason, ZigBee may ultimately be better suited for automation in commercial and industrial environments, such as hospitals, office buildings, and factory floors.

A few years from now, however, the story may be very different. Strategy Analytics predicts that the market for wireless mesh networking chips, including ZigBee, as well as Z-Wave and other proprietary solutions, will reach tens of millions of units annually by 2008.

On the other hand, the hurdles are not trivial: the market is relatively new, there are several competing standards, and there is no pressing consumer demand for home automation. “This stuff feels poised for takeoff — but three years from now, it may still feel poised for takeoff,” says Saffo.

One thing is clear: As more and more consumer products gain sensing capabilities and start interacting with the world around them, the value of networking them will grow. And ZigBee could be the way.

Dylan Tweney is a writer and editor in San Mateo, California.

Link: Esperanto for Toasters

Link broken? Try the Wayback Machine.

Malevolent design.

Sure, intelligent design is bad science, because its central propositions can’t be tested — more precisely, there is no test that could show them to be false. But doesn’t it give you an extra frisson of schadenfreude to know that ID is also bad religion? J. M. Tyree points out ID’s theological underpinnings, and summarizes some of the best arguments against it, going back to Hume.

Despite the new cloaking device of pseudoscientific language, ID is actually a recent mutation of one of the oldest, most persistent, and most tempting of religious ideas, the so-called “teleological argument” or “argument from design.” … The most devastating objection is that even if you assume the world was designed, it does not appear to be designed by a very nice deity. … The Designer who so Intelligently Designed our world, in theory, could be malevolent or capricious just as easily as he could be all good. He might have designed us intelligently, but for the purpose of watching us tear each others’ throats out.

In other words, God might be Cthulhu … or maybe even the Flying Spaghetti Monster.

I rided a bike to school.

by Clara (transcribed, and annotated, by a teacher at preschool)
diagram of preschool, by Clara


Once upon a time I rided a bike to school. Cause it is hard to get to school on the freeway cause there is traffic. Traffic is so you can’t get through. It never ever lets you. Traffic is cars and they make sure you don’t get to school and you don’t get any energy. Energy is so you can play. Adults use energy to make sure kids don’t get hurt. Riding a bike feels like we’re on a roller coaster. It feels like we’re on a train in the cars. My body moves both sides. My job is to hold on tight to my friends so they don’t fall. Stuffed friends. Daddy’s job is to pull all my straps so I don’t fall down.

My favorite thing about riding a bike to school is moving and going so fast. I like to play. You get to play when you get to school.

The End.

10/11/2005

Palm Today, Gone Tomorrow

Palm lovers, get out your hankies. Thisstory is a real tearjerker.

The company that single-handedly createdthe personal digital assistant with the Pilot 1000in 1996 is at a crossroads. Down one path liesWindows. Down the other path: stagnation,decay, and perhaps death.

If you buy a Treo next year, there’s a goodchance that it will be based on the WindowsMobile operating system, not the venerable PalmOS that has powered all prior Treos and PDAs.At press time, Palm had made no announcementabout a Windows-powered Treo, and Palmrefused to comment for this story. But all signspoint to an imminent platform switch.

FOCUSED ON PHONES
Let’s get one thing cleared up right away: Palm’sdays as a leading vendor of PDAs are over. Infact, the company’s share of the worldwide PDAmarket has been steadily shrinking from its highof 68 percent in May 1999; it currently stands atabout 18 percent, according to Gartner.

What’s more, the PDA market is increasinglyirrelevant. Today’s phones are more capable andcan hold more data than 20th-century phonescould. Why carry two devices when one will do?Palm has seen the writingon the wall. “There’s noquestion that the traditionalPDA business has declined,”Palm CEO Ed Colligan saidin a June teleconference.Accordingly, the companyis putting its most intenseefforts into the smart-phonebusiness.

“Ever since Palm acquiredHandspring, it has reallyfocused its resources onbecoming a stronger playerin the smart-phone market,”says Todd Kort, a principalanalyst for Gartner.

PALM TUCKERED OUT
While its PDA market founders, Palm’s smart-phone businesshas taken off. The Treo has been a tremendous success, drivingPalm to a whopping 50 percent of the U.S. smart-phone market(though it holds only 5 percent of the market worldwide)and helping propel the company through two solid years ofunbroken revenue growth. In retrospect, it’s lucky that Palmbought Handspring, the developer of the Treo, in 2003.

Unfortunately, the underlying operating system isn’t doing sowell. Palm OS 5.4 is a dead-end street. It has poor support formultimedia features and lacks multitasking capabilities.Windows Mobile, by contrast, has been multitasking foryears. It’s also got a built-in web browser, extensive support foraudio and video, and corporate-friendly security features.

Palm OS developer PalmSource (which was acquiredby Japanese software developer Access in September)attempted to bridge the features gap last year with a brandnewoperating system, Cobalt, which was multimedia- andmultitasking-friendly. But the OS was a fl op, requiring toomuch memory and processing power to be practical, and nocompanies ever released a Cobalt-based device.

Implicitly acknowledging the failure of Cobalt, PalmSourcehas stated its plans to switch to a Linux-based architecture. Buta commercial version of the new Linux OS won’t be availableuntil late 2006, with devices based on it unlikely to be readyfor sale before 2007. Palm can’t afford to wait that long.

END GAME
M:Metrics senior analyst Seamus McAteer predicts that Palmwill announce a Windows-based Treo by January 2006.Over the long run, that will spell the end of Palm OS-basedTreos. “Who wants a Palm-based Treo when the company hasannounced that it’s migrating?” says McAteer.
Without a major upgrade to Windows, the company maywell be dead in the water, increasingly unable to compete withsmart phones that offer more whizbang features.
But it’s a risky move. Many of Palm’s customers have stuckwith the company simply because it’s not Microsoft. “Goingover to a Microsoft operating system will not win the companyany friends among its current customers,” says Gartner’s Kort.And with Palm accounting for 60 percent of PalmSource’srevenue, losing even a fraction of that business would be aserious blow for PalmSource.

Still, Palm’s executives aren’t likely to spend a lot of timeworrying about the fate of their sister company. Palm has towalk a delicate line between a dying operating system and apack of anti-Microsoft zealots. Whatever decision they make,they’re likely to piss off somebody. -Dylan Tweney

Link: Palm Today, Gone Tomorrow

Link broken? Try the Wayback Machine.