A Letter in Scarlet*

A language of fingers

The MEMS Executive Congress last November had it own share of eye candy. It wasn’t just the Monterey Bay (and dinner in the Aquarium) that tantalized the 280-odd attendees. Rather, it was the MEMS Technology Showcase, featuring some dramatic applications of micro-machines themselves:

• MOD Live from Recon Instruments is an interactive display for GPS-enabled ski goggles. It’ll map your route on the slopes, tell you where you’ll hit bumps and crevices in the snow, and monitor your performance.
• The Sunrex Air Mini Keyboard is a remote control with integrated keyboard which supports PayTV, PC-connected home theater and other interactive media applications.
• Sys-Evo is an electronic surfboard attachment that monitors wave height, wave direction, distance traveled, and the duration of the wave — as well as the surfer’s actions (including waiting time, paddling, riding and time waiting to catch a wave.)
• R0R3 Devices' latest contraption is a MEMS-based heart rate monitor which correlates chest strap electrocardiograms (ECG), with periodic motions such as running, push-ups or jumping jacks. (These are the folks, remember, who engineered “Guitar Hero.”)
• The Raisin System from Proteus Biomedical combines medical monitoring, and daily pharmaceutical administration with social networking. It doesn’t threaten to tell your Facebook friends if you’ve missed a pill — but something close. The Raisin System is based on a chemical sensor called the “Proteus Ingestible Event Marker (IEM)” and the HealthTiles application environment.
• And MicroVision is back with the “ShowWX+ HDMI,” a handheld pico projector that lets users project images, presentations and videos as large as 100 inches across from an iPhone, iPad or laptop onto a screen, a wall or any other available surface.

The market research services will have field day tabulating the market potential for these eye-catching but somewhat niche devices. Gartner’s own revenue projections focus tightly on semiconductor MEMS, and are much more conservative than those counting digital light projectors (DLP) and/or inkjet nozzles. Still, revenue projections for micro electromechanical systems (MEMS) devices in cell phones, it should surprise no one, resemble the proverbial hockey stick. In a dismal 2011, it was among the only semiconductor devices that exhibited 16% growth. And all projections suggest MEMS sensor revenue will continue to grow in the foreseeable future.

Where’s Elmo?

There are two glamor applications supporting the use of MEMS devices in mobile handsets and tablets. One is the support for the location-based services some people are starting to call “augmented reality.” Global positioning systems (GPS) depend on a clean “line of sight” between your portable navigation device (PND) and the position-detecting satellites. You lose this line of sight inside a department store or a shopping mall. MEMS motion sensors can serve as footstep counters, keeping track of where you are inside this canopy. A magnetic sensing element (like a magnetometer) adds a digital compass to this tracking technology.

Robert Bosch, the ostensible leader in MEMS accelerometers (particularly airbag triggers) used the MEMS Executive Congress to show the parts it was developing for cell phones with location-based service apps. Bosch is packaging three devices — a MEMS accelerometer, gyros and magnetometer — in the same integrated circuit package. Augmented reality adds an application layer to the dead reckoning provided by this combination: The shopper entering a food court for the first time, believing there was a Starbucks close by, but unable to spot it, need only snap a picture of the court with the camera portion of his smartphone. An overlay, perhaps with an arrow, superimposed on the picture you’ve taken will point out the Starbucks in that scene.

Some application developers, like Qualcomm, regard this “where’s Elmo?”scenario as a fairly primitive application of augmented reality. A more developed AR application might take a picture of a Chinese restaurant menu in Shanghai, and the overlay will give you the English translation. In the demos we’ve seen, Qualcomm’s applications processors can place an avatar within the Starbucks scene you’ve captured, and guide you step-by-step through the purchase of a latte (or challenge an imaginary barista to a stick fight).  

MEMS will also be the enabler of a technology we’ve come to call “gesture recognition.” Thanks to iPhones and other screens equipped with capacitive touch sensors, we’ve begun to learn a new language — the language of fingers. Think of this: You hold your thumb and index finger together, place them on image on your touch screen, and spread your thumb and index finger apart. The applications processor understands that as intent to zoom-in on your picture image. Conversely, touch the image with your thumb and forefinger, and squeeze them together. The applications processor understands this as intent to zoom-out of the picture by decreasing the size of the image. Or, do something else: swipe a finger across the image, right to left, and watch the image (or icon) leave the screen, quickly replaced by another… I don’t want to do a thing on the genius of Apple, Steve Jobs, blah blah blah, but it IS something to think about: Did this language of fingers exist prior to the introduction of the iPhone?

This is really just a beginning.

Thanks to a proliferation of cameras pointed at us (on our laptops, tablets and cell phones), and the increasing sophistication of visual pattern recognition — as well as a proliferation of MEMS sensors — we’ll have entirely new ways of interacting with our computing equipment. We’ll have to think twice before we grimace, shrug, wave our hands, clench our fists, or use our middle finger.

The MEMS devices in cell phones and tablets are the early enablers of gesture recognition. Already, the embedded gyros and accelerometers in our mobile handsets allow them to be used as gaming attachments. We can swing our smartphones as if they were the handles of tennis rackets, while a pico projector with the gaming software will project the trajectory of the imaginary tennis ball we’ve hit over the net into our opponent’s court. Incorporated into tablet computers, the MEMS motion sensors convert your tablets — the entire slate — into a steering wheel for a projected racetrack game, or an assault weapon for a first person shooter.

Careful what you wish for

The rotational sensors in the Amazon Kindle Fire work pretty well, but the touch screen is half asleep. You can turn the pages of your eBook fairly easily with a finger swipe, but moving from an eBook app to (say) web browsing requires several finger taps before the machine registers that it’s supposed to do something. .

I had obtained my first Kindle with the help of Freescale, who makes the iMX processor used in most of the Kindles (and several other eBook readers). As a long-term book lover, I was surprised and amazed by how easy it was to get used to Kindle… No more sore shoulders from the hardbound books in my briefcase… The Kindle gives me access to hundreds on slate as small and as easy to carry as the steno pads I use for note taking… and, as a consequence, I feel I’m reading quite a bit more.

I couldn’t resist the Kindle Fire, which is reputed to do almost as much as the Apple iPad — and costs $300 less. While I really like this Kindle as well, I’m more aware of its short comings. Unlike the original Kindle, which has universal WiFi, the Fire depends on a “hot spot” connection… a problem since most of your content (including the book you’re reading) is in “the cloud,” which means you’ll wind up paying a fee at many airports. This is not a problem inside my house where everything is connected to my personal encrypted wireless network, but a problem everywhere else.

The Kindle Fire is surprisingly heavy, and the width of the touch screen doesn’t give a convenient place to grip it. A problem if you want to hold the thing with one hand on a crowded bus, since the pressure of your thumb on the screen will be interpreted as an attempt to turn the page. And the screen's frame is otherwise too narrow to grip it firmly without touching the screen.

And the battery life isn't great. You can get about a week with intermittent eBook use; much less, not even a full day if you use the Internet for streaming media. But the machine does excel as a media player. Powered by one of Texas Instruments’ OMAP processors, the 4430, the Fire does a terrific job with MP3 music playback, JPEG picture display, or movie viewing. The sound is vivid and the screen is as enticing as any portable media player I’ve seen. Using the finger language I talked about — thumb and forefinger — I can expand or collapse images with fairly high screen resolution… which is ideal for (say) showing pictures from Europe.

Can’t rub two dimes together…

I apologize to readers for my long absence from this space. I had spent a good part of the fall — four weeks in fact — two weeks in Europe (Berlin, Prague and Paris) and another two on the east coast (the Lake Country of New Hampshire, first, and the suburbs of New York). I was pursuing a writer’s obsession that I could neither quench with overseas travel nor put to rest with isolation in rural New England. All I can say now, if I ever have the time to finish this novel, it will be at once as steamy and philosophical as anything Philip Roth turns out these days.

As it turned out, a colleague was interested in hearing about my time in Berlin, the scene of so many John le Carre spy thrillers. “Glad you have the resources to travel,” she said, a reference to the economic downturn so many people were feeling.

“I did cash in on some airline mileage I had accumulated,” I explained, almost apologetic. “I’m not rolling in dough.”

“You know what this is?” she said, showing me her hand, her thumb working back and forth against index and second finger.

 I thought her hand gesture pre-dated the iPhone by several centuries, a reference to what some cultures would call “baksheesh.” But I played dumb in case a different story emerged. “No, what is it?” I asked.

She beamed at me as if the answer were entirely obvious: “The world’s smallest violin!” []

Tragedy strikes the analog world… twice

I’ve come to dread middle age, where ─ despite your attention to diet and exercise ─ life threatening health issues seem par for the course. “I resent getting older,” I told a close friend (15 years ago), “every chest pain is a heart attack; every headache is a stroke; every stomach ache is cancer.” No jest. I’ve become wary of emails, of the EE Times Newsletter, the same way my father would warily peruse the back pages of his local paper, The Asbury Park Press. He could never tell when a friend would show up on the obit pages, but over time the probability increased, and, now watching editorial colleagues and mentors pass on, I experience the same trepidation.

It was with great remorse that I read of the death of Linear Technology Corp’s staff scientist, Jim Williams, and then, Monday, the death of National Semiconductor’s staff scientist, Bob Pease. I picked the news off my Blackberry just after the plane I was on landed in Dallas. “Oh no!” the people around me heard me exclaim.

Both Jim and Bob were ostensibly stroke victims. Both were analog design geniuses with many devoted followers among the engineering community. Both were quirky, charismatic and iconoclastic personalities. These were colorful people, about whom many stories will be told, for years to come.

Jim Williams was suffering from Parkinson’s tremors when I saw him at Paul Rako’s “analog aficionados’” party in February. His boss and long term mentor, Bob Dobkin, told me Jim’s condition didn’t impair his work; it was just when he sat still that the tremors became noticeable. I don’t know whether Jim’s Parkinson ’s disease had any relation to the stroke that killed him.

Bob Pease came to Jim’s memorial service at the Gallo Mountain Winery in Saratoga. But coming down the hill, his Volkswagen Beatle left the road and crashed into a tree. Bill Schweber’s account seemed to suggest that Pease had some sort of stroke himself, and was dead before the car actually crashed.

 The nose cone of a Minuteman Missile

Williams liked to build stuff, clever circuits that others couldn’t quite dream up. On one visit with him in Palo Alto, he showed me something distinctly illegal: a cell phone jammer. The machine, built into a briefcase, would spread radio hash across the dominant cell phone spectrum. It used to irritate him when someone entered a coffee shop, speaking loudly on a phone.  He thought it was rude and inconsiderate of the other coffee shop patrons, and he would turn on the jammer. He’d watch as the talker would suddenly stop, take the phone from his ear, and stare at its display screen. “Why do they always look at the screen?” he laughed, as if he were reviewing a tape from candid camera.

I had visited Jim at his home to take pictures a piece of wall art I had come to think of as “The Tapestry.” That’s certainly what you’d think it was if you saw it from the sidewalk, looking through his living room window. But up close, you’d recognize the Tapestry as an elaborate display of ancient electronics. These were circuit cards from the nose cone of a Minuteman Missile. There were arrays of color coded resistors and capacitors, transistors with gold and silver metal caps, all on green phenolic circuit cards.The pictures (different views of the circuit cards) were published as cover art for EE Times’ Planet Analog magazine supplement.

The colors of the electronic components were so vivid because the electronics had been soaked in liquid Freon. The missile was permanently armed, Williams explained, but the electronics ran so hot, that Freon was required to prevent the machinery from burning itself up. Without the coolant, the system would last about 10 minutes before burning up ─ roughly equivalent of the flight time of the missile to its intended target.

Jim had acquired the nose cone electronics during his student days at MIT, a period I wrote about in the “Mixed Signals” newsletter (a paper version) I once published. That article, too, described some of the goofy circuits Jim used to build. Published in the days before anyone knew how to use the Internet, the Ventura Publishing files are forever lost.  But the tribute article I wrote was appropriately entitled: “The Boy Who Loved Electronics.”

The three-cup brassiere

If Jim Williams seemed “the Boy,” well into his sixties, Bob Pease, bearded, regal was more your stern father figure. As I wrote on the PlanetAnalog website once, Pease took analog circuit design very seriously.** He was concerned that his colleagues “got it right,” and would often send them handwritten notes with barbed comments and suggestions. Dan Sheingold, the long term editor of Analog Devices’ “Analog Dialogue” was among those glad to have received what became known as a “Pease-o-Grams,” and proudly saved it.

Pease, if the truth be told, did not suffer fools very well. Early in his career, Pease had a run-in with the president of the company he worked for (Teledyne Philbrick), and made it absolutely clear how he felt about attempts to micro-manage engineers. Philbrick’s president, Bill Earley, an appointee of the Teledyne corporate parent, may not have known much about the analog engineering role when he publicly remarked (apologies in advance to the women in my audience): “If we let engineers talk to customers, we’ll wind up making three-cup brassieres.” The remark must have enraged Pease, whose successful 4701 voltage-to-frequency and 4702 frequency-to-voltage converters were by-products of interactions with customers.

Guess what Earley received for Christmas? According to the late Frank Goodenough, the Electronic Design magazine analog editor, who worked at Teledyne Philbrick at the time (circa 1975), Pease wheeled it in on a scope cart during an office party. Cameras were snapping away. Under one cup of the bra was the 4701; under the opposite cup was the 4702. And under the middle cup was Bob’s resignation from Teledyne.

The account published by Pease in Electronic Design magazine was less venomous and more politically correct than the version that appeared in the “Mixed Signals” newsletter. The office party was actually a “Golden Camel” awards ceremony, the engineering team’s acknowledgement of management support and/or meddling. It was a “roasting” based on the parable of a desert Arab who makes the mistake of letting his camel put his nose in the tent. In Pease’s account, the artifice that carried his resignation was a camel’s back with three humps.

One other thing that Williams and Pease had in common: They loathed the EDA tool vendors who believed automation could supersede personal analog design knowledge. Pease’s familiar refrain was “why use Spice if it’s going to lie to you?” And Williams, in a luncheon address at the one of the analog conferences I used to organize for “Computer Design” magazine, flashed a picture of a computer terminal. “Is this the way you do your engineering?” Williams asked. A glass of red wine was positioned close to the computer screen, and a rose was draped across the keyboard.  The audience laughed.

Williams and Pease were colorful personalities whose antics will be remembered for a long long time. As I wrote years ago, at the death of Frank Goodenough, with the passing of these people ─ Bob Widlar, Art Fury, now Jim Williams and Bob Pease ─ we enter a new and entirely different emotional era. The engineering world will be painfully dry without them. []

 

**A correction to the original EE Times' article mentioned here (and I'm grateful to Dan Sheingold for looking over my shoulder at this): ADI’s seminar books were re-published by Newnes. These include “The Op Amp Applications Handbook,” by Walt Jung, “The Data Conversion Handbook" by Walt Kester, and "The Linear Circuit Design Handbook" by Hank Zumbahlen. Much of ADI's seminar material is now available online through its website with podcasts, "Circuits from the Lab," and "The interactive Engineer Zone." See http://www.analog.com/en/index.html

Amazing but scary

I was in love with the Panasonic DMC-FZ35 digital camera long before I plunked down my credit card. I would go regularly the Digital Photography Review site and drool over the sample images (all those wonderful shots of London Bridge, the city skyline, and colorful shopping plazas within). And I would periodically visit Best Buy, Amazon and J&R Music World to check pricing. And behind me ─ bet on it! ─ was a robot salesman, watching my clicks, recording my interests, sending me advertising and ultimately, a year after my first explorations, closing the sale.

To say I got a $400-camera for $220 would be a bit factious. Panasonic had brought out an update to the FZ35, the FZ40 (for about $300, street price), though user reviews suggested that the FZ35 was still the better choice. But in an effort to clear inventory on the older model, Amazon and the like dropped their prices, and sales approaching Christmas were infectious. I had to make a judgment call as to how low the price would drop before inventory was depleted, and I think I called it right. The day after I made my purchase, the robot-driven side panels next to my AOL email messages were showing the FZ35 and saying things like, “Act now! Only 5 left at this price.”

For those who haven’t quite caught on as why their gmail, AOL, Yahoo, or Hot Mail services are free, let me remind you that it’s paid for by advertising. And thanks to spiders, software robots and other scanning mechanisms, that advertising is very precisely targeted. The spiders will root through your emails, keep track of the websites you’ve visited, and likely compile a profile. If you have a special hobby, are interested in buying something, the machinery knows about it. Check the side-panel advertising to the right of your Facebook profile: How do these people know you’re cyclist? A cancer patient? Or getting married?

I took a survey about my movie preferences on Netflix, knowing my preferences would help them make recommendations to me. They gave me a not-overly-long list of movie types and titles, and asked me to rate them on emotional impact. I tend to think foreign films are more interesting than American films; I like science fiction (concept films, at least), but I have a limited tolerance for movie violence. (The "Millennium Series," which I wrote about last time, was a stretch for me, not because of the subtitles, but because of the sadism toward women. Alfred Hitchcock’s classic “Psycho” was similarly a stretch for me, though Hitchcock intended it to be a stretch for everyone.) I think there were 15 films the Netflix survey asked me to rate.

The outcome of the survey is that, now, Netflix totally “gets me.” Every time I go to the site, looking for a DVD to watch, Netflix’s recommendations are 100%. Every film title they recommend is either something I’ve already seen and enjoyed, or one which has been high up on my list of things I’d like to see.

But Netflix solicited my cooperation. It’s a little disconcerting when AOL or Yahoo comes down unbidden on your consumer preferences, making you wonder whether they know more about you than your spouse does. As one listener protested in a radio perspective piece, “this is amazing... but scary.”

Most call it stalking

Oddly enough, few people are screaming about invasions of privacy. I might have mixed feelings about the ad service that tracked my camera preferences, but then I’m also on record seconding Scott McNealy’s conjecture that privacy on the Internet is a pipedream

In an article about software ad servers, a CNET reporter claimed ads posted by advertising service Specific Media didn’t offend her enough. As a reporter with a byline and a following, she had expected the spiders to have collected a lot of information about her. To say she was interested in technology, the Internet and digital cameras describes a lot of us, but she had no idea why the ad server said she was interested in “heavy construction. ”I must have hit on some housing or building site to trigger that one,” the reporter, Elinor Mills, concluded.

“Behavioral targeting tracks and analyzes your online behavior, tracking what you search for, what websites you visit and what services you use. This could be used to determine what ads are shown to you and where, even having the same ad follow you from site to site, a practice known as retargeting (or remarketing),” Mills wrote. “Outside the advertising industry, most call it stalking.”

The Huffington Post offered some pointers for getting these robots off your tail, but acknowledged it maybe a losing battle. “The most straightforward approach to combating targeted web ads is via the Network Advertising Initiative (NAI), a cooperative of online marketing and analytics companies. The NAI offers some “opt out” tools for killing cookies and other tracking markers. The Huffington Post also points to the PrivacyChoice.org for a larger assortment of tracking killers and smoke screens.

As a reporter with a very public profile, I admit to having mixed feelings about this exposure. I cannot change hats (literally) without somebody noticing. As far back as 1998 and 1999, when my then employer, CMP Media, began setting up its web portals, Internet jocks in the company imagined how USEFUL the websites could become to our visitors if only we knew more about who these visitors were and what they were interested in. Engineering readers were resistant to register to obtain access to the websites, but following someone’s click patterns could tell you a lot about what content you should be providing to keep that visitor coming back to your site. If visitors keep coming back to (say) your microcontroller pages, maybe you could invigorate your site with more architecture and software subroutine choices – and get a programming tools advertiser to sponsor this.

Jump ahead 12 years: pattern recognition, along with the impact of social media is high up among Gartner research agendas for 2011. We see this not as a threat, but as a mechanism to help IT system administrators detect and rapidly respond to indicants of change. Think of the possibilities: If you’re music critic at the increasingly unmanageable South-by-Southwest music festival in Austin, you just follow the real-time text message streams to find the best musical acts and venues. If you’re a stock trader, you can make enormous amounts of money on penny swings of a particular stock – if you can catch them fast enough. (I used to wonder why all these $500-million fund managers  I was presenting to were all 26- and 28-year-olds, until I realized these young geniuses – fresh out of Wharton or Harvard Business School – wrote the complex mathematical models that would help them capture these patterns.) Conversely, early pattern recognition could help prevent fraud and deceit.

Fast on the trigger

Still, I cringe at the speed with which ad servers can react (and am steeling myself for the ads this blog will bring). We should be grateful, perhaps, it’s a robot reading your email, and not a real person. A lady friend of mine uses the online dating service, match-dot-com, as a vehicle for meeting available men in her age group. I think it would not unfair to say that most people advertising for social connections this way do not live up to their promotion; the men are generally older, heavier and less financially solvent than their pictures and online profiles seem to suggest. Sometimes people live up to their profiles, my friend tells me, but beyond a handshake and/or a peck on the cheek, it’s often hard to get a romantic relationship going.

Still, my friend – a former actress, model and press agent -- has a great sense humor and irony. She takes a you-need-to-sort-through-a-lot-frogs-before-you-find-the-handsome-prince attitude, and has some marvelous stories to tell. However you interpret her online profile, there is no doubt she is looking for someone handsome, intelligent, and successful in his career.

Thus, it was with great interest that I read a recent email from her, saying she had made the breakthrough connection: Prince Charming himself. They were exchanging lengthy emails, and getting set to meet. She was so thrilled. He was a world-famous economist, she wrote me, widely published, an advisor to corporations and congressional subcommittees, a frequent guest on the McNeal Lehrer News Hour and other NPR interview programs.

I gave my friend the benefit of my doubt: highly successful professional people do indeed show up periodically in online dating sites. “Maybe I ought to meet him,” I wrote back, partially jesting and partially thinking this guy would make for interesting conversation, if not become a keynoter for one of the conferences I wind up supporting.

I’d like think that what happened next was a coincidence; that there wasn’t a spider reading my mail.

The side-panel advertising on the very next email I received said this: “Meet available singles in your area, on match-dot-com.” And the pictures accompanying the ad were head-and-shoulder shots of handsome young men.[ ]

 

Why I need a Kindle

On impulse, I picked a bookplate off the sidewalk near my house. It was the title page for something called “The Future of the Book,” edited by linguist Geoffrey Nunberg and published in 1996. It was sopped with rain, but I thought to carry it home, dry it out and have it framed. It struck such a nerve.

Like a lot of educated people, I’ve been in love with books. Hardbounds. Paperbacks. Fiction. Non-fiction. Coffee table art books. I love not just to read them and absorb their contents, think about what their authors went through to create them, but also to keep them around me, as if they were old friends (or new ones). I certainly couldn’t ride a MUNI bus without spending some time with a book. (“Oh: You’re one of these guys who reads for fun,” a disillusioned school teacher once teased me. But she spent her days with kids for whom cracking a book was a torture, and I could see how my hobby bemused her.)

I dumped 2/3 of my books ─ just left them on the street ─ when I moved from Long Island to San Francisco ten years ago. They were mostly college texts and cheesy paperbacks (including some Peter Drucker and Ayn Rand titles) which I don’t particularly miss. Still, I love the image of myself surrounded by books, floor-to-ceiling, as in the typical book jacket author’s photo. I love places like Bell’s Books in Palo Alto and Powell’s Books in Portland, where they group books by authors and don’t make too many distinctions (except in price) between what’s new and what’s used or Kramerbooks and Afterwords in Washington DC, where they might even feed you. San Francisco’s Green Apple Books is just around the corner from me, but I have to be judicious about dropping in there: It could be a $200-bill.

Though not quite floor-to-ceiling, I have books in pretty much every room of my house. Living room, bedroom, and front hall ─ even the kitchen and bathroom. But as much as I like having books around me, it occurs to me I won’t be able to take these with me when I move. Even if I don’t wind up in a London studio, with children and home ownership behind me, it’s time to think about downsizing. As much as I love my books, I’m just not going to be able to take them with me.

 

Love and Death above the Arctic Circle

I wasn’t thinking about relocation on my last trip to Europe. (Okay, maybe I was, but that wasn’t what got me thinking about a Kindle.) The 10-hour flight was a great time to catch up on some reading. I had three books with me, and as I’d finish one, and move onto the next, I’d wonder, “Okay, what do I do with this?” My briefcase was pretty heavy.

I’m ordinarily not a mystery fan. I know people who gobble them up like barreled peanuts, cracking the shells, crunch-crunch-crunching briefly on the seeds, tossing the shell, and reaching quickly for another. It’s a genre I needed to learn. (I would hesitantly identify science fiction as my “junk reading,” though few fans of William Gibson would identify likely put his books and the same consume-and-toss category as peanuts.)

Rather, following the incredible popularity of the Stieg Larsson Millennium series (beginning with "The Girl with the Dragon Tattoo"), I discovered a whole library of somewhat bleak detective fiction set in Scandinavia ─ Sweden, Norway and Denmark ─ as well as Finland and Iceland. Wisconsin Public Radio underlined this genre in a radio program called "The Scandinavian Death Trip." The Nordic Noir novels were easy to consume (fast reading), compelling in their own way (you want to know who the killer is), but I couldn’t help wondering what do they say about life ─ modern life ─ in these cold and dark places.

In all the Nordic Noir novels I read, the cold and dark of the place (especially in the winter, where the sun sets one day and doesn’t come up for six months) is a character, a palpable presence, dark and foreboding. In the Finland depicted by James Thompson in “Snow Angels,” for example, people gage the temperature according to the speed with which their nostril hairs freeze. Ski resorts with luxury condos have sprung up in northern Finland (far from Helsinki). On weekends, the icy roads are clogged with German-made SUVs. But the working-class people who live there year round often turn to drinking when the winter sun sets; it makes them surly and sets the stage for violence.

The changing demographics of these northern countries (the place of ethnic minorities, for example) seems to be a common theme in the mystery novels I read. In Thompson’s “Snow Angels,” the first person murdered is a Somalian immigrant, leading the detective-hero to ask whether this was an act of racial/ethnic prejudice ─ especially as politicos race to keep the crime away from the news media. In Arnaldur Indridason’s “Arctic Chill,” set in modern Reykjavik, the murder victim is a Thai immigrant, and the first suspects make public noises about “Iceland for Icelanders.”In Karin Fossum’s “The Water’s Edge,” the prime suspect is a pederast, but the hunt for the child killer forces the detective-hero to ask how Norwegians actually feel about homosexuals, “people on benefits” and other minorities.

The woman who took me to see the film version of “Girl with The Dragon Tattoo,” and the Finnish woman who took me to see the sequel, “The Girl who Played with Fire” had a lot in common with their admiration for Lisbeth Salander, the main character of Larsson’s series. She was more than an expert computer hacker. She was the victim sexual abuse, depicted in often uncomfortable detail in the films. But she doesn’t wimp away; she extracts a carefully-choreographed revenge, and the movie audiences cheer when she succeeds. I don’t know what this says about the status of women in modern Sweden, though it strikes me as a long way off from the painful psychological introspection depicted by Ingmar Bergman in (say) “Scenes from a Marriage.”

I lent Thompson’s novel, “Snow Angels,” to my Finnish friend, who spent her early years in Lapland, above the Arctic Circle. She had to read in fits-and-starts, she said; Thompson’s descriptions made her homesick. She remembered with nostalgia the Finnish custom of removing your shoes when you entered someone’s home. She cringed at the Finnish curses Thomson repeated, and remembered seeing the Northern Lights in Finland’s star-filled winter sky ─ “which is why,” Thompson’s detective hero reflects, “I believe in God.”

But my friend had to wrestle with the notion of Finland as destination for Somalian immigrants. The whole notion seemed strange. The native people, like the Swedes, were predominantly blonde-haired and blue-eyed. She never even SAW a black person up close, she confessed, until she emigrated to the U.S. in the 1950s.

The Thompson novel would make a great movie, I think, gauging by current standards. The detective hero gets shot in the face at the end, but survives, probably to make his way to the inevitable sequel. (I can’t help thinking of Bruce Willis who plays roughnecks who survive bloody beatings film after film, though the part of a world-weary Finn might be something of a stretch.)

In the end, mystery writing seems governed by formulas: Begin with a horrific crime (the murder of a child or the mutilation of beautiful movie star). Throw in an aging, cynical (seen it all before) detective, one who may drink a lot and have some personal involvement with the suspects. Mix in a bunch of clues, including red herrings, often earnest theories about what the forensics of the crime suggest, and expand the number of suspects. End with the murderer, explaining or lamenting his crime, in elaborate detail, even as he threatens the life of the detective hero.

I faithfully lugged these books to and from Europe, and added them to my floor-to-ceiling collection when I returned home. But I had in fact absorbed the formula.

 

The Future of the Book

Freescale, whose iMX processor forms the heart of the Kindle, claims the largest penetration among eBook readers. This is area to watch closely, not just as bookworms like me convert to electronic media, but as permutations of the tablet computer intrude on eBook territory. Analysts looking at the Apple iPad had earlier concluded its color screen and multimedia playback capability could cannibalize eBooks ─ especially as the tablets' form factors morphed toward 7-inch screens. Strap hanging commuters, for example, could hold this kind of reader in one hand on the bus or train to work. But this does not portend a shrinking of the iMX market. iMX-based tablets, using the Android operating system, are starting to appear in Asia.

Two trends will affect the future of the eBook and Amazon promises to play a role in both cases. On one hand, the pricing on eBooks (as expected) is plummeting. Amazon reduced its pricing down to an accessible $139 before Christmas, in an effort to stimulate sales. By all estimates they were successful. One rumor has Amazon moving as many as 3 million Kindles just in the month of December. Wall Street analysts had projected that Amazon will have sold between 4 and 5 million Kindles in 2010, but that number could be as high as 8 million. But it looks as if 2011 will be an even better year for the Kindle.

The other effort, announced just before Christmas, enables Kindle books to be displayed on other platforms, such as portable computers (and tablets), and iPhones. This totally reduces dependence of the media on an eBook hardware, which many analysts believe will turn into a giveaway in any case. The business case for eBooks starts to look like cell phones, in which you pay for the subscription, while the service provider offers the hardware for free.

I can certainly see myself reading books from my Blackberry in an airport lounge or on a standing-room-only commute downtown. I’ve actually become quite ecumenical about my book consuming habits. As one example, I’ve downloaded some non-fiction titles from Audible.com. Randall Lane’s expose on the financial excesses of the last decade "The Zeroes," and Douglas Frantz’ accusation of America’s somnambulance with regard to Pakistan’s theft-and-resale of nuclear weapons secrets, “The Nuclear Jihadist.” Though the stories are disturbing, only a single earpiece is necessary to keep up with the narration, which keeps me company as I prepare a gourmet meal for friends, or cycle weekends through Marin. []

Asia’s best kept secret

It’s been 14 years since I last visited Seoul, a place I like very much, but I admit to being increasingly troubled by what’s going on (or not going on) on the Korean peninsula: Specifically, the mounting tensions between North and South. There are some 23 million North Koreans, living on a per capita income of something like $735 per year, many hungry and close to starvation, and likely cowed by their military government into believing they under siege. The mentality of North Korea’s government is probably closer to World War I than World War II: They see themselves in the trenches against Americans and South Koreans, who, any minute now could decide to go “over the top” and come streaming across the DMZ. As the shelling of Yeonpyeong Island proved, they are not averse to slinging artillery shells across the border at South Korean targets, just to remind us they’re there, waiting. One commentator suggested that the shelling was meant to distract the North Korean people from their loathing of Kim Jong Eun, but that doesn’t make them any less dangerous: We don’t know yet whether these folks have any self-control with regard to the types warheads they put on these shells.

North Korea’s nuclear armament capability is a lot more substantial that anyone would have admitted, prior to Professor Hecker’s visit, considering their roads and infrastructure are reported to be in a horrible state of disrepair. The purpose of showing off their uranium enrichment capability was not so much as to intimidate the West (North Korea had already successfully detonated a nuclear weapon in October of 2006), but rather to advertise to nuclear wannabes (like Burma) that they have stuff to sell. When you combine this obvious weapons development with the hostile, belligerent language the North uses toward the west, and repeated assaults on South Korean targets, you may have justification for a strategic strike against the North’s nuclear facilities, something similar to what the Israelis mounted against Syria in 2007 (and decades earlier against Iraq). But this led National Public Radio commentators to suggest this would bring on a nasty retaliation against the South, possibly a leveling of Seoul with more than two million casualties.

This is damn scary, I think, but how close does it come to the things we’re worrying about right now? How many Americans, on the down side of the economic spectrum, obsessed with unemployment and job growth, will empathize with two million South Korean nationals? Does this sound callous or accusatory? Let me bring the situation a little closer — at least, for my colleagues in Silicon Valley: Some 50% of the world’s DRAM production is within a sloppy missile toss across the DMZ. (I said a sloppy toss, not a precision one.) Never mind nuclear warheads: The North Koreans, with the push a button, could severely cripple the worldwide electronics industry for many years to come.

The making of an electronics juggernaut

Apart from a 300mm facility in Austin, Samsung Electronics manufactures the bulk of its memory chips on the South Korean peninsula. The consumer electronics juggernaut, with roughly $121 billion in revenues in 2009, is the world’s largest TV supplier, a supplier of its own TFT-LCD displays, and among the first to promote and use white LEDs as LCD TV backlights. And, by-the-way, it is the world’s largest semiconductor memory supplier. Its semiconductor business, which includes graphics processors and cell phone chip sets, grew almost 60% to $28.3 billion in 2010, according to Gartner’s preliminary market share estimates. The company was one of the few to retain capital equipment investments during the recession — investments which will enable the company to retain leadership in DRAM and NAND Flash production — even as the fabrication technology moves below 40nm.

I'm not sure if anyone wants to retrace the history of Samsung Electronics, how it came to be such a worldwide powerhouse. Samsung certainly benefitted from a accumulation of investment capital in big, clannish chaebols, but also from a protectionist environment which forbid foreign companies from taking holding positions in Korean concerns — or from importing their own goods into Korea without prohibitive tariffs or hefty technology concessions.

I do not fully understand the transfer and possession of intellectual property that allowed Samsung to displace the Japanese as the most formidable supplier of memory chips. We do know that the South Korean government effectively required foreign telecommunications equipment manufacturers to license semiconductor technology to Korean companies in return for access to the Korean market. We do know that the 1990’s recession in Japan seemed to encourage downsizing and to squelch off capital equipment investments in Japan, while Korean investments continued. We do know that processors and memory are the most capital-intensive parts of the semiconductor business — requiring many billions of dollars in reinvestment each year — and the number of manufacturers capable of keeping up with the geometry shrink can now be counted on the fingers of one hand. But Samsung was already on its way to a leadership position in memory in the late 80s, thanks to a 1983 technology transfer from Micron in the U.S. and Sharp in Japan. Samsung became the first Korean manufacturer to produce 64-Kbit DRAM chips. It later partnered with other Korean manufacturers, brought together by the government-sponsored Electronics and Telecommunications Research Institute, to produce a 1-MBit DRAM (and later, in 1989, a 4-MBit device).

A Korean National Tragedy

At the time I traveled in South Korea, in the mid-1990s, Korean manufacturers were so married to memory production that the local press was treating the fall of memory prices as a Korean National Tragedy. Daewoo and Lucky Goldstar (the predecessor to LG Electronics), were still in the semiconductor business — at least, they were using EDA software tools to design ICs. I was the technical program organizer for what was then called the EDA&T (Electronic Design Automation and Test) Conferences, a traveling conference and exhibition put on the Asian Sources Media Group. Over dinner at an elegant hotel restaurant, I chaired a focus group for engineers and project chiefs — EDA tool users — at Samsung, Daewoo and LG. I was probing them to see, first, how familiar they were with the publication I worked for (“Asian Electronics Engineer”) and, next, what would it take to get these authorities to submit papers for the conference, and encourage their colleagues to turn out to hear them.

Apart from me and the Canadian who set up the dinner, everyone at the table —ostensibly unrelated and meeting each other for the first time — was named “Mr. Park.” Park, it turns out, is common name in Korea. (The country’s long-term president, the one credited with the development of South Korea’s industrial development and export philosophy from 1961 to 1979, was Park Chung-hee.) Other ubiquitous Korean names include Kim and Lee, as common in Korea as Smith, Jones and Williams in the States. I would address my focus group participants individually with their company affiliations — “Mr. Park from Samsung” — and each time they’d chuckle at my bewilderment.

Global Sources' (then called Asian Sources) promotional handle for American and European semiconductor and components suppliers was “we’ll show you how to market in Asia (particularly in China).” The formula was simple: advertise in our publications, present papers at our conferences, buy space at our exhibitions, and we guarantee you an exposure to Asian system builders who may become consumers for your products. In addition to providing `editorial support for “Asian Electronics Engineer” and “Electronic News for China” (later: “EETimes Asia”), I would recruit speakers from American and European semiconductor makers to give talks at Asian Sources’ conferences. There was an IC applications conference in China (the IIC, which still runs today), and a multimedia conference that would run in Taiwan. The EDA&T Conference, which I chaired several years, would open in Beijing, shift to Seoul and finish in Taiwan, two days in each location.

As semiconductor companies became more sophisticated about marketing in Asia, setting up their own offices in Hong Kong and Shenzhen, my recruiting job became complicated, as I had to work, for the first time, across widely divergent time zones. Where I once called California companies from New Jersey to recruit technical speakers, I was instructed to call Asia, 12 hours removed. It was the overseas sale offices who were increasingly empowered to make decisions about what conferences and trade shows to support. But initially, before those offices sprang up, my recruitment job was easier. Companies, in fact, proud of their participation in IIC or EDA&T would ask me for advice about how to word their press releases. “Should we refer to EDA&T ‘in South Korea’?” one press agent asked, “or is it ‘in Korea’?”

I had to think out loud on that one: “The United States government does not recognize North Korea,” I suggested. “Therefore it would appropriate to say ‘the Republic of Korea’.”

 

 

Reporter’s Notebook: The European Outlook

London in November was like summertime in San Francisco: Mostly overcast, often foggy, drizzly and cold. I didn't have a sunny day in the week-and-a-half I was there. Gartner's Egham office, outside of London, was a short walk from my hotel, but I had to cross a busy roundabout to get there, and on foggy mornings, I worried that a speeding car, invisible with its headlights turned off, would suddenly pop out from the curtain of fog and turn me to road kill. Never mind the warnings, painted on the asphalt street surface, reminding dummies like me to "Look Right" when crossing.

The Runnymede-on-Thames, where I was staying, overlooked the river, little more than a creek at that point. Gartner coworkers jested that the river would overflow its banks in the spring, generating dampness, mildew and bugs on the hotel grounds. The Gartner office complex (the company's European headquarters) was a surprisingly modern atrium, considering it lives on the outskirts of a dreary factory town southwest of Heathrow. It was close enough to the airport to see airplanes (when the fog lifted) hanging almost motionless in the sky (imagine some bowdlerized version of Gershwin's "Rhapsody in Blue" going through your head). Nobody would equate Egham with Mountain View or even Freemont, but the Blackberry folks, Research in Motion, and Future Electronics have sizable complexes there.

I was in Egham lending what I assumed would be applause and moral support to the presenters at Gartner's European semiconductor briefing (though not actually presenting myself). I had spent the pevious week at the Electronica trade show in Munich. As it turned out, I took more calls from the U.S. while in Egham than I did from European clients, though I had had a busy week before at Electronica. I wound up taking day trips by train into London's Waterloo Station, as easy as taking a commuter train into Manhattan from New Jersey (about 35 minutes in either case), though not the same as staying in the City.

My time in the U.K. might be considered a "London injection," something of a temporary fix against a better-than-vague possibility of actually moving there. My interest in living in the U.K. goes beyond living among the Brits, whose customs, mannerisms and accents on the language never fail to amaze me (see for example, Sarah Lyall's "The Anglo Files: A Field Guide to The British"). I have an interest in the European members of the semiconductor industry, and, thinking about the possibility of using me as an "ambassador" to those companies, my management is considering my request to transfer.

 

When the world stopped dead

Despite a heavy rain the day I landed, I did see sun in Germany, a warm autumn glow with the sun hanging low in the sky. The mood at Electronica, like the weather, was crisp, sunny and upbeat, with semiconductor suppliers congratulating themselves for bouncing back after a difficult recession. The American companies gave interesting and informative press conferences and presentations, but it was the same sort of briefings you would see from companies like Fairchild or Freescale at "analyst days" in San Jose. Texas Instruments, whose sometimes colorful marketing executives were parked at same hotel as me, promoted the manufacturing muscle it built and strengthened in 2010 - a message I had heard before. It was the European companies, harder to get close to in the States - Infineon, STMicroelectronics, NXP and Austriamicrosystems - that interested me the most in that context.

Infineon, finally out of the DRAM business, staged a spectacular comeback in 2010, concentrating on strengths in analog, power management and discretes. My own recently-published preliminary market share analysis, shows Infineon's revenues growing almost 43% in one year - more than 10% better than the semiconductor industry as a whole. Infineon's CEO Peter Bauer claimed recently that, contrary to the slowing of revenue growth we're seeing among semiconductor suppliers, his company's orders continue to pour in, mostly from automotive industry suppliers, with no end in sight. "Infineon has the ear of companies like Audi AG and China's Chery Inc. that are looking to replace more expensive mechanical parts," Bloomberg reported.

But Bauer was verbally attacked by one his own customers during a CEO panel on Electronica's opening day. The panelists, Peter Clarke reported in EE Times, seemed to feel confident at weathering a two-year storm when a systems vendor, later identified as an engineer from Robert Bosch, shook her fist at Bauer and accused him of letting the customers down with faulty allocation schemes. The other CEOs - including STMicroelectronic's Carlo Bozotti, NXP's Rick Clemmer, and Freescale's Henri Richard (filling in for Rich Beyer) rallied to Bauer's defense, as if an attack on one was an attack on all of them. "You can't blame us for failing to match the speed of the recovery," Carlo Bozotti reminded the audience then, and later, in a mid-week reception. "With formerly solid institutions like Lehman Brothers slipping into bankruptcy, it seemed is if the world stopped dead." There was no buffer, no inventory to draw from when demand suddenly exploded.

This will make you blind

My copy editor at Gartner was thrown by a statistic that appeared in one of my reports on data center expansion: “Internet traffic,” I wrote, “accounted for a total of 100 terabytes of data in 1998, and in 2009 accounted for 160 terabytes…”

“Is that per year?” she asked, failing to register what was on the page. It was quite unbelievable.

 “No,” replied, “that’s 160 terabytes per second.” When you think about it, it’s a 31 million time increase in the amount of data accessed from the internet. How many terabytes is that per year?

BBC radio has made gestures toward previously unimaginably-large numbers. The entire rendering of the Avatar film (with its multiple layers of 3D content) took up 1.2 Petabytes to store, the BBC claimed.

1.2Petabyte = 1 228.8Terabyte = 1 258 291.2Gigabyte

This is the equivalent of 500 hard drives of 2 Terabytes each or 32 years of MP3 content; that is, if you streamed all the digital content through an MP3 player, it would take you 32 years to get through it. The movie was shot and rendered in a dedicated data center in Miramar, New Zealand, (the same place where King Kong and Lord of the Rings was rendered).

Information Management reported that the computing core – 34 racks, each with four chassis of 32 machines – consisted of roughly 40,000 processors and 104 terabytes of DRAM. The blades read and write to 3 Petabytes of a Fiber Channel disk network area storage.

This is not BBC's first commentary on immensely large data networks. In 2009 it launched a content distribution system called “iPlayer.” During peak loading times, the iPlayer content distribution network passes 7.5 Gbits/s ─ which amounts to 7 Petabytes-worth of data every day.

The numbers are hardly record-breaking, the BBC admits. Hulu, the video website backed by major studios, streamed 17 Petabytes worth of video in 2008 while Youtube achieved 31 Petabytes over the same period. By 2012, Cisco VNI estimates that the monthly bandwidth consumed by online commercial videos will reach around 5 Exabytes (that's 5000 Petabytes).

Exabytes? I swear I heard the BBC reference a “Jetabyte,” presumably the next order-of-magnitude beyond a Petabyte, but I can find no reference to it with Google on the web.

Even smaller numbers make you crazy

When I first joined Gartner from EE Times, I had nightmares I was drowning in numbers. I told Colin Johnson once that I worried “I had died and gone to hell.”  I had some facility with Excel spreadsheets, when I came over, but I found myself looking at the sort complex forecasts models (with dependencies) that could make the average person go blind. Mark Skillings, then with Analog Devices, questioned some of my forecast assumptions in a face-to-face meeting we had one time in Norwood. I was sitting across from him and turned my laptop so he could see what was on my screen. He studied it for a couple of minutes, then rubbed his eyes and moaned.

As an electronics industry reporter, I had a lot of exposure to many companies’ products and technology. But as a market analyst, I’m really looking at what I call “the seedy side of the industry.” I’m trying to define how much money is there ─ how much money is out there, really. As I learned earlier as a marketing manager in the semiconductor industry, what a company really makes money on is quite different from what gets them on the cover of an electronics trade magazine. I remember proposing a 12-bit digital-to-analog converter to my management at Signetics [later, Philips Semiconductor and now NXP]. I figured we could use the charge-balancing technique developed for audio D/As to build a voltage-output device. I figured (naively) we could get the part out the door in 6 or 9 months, and, for industrial use, we could get $20 for the part. I had a market research report I bought from Frost and Sullivan, showing data converter revenues going up and to the right.

“Do you really think the data converter market will increase that way?” my management protested. “You really think we can get this part out at the top of the learning curve? What happens five years from now when the ASP for your part has dropped to 38 cents? Can we still make money on it? Why don’t you take the TAM down 20%? And while you’re at it, re-do your RIO calculation.”

The truth is my data converter products never made any money for Signetics. The parts that made money were jelly beans, cheapo widgets that sold in million-piece lots. Dolby B noise cancellation devices, for example. The RS-232 line drivers and receivers, sold into computer terminals and the newly-emerging personal computer market, were the only parts that made any money for me.

I told my reporter friends I had turned into a “bean counter” at Gartner. And, sure enough, in one of my dreams I was climbing the Marin Headlands (overlooking San Francisco) on my bicycle, and opened my saddlebags when I got to the top. It didn’t surprise me that they were filled with white beans; what did surprise me was why they hadn’t changed during the ride. In the latest incarnation of my numbers dreams, I’m feeding my cats (I have two) and I’m pouring their kibble into the cells of what seems like an excel spreadsheet.

But the financial modeling has proved to be interesting, albeit complicated. In a somewhat shocking expose of the financial wheeling-and-dealing we’ve seen over the past decade on Wall Street, “The Zeros”, Randall Lane described how young fund managers would make many millions of dollars for themselves and their clients catching penny-shifts in somebody’s stock price. They were invariably 26- and 27-year-olds (young geniuses) with mathematical modeling skills taken from Harvard, Stanford, and Wharton School of Business. The nasty stuff that Lane describes is not just their ability to profit from penny-shifts, but also their ability to manipulate clients into making the kind of expensive purchases that would produce just those shifts.

In my own models, I see major perturbations in the market when the price of a power transistor goes up-or-down 1/10^th of a cent. The volumes are so incredibly large: some 48 billion power transistors will be shipped this year (more than half of them MOSFETs), and 109 billion small signal devices. Together, a 1/10^th of a cent price drop can bring down the transistor market by over $150 million.

Semi revenues are small, even with large numbers

I’m seeing similar sensitivities for LED bulbs in lighting applications (though not for the LED drivers). The LED fixture makers are claiming the devices have higher energy efficiency than current lighting technologies, have 25 or even 50,000 hours of useful life, and provide few toxic substances (like Mercury) when you go to dispose of them. Claims for LED lighting, like Samsung’s assertion that its LED TVs save energy 40%, or cost savings for the Toyota Prius, may not look that great on paper.  

LEDs certainly offer energy savings in street lamps and halogen down lights, and, coupled to government stimulus incentives, we’ll see rapid deployment of LED street lamps in a number of cities (including San Jose). But the lighting efficacy of LED lighting (if the truth be told) is roughly equivalent to fluorescent lighting, when you take into account the efficiency of the dc-dc and ac-dc converters required to power the LED arrays and strips. But the price tag (dependent on non-existent LED manufacturing-and-test efficiencies) is considerably higher for LED lighting. You will get some swapping of (say) T8 fluorescent tubes by T8 tubes (using as many as 300 low-power white LEDs per tube), but at the volumes required to replace legacy tubes in commercial offices and factories, you don’t get a significant uptake until 2017 when the cost of a T8 tube with LEDs drops approaches $5. The semiconductor content will be about 75 cents. (The current price of a fluorescent T8 tube is about $3.50, and it will be considerable less by the time we get to 2017.)

Home lighting – incandescent bulb replacement – though equally if not more price-sensitive, will actually not reflect a replacement pattern that keys on price. The average consumer goes to Home Depot and sees a sign: “LED light bulbs. Normally $24.95. On sale, $19.95”. And the consumer thinks, “Oh: LED lights. They're supposed to be energy efficient. $20? I can afford that.” So he buys ONE (and maybe struggles with the “infant mortality” associated with the new Asian-manufactured product… “20 dollars for this?!”). But in a couple of months, he may be ready for another Home Depot sale. Thus, the LED bulb revenues (not the LEDs) could exceed $1 billion by 2013 when the price of an LED-based screw-in bulb drops below $10. But the consumption curve shoots upward when the bulb price drops below $5. (Remember: LED light bulbs will forever be chasing compact fluorescents ─ which, if Philips has anything to say about it, will have a long life ─ down the price curve.) But, think about this large number: by 2020, even with a limited penetration, both the commercial and home lighting markets will consume roughly 4 trillion low power LEDs. [ ]

Coming to a 7-Eleven near You…

In Japan last summer, hardly a day went by without some announcement on electric car production appearing in the press. “Here look,” my Japanese hosts would beckon toward a newspaper article they were reading, as if I could read the Kanji symbols. “Toyota is forming a battery partnership with Panasonic,” Yamaji-san would translate. “Mitsubishi is investing in IGBT production.” “Nissan is in a development deal with EnerDel.” And so forth, every day I was there.

 

The Japanese liked my forecast numbers for power transistors, but they thought my forecast for electric car production — roughly 600,000 vehicles in 2014 — was the pits. How could I be so blind to the assault that was mounting? they wondered. Electric Vehicle (EV) production in my forecast was gated, frankly, by the cost of the cars. Way too exensive. If General Motors was intent on selling their anemic Chevy Volt for $40,000, there will not be a lot of these vehicles on the road by 2014. For that kind of money, you can buy a Mercedes. The 600,000 EV forecast I had assumes only one percent of all the cars going out that year are plug-in electric vehicles, and despite all the furfural around the Chevy Volt and the Nissan Leaf, the analysts I’m reading say battery-powered electric vehicles will only account for 10% of car production in 2020.

 

The culprit is the battery, quite a bit smaller and lighter than the 3000-pound lead acid monstrosity that powered the EV1, but still big, expensive, and not nearly as potent as a tank full of gas. As battery maker A123 pointed out at last year’s Applied Power Electronics Conference (APEC), a single gallon of gasoline contains roughly 36 kilowatt hours (kWh) of useful energy — albeit modulated by a power train that’s only 20% efficient. The battery for the Nissan Leaf, in contrast, contains only 24 kWh of useful energy. And if batteries cost $1000 a kilowatt hour, speculated Greentech Media, the battery alone would cost $24,000.

 

Much of electric vehicle development has focused on increasing battery capacity and lowering its cost. There is some progress to report: At $400 a kilowatt hour, Greentech Media estimates, the battery pack for the Nissan Leaf would only cost $9,600. The newer batteries are multi-cell lithium ion (Li-Ion) with as many as 200 to 1600 separate cells strung together in serial and parallel strings. Lithium-ion charge controllers — what many power management IC suppliers call “safety circuits” — are used to balance the charging levels of each lithium-ion cell. Under certain conditions, lithium batteries can be incendiary. Computer battery packs have been known to catch fire. Thus, the charging levels on the automotive batteries must be carefully controlled. While it is possible to speed up charging by using higher voltages, it is important to avoid overheating the battery cells. The process will never be as fast as (say) putting gas in your car at a filling station.

 

On my way to Japan last summer, I read about a shift in the thinking process regarding electric car batteries. Instead of a 300-mile battery that takes 12 hours to charge, a consensus was growing around a smaller battery that could be charged in 45 minutes, supported by a network of charging stations. As the Japanese told me last summer, anticipate at least one charging station in the parking lot of every 7-Eleven store.

 

 

GM’s great comeback

 

The Chevy Volt did indeed seem like an oversized go-cart in the mockup that was presented at the Freescale Technology Forum in June. The body was removed so that everyone could see the T-shaped battery which ran, like the transmission chain of an ancient rear-wheel drive car, down the center hump of the vehicle, and across the underside of the rear passengers’ seat. Freescale’s head of sales and marketing, Henri Richard, was filmed taking a test drive in the vehicle, which looked stylish and modestly trendy on film, and reported that the vehicle was quiet, peppy and fun to drive. But to the casual observer, this vehicle looked like it would utilize an awfully large battery just to go 40 miles.

 

Well known by now, thanks to press coverage at the Plug-In 2010 Conference, the Volt embodies an entirely different philosophy than the Nissan Leaf. Though it will seat five passengers, the Volt is intended to be a short-distance commuter car: You drive it to 20 miles to work, drive it home 20 miles home, plug it into the electric outlet in your garage to recharge. If you’re worried running out of juice on a 35 mile trip, the Volt just happens to have a gasoline engine, which takes over when the battery has discharged. If you go as far as the battery charge will take you, and then figure out how much gas you’ve used to get you home (or to the next charging station), you’ll get something like 150 miles per gallon.

 

But the need to carry what GM refers to as “dual power supplies” makes this electrical vehicle unnecessary expensive. What Obama has termed "GM’s Great Comeback" looks like a terrible miscalculation to me.

 

Nissan Leaf, in contrast, has a lithium-polymer battery that goes for 100 miles. The battery can be charged to within 80% of its capacity within 30 minutes, Nissan claims. It’s also promoted as a five-passenger sedan, and sells for $32,800.

 

To my mind, this too misses the mark. If you’re going to spend this kind of money, why not opt for a Lexus, a Mercedes, or (if you’re environmentally conscious) a Toyota Prius? One recommendation is to hold out for one of the $49,900 sedans announced by Tesla Motors at this year’s APEC. The Model S, scheduled for production in 2012, will do 0-60 MPH in 5.7 seconds, but will consume 900 amps in the process. Its lithium-ion battery has a highway driving capacity of 300 miles… but it takes at least 45 minutes to charge.

 

Imagine if you will the surge on the electrical power grid when you plug these cars in at night: the load for 3 cars on the power line can exceed that of an entire home. A neighborhood with many electric vehicles (picture Atherton, California), all charging at the same time, can overwhelm a local transformer. There are two alternatives to electrical overload: Either expand the available power (not likely to occur), or somehow control charging to avoid demand peaks (difficult, but doable). Attention is shifting, consequently, away from the electric cars, and onto the charging infrastructure needed to support them.

 

 

Charge your electric vehicle from your cell phone

 

Both the Chevy Volt and the Nissan Leaf will have IP addresses and cellular modems, Qualcomm’s Ignacio Contreras told me at their recent analysts’ meeting. While your electric car is charging in the garage, you can call it up on your phone, and tell it, on a hot day, to start your air conditioning. You don’t want to burn battery power cooling off the passenger cabin, Contreras said; you want to do that while it’s connected to the Grid. The cellular connectivity, Qualcomm believes, would similarly allow you to monitor your car battery’s charge level.

 

Qualcomm is partnering with ECOtality, a developer of internet-based software for electric vehicle recharging, The software platform will allow drivers to remotely check the charge status of an electric vehicle at any ECOtality charge station. The software will allow electric vehicle owners to program the recharging time in advance to avoid peak energy charges, and will send text or email reminders when recharging is needed. It will also help drivers locate the nearest charging station when they are on the road.

 

ECOtality’s EV project, supported by the Department of Energy with American Recovery and Reinvestment Act (ARRA) funding, which plans to install 14,650 220Vcharging stations in public and commercial sites across six states (Arizona, California, Oregon, Tennessee, Texas and Washington). ECOtality claims this would support the 5,700 Nissan Leafs and the 2,600 Chevy Volts likely sold next year.

 

Coulomb Technologies’ ChargePoint Network, similarly, is attempting to proliferate a network of 110 and 220V charging stations — roughly 5000 in 2011 — with $15 million in DOE stimulus funding.

 

Better Place not just charging the battery, but swapping it out. The demo on the Better Place web site shows a battery being replaced on an automated cradle in about a minute and 20 seconds. The customers leave their depleted batteries with charging station, and pay only for the charge on the fresh one.

 

Better Place has already $350 million in equity financing from an investor consortium led by HSBC. A good part of the money will be used to stock on those big, expensive batteries. q

Separating reality and dreams

On balance, I found the recently-released sci-fi thriller, “Inception,” satisfying and thought provoking. The promotion for this billed it as “James Bond meets ‘The Matrix’,” and the film maker (Christopher Nolan) certainly delivers. The plot follows a team of corporate super agents, led by Cobb (Leonardo DiCaprio) as they storm a highly armed fortress, implant something of a potentially world-changing nature inside, and try to escape before the whole thing blows up. Described this way, you have a conventional Hollywood bang-bang-shoot-‘em-up caper movie. (The 12-year old in me can’t wait to see the forthcoming Sylvester Stallone, Arnold Schwarzenegger, Jet Li fight extravaganza, The Expendables.” Guaranteed: it uses the exact same formula.)

What makes “Inception” so different and interesting is that the armed fortress the team is trying to penetrate is not a physical thing; it is the deepest chambers of a person’s mind (Cillian Murphy), identified as the world’s 2^nd richest man. And the world-changing thing the team is trying to implant is an idea. The team “goes in” while the victim is sleeping, sedated in the first class cabin of a 747 on a 15-hour flight from Singapore to the US. From the word “go,” the corporate agents are dealing with the sometimes bizarre landscapes, characters and emotions that can only populate a dream world.

 

I have criticisms, to be sure, and understand why the film had such mixed reviews. There’s a lot of unnecessary exposition at the beginning, trying to tell us that the architecture and time lines for a dreamscape are likely to be different from what we think of as “reality.” We know this, without being told. The inner landscapes of the DiCaprio character, in fact, and the young “architect” he hires (Ellen Page), are 1000 times more interesting than the dreamscapes of the man they’re trying to penetrate. The architect imagines entire cities folded in half (wonderful special effects) and the DiCaprio character is tormented by images of his wife, unfortunately deceased, who nonetheless appears repeatedly in his dreams as a villainess and saboteur.

 

The dreamscape of the world’s 2^nd richest man, in contrast, is actually quite Hollywood conventional. It’s populated by “security” personnel, with pistols, machine guns, hand grenades, SUVs and armed fortresses. Thus, penetration of this dream world is an endless series of fistfights, gun battles, car crashes, and explosions. Sure: they make you squirm in your seat, as Hollywood fight scenes can make you flinch and squirm. But half the time, you can’t tell who’s punching whom, who are the good guys, who are the bad guys, who’s winning, or why we should care. But you do loose track, with all this shooting kicking and punching you’ve witnessed, all these bizarre-yet-familiar landscapes and cities you’ve visited, the heart-wrenching romance you’ve experienced, the deaths you’ve witnessed or been a party to, that this mayhem has gone on entirely inside someone’s head. And what a jolt it is — a mental jolt — to awaken in the first class cabin of a 747, and hear the voice of a stewardess offering a hot towel, a refresher for the descent into San Francisco.

 

 

While humanity sleeps

 

I had actually written about the idea of reality as a collective illusion years ago in a Planet Analog column, about "The Matrix" film series. The Matrix is a full-blown science fiction epic. Set in the distant future, it portrays a war between intelligent machines and the surviving humans. The humans are put to sleep while their energy is harvested. A software program is injected into people’s brains, a program which makes the sleeping human think he (she) is living a real life. The collective consciousness suggested by this programming is called the Matrix. (The only way I got away with posting a film review in the context of a technology column was by asking my readers to click their heels together three times and repeat after me: “The Laws of Physics.” Get ready.)

 

 

“Inception” is actually more of a “concept film” than science fiction. It doesn’t show us (say) a dystopian future. It simply takes a feature of the world as we know it, and sets it on its ear. In this case, it’s the stuff that remains within the inner spaces of our minds, and how we access this in dreams. We live in our own dreams, sometimes vividly. But, can we live — fully conscious — in somebody else’s dream? Cobb and his wife (Marion Cotillard) — the real love interest in this movie — actually shared dream states. They became so intertwined in each other’s dreams, each other’s cityscapes and emotions, that it became difficult to discern which was a dream and which was a conscious state. Cobb, the DiCaprio character, began to utilize handheld tokens whose behavior — their responses to the laws of physics —would cue him as to whether he was awake or dreaming. His wife refused such references, and would return repeatedly, even after she’s dead, imploring him to join her in the dream that they shared. Though he knows she’s a figment of his subconscious, he can’t let go of her. He still loves her, still longs for her, and that’s what gives her so much power.

 

 

Some movie critics pointed to this subplot as a distraction, but I found this infinitely more thought-provoking than the caper theme. It started me thinking about the boundaries between conscious waking states (what we think we’re seeing and experiencing in “reality”) and what happens in dreams, and how one state of consciousness invades another.

 

 

A World Yet to Come…

 

I can’t tell you very about the thinking process of suicide bombers or Islamic terrorists… except… there seems to be a denial or rejection of THIS world, in favor of The World Yet to Come. In Salman Rushdie’s novel (a suspenseful page-turner), “Shalimar the Clown” is a Taliban-trained assassin. This world is not real, they tell him; it is a dream state. His job is to clear the space for the world that is to be. The thinking process encourages a lot of bloody sacrifice, and Shalimar does his work without fear or remorse.

 

 

I’m wondering what similar thinking exists (say) in Jewish philosophy. At my father’s funeral, I read paragraphs from the Isaac Bashevis Singer story, “Gimple The Fool” which my father had liked. Gimple appeared for much of his life a gullible soul. As a youngster, his classmates on the shtetl gave him goat turds to eat, telling him they were raisins. As a young man, the “charming young virgin” he was convinced to marry was a lame toothless hag, and highly promiscuous as well. Yet, she took tender care of Gimple. He came to love her and missed her terribly when she died. Thus, there was a schism between what “the Fool” thought and felt, and what the rest of the world perceived. In the end, as his health began to fail, he looked forward to joining his wife in the afterworld. He shed his earthly belongings, including his clothes. Piss on THIS world, he said (literally).

 

 

I think my father liked the “piss on you” part (he was not the least bit religious) but I was still wrestling with the life-as-a-dream issue when he died. I loved my father, and, though he was perhaps my most severe critic in life, I’m sure he loved me. I have a vivid memory of him, one night after we moved from a housing project in Manhattan to a small suburban home on the south shore of Long Island. It was just after WWII when veterans returned from Europe and housing was scarce. My father got a GI loan which enabled him to purchase the house (for less money, in fact, than I paid for my last two cars). I was 4 or 5 at the time. I must have been nervous and agitated about the move; I remember my father talking to me, spending time with me, calming me by his presence. When we first moved, my younger brother and I shared a bedroom, downstairs next to our parents’ bedroom. Later, our beds would get moved to the upstairs bedroom. But, initially, this proximity to our parents probably had a calming effect. On the night I remember, my father came into our room. He gave us some chocolate and read us a story.

 

I remember falling asleep after that feeling very happy. I don’t remember what happened the next day. In fact, I don’t remember waking up. Thus, the seed of an idea was implanted in my head, an idea that was to become more vivid as my childhood and early adolescence became psychologically difficult. The idea was that I was still asleep in that first floor bedroom. That what I was experiencing was a bad dream. And that any moment now I would wake up. My father would be there. And I would be happy.

 

It is this very thread — this lingering doubt about the reality of what we are seeing —that “Inception” so successfully pulls. Even after coming awake in the cabin of an international flight, passing customs in San Francisco, you’re never quite sure whether the protagonist has returned to “reality” or entered another dream world. You’ve been there: The customs agent examines your passport, gives it a stamp and hands it back to you. “Welcome home,” he says. Why does that feel so ominous? Is it overcast outside? Why is the customs hall so dark? The scene is so vividly portrayed, that you, as the movie viewer, have a difficult time — right up to very last frame of the film —discerning if what you’re seeing is “real,” or just part of somebody’s dream. q

Google Your Body with Your Phone

Hugh Herr, from MIT’s Media Lab, gave one of the more inspiring keynotes I’ve heard in a long time at Freescale’s Technology Forum (FTF) in Orlando last month. There were roughly 4000 people in the audience. Dr Herr, equipped with a lavaliere microphone, and dressed in a golf shirt and slacks (totally appropriate for the mid-June Florida humidity) paced back-and-forth across the stage as he spoke. He seemed a bit nervous on video close-ups, but there was nothing unusual or distracting in his gait. Consequently, it was a bit of a shock when he revealed — five minutes into his talk — that he had no legs.

 

He lost his legs 20 years ago, Herr told us, from a mountain-climbing expedition on the back side of Mt. Washington in New Hampshire. Mt. Washington is only about 8,000 feet high [a high heart-rate challenge for cyclists in the summertime], but a bit short compared to California’s Mt. Shasta, Washington’s Rainier, or Alaska’s McKinley. As the highest peak on the eastern seaboard of the U.S., however, Mt. Washington has absolutely no windbreaks between itself and the Arctic. Winter winds howl through at 100 MPH and temperatures have been known to drop below -45 degrees. It was in this environment that Herr and a friend were stranded for four days in deep powdery snow. At the point at which they were ready to give up living, in Herr’s telling, he and his friend were discovered by a lone woman on snow shoes, and transported by helicopter to a prominent medical center, where doctors labored in vain for days to save their severely frost-bitten limbs. Herr lost both his legs below the knee.

 

The story is significant, because it becomes a prolog to Herr’s life work, for which he was named one of the world’s top 10 inventors in healthcare by “Time” magazine in 2004. Both at MIT and at his own company, iWalk, Herr specializes in biomechatronics, a discipline (not unlike robotics) which uses machinery and electronics to replicate the fine motor activity of human limbs and torsos. He has, for example, modeled walking, running and standing for various clients (including Olympic-class runners and runway fashion models). He can tell you how much pressure is put on the heel or the balls of your feet, which way the toe points or flexes climbing steps, and what your calf muscles or knee joint is doing to facilitate that. He says that your legs are 2.4% of your body weight, and signals to them from your brain have a 50 millisecond delay coming down your spinal cord.

 

Herr’s own artificial legs (he rolled up his pants legs to show us), were not simple mechanical prostheses; they were intelligent robots with carefully-weighted joints and musculature, and six degrees of freedom. Next to a motor (and hydraulics) at the angle joint, there were 12 environmental sensors, including (wait for the commercial) several Freescale 3-axis semiconductor gyros, and 5 PIC microcontrollers. The controllers, Herr remarked, had their own IP addresses ─ someday accessible by Google. Thus, the fine motor activity of his legs, Herr claimed, could be tuned with a cell phone.

 

 

Minority Report

A number of MEMS tracking services have noted the increasing use of flow and pressure sensors in medical electronics, devices implanted inside the body, transmitting data signals wirelessly, enabling doctors to see, through new-generation monitors outside the body, what’s going on within. I don’t think anyone has projected the revenue potential for micro-electromechanical systems in prosthetics or robotic body extensions. Gartner’s own analysts were in fact pretty sanguine about the future gesture recognition in the question-and-answer session that followed our semi-annual semiconductor briefing in San Jose last month.

 

Because of our focus on semiconductor devices (micro-machines etched in silicon), there’s chunks of the total MEMS market we undoubtedly miss. We don’t count ink jet nozzles, for example, though we do count the driver ICs (as application-specific analog). We don’t formally track digital light projectors (DLP), for that matter, a shrinking business in any case. Thus, our MEMS forecast numbers are likely to be much smaller than our competitors' in market research. But, what we do track — accelerometers and gyros — are pretty well-honed with market share compilations and forecast projections. (We can even tell you which companies are buying the parts.)

 

We are, for instance, seeing enormous unit and revenue growth for silicon motion sensors in cell phones. Rotational sensors will distinguish between portrait and landscape viewing modes for the LCD screen, and 3-axis gyros will serve as footstep counters supplementing GPS receivers for location-based services (e.g., where’s the nearest Starbucks) in handsets and portable navigation devices.

 

No stranger to the trend, Freescale used the Technology Forum (its own mini trade show) to introduce a family of 3-axis gyros called “Xtrinsic.” The Xtrinsic line will marry silicon motion sensors with signal conditioning circuits and microcontrollers in the same package. The idea is to bring on-site intelligence to the motion sensing function. It is a necessary step for transforming simple motion sensing to full blown gesture recognition.

 

 

The Future of Gesture Recognition

Remember Tom Cruise swatting images around on an otherwise transparent glass template in the Steven Spielberg collaboration, “Minority Report?” A stage actor used the same motif, pulling on a tight-fitting glove and using it to swat video images around the stage at the opening of the Freescale Technology Forum. Despite the science fiction imagery, the sequencing of video screens was triggered by motion sensors in a real commercial device. The Gesture Glove, developed by AnthroTronix is used to transform sign language into computer-readable text.

 

STMicroelectronics, demonstrated their own 3-axis gyro during a financial analysts’ meeting in London in June. Taking a page out of the Wii, STMicroelectronics’ video showed a game of racket ball, with players swatting an imaginary ball back and forth on a projected screen — using their cell phones as paddles. More useful was a demonstration of handwriting and signature recognition in which the sensor would follow the swath of your hand or finger motions, recognize the points at which you change direction (or make an underline) and matching that with your digitized signature. It was the same principle, by the way, as used in the handwriting recognition software (“graffiti”) on the Palm Pilot.

 

Oddly, the subject of gesture recognition never came up during the MEMS Industry Panel I chaired at the GlobalPress Summit in April. Kionix's Eric Eisenhut mentioned the subject to me in a series of phone calls we had before the panel, but the subject had barely a mention during our panel discussion. Eric worried how to turn hand gestures—like swipes or upheld palms—into meaningful statements of intent (or emotion). This requires not just a lot of intelligence, but a lot of programming with “if, then” loops.

 

I think we (myself as well as the panelists) were still dazzled by a keynote presentation from Nokia at the MEMS Executive Congress in the fall of 2008. Nokia’s Vision of the future cell phone went beyond the combination phone, PDA, email processor and Internet browser we have today. The phone would have a malleable form factor, and the ability to morph. We needn’t stuff it in our pockets or wear it on our belts. We could, for example, make a bracelet or piece of jewelry out of it; wear it on our arms or around our necks. (As far-fetched as this sounds, Jabil Circuit in China, one of the world’s largest contract manufacturers, is quietly acquiring companies with design and material expertise that would allow it build electronics with malleable form factors.) But where the phone becomes something you wear, responding with intelligence to movements and gestures, it can be not just an extension of your body, but an intelligent part of it. q