Can This Man Work Magic?
In a conference room on the sixth floor of the vast glass-and-steel building housing Nokia’s research labs in Helsinki, Igor Curcio, an Italian-born computer scientist who specializes in signal processing, demonstrates a prototype of one of his favorite projects. Called Director’s Cut, it is an automated, crowdsourced video production service. A small group of people who go to an event—a concert, a wedding, a soccer game—can capture video on their cell phones and then upload it to the service. Its computers synchronize the videos with the help of their time stamps, analyze sharpness and exposure to get the best pictures and sound quality, and decide what to include in a final composite video by figuring out how many of the cell-phone users got pictures of the same thing.
In a demonstration video made from a live concert in Finland, quick cuts from one camera angle to another give a professional look to the three-minute show. Multiple audio tracks create stereo sound. The clip looks ready for broadcast on low-definition television.
What’s not clear is who would want such crowdsourced video, or whether Nokia could make a business out of it. Curcio, who has worked as a radio-station disc jockey, suspects that people would enjoy the opportunity to create more professional-looking videos of weddings, family reunions, concerts, high-school sports events, and other shared experiences. He also says a service that encourages people to use bandwidth uploading pictures and video is likely to be attractive to Nokia’s crucial partners: the wireless carriers that sell its phones. But it may not be possible to provide this service to the public. Nokia’s lawyers are concerned about the digital rights to audio and video from concerts, sporting events, and even weddings where music is played.
You might say that projects like Director’s Cut are all too common at the company’s labs. Nokia spent 2.9 billion euros on handset R&D last year—far more than any competitor, and twice as much as Apple spent on R&D of all kinds. But in recent years, such lavishly funded efforts have failed to produce any technology that would allow Nokia’s products to compete with the iPhone and other new smart phones. “Nokia was the king of the mountain, with the largest budget,” says Howard Anderson, a venture capitalist and a senior lecturer at MIT’s Sloan School. He says the company was “blindsided” by Apple’s iPhone, which was designed primarily as an Internet device while Nokia and others were designing phones with add-ons. Jason Armitage, a London-based mobile-media analyst at the Yankee Group, says that Nokia adopted a classic strategy for dominant players: it segmented the market and provided different products for every customer class. “It missed seeing the opportunity to invent a whole category with a single killer design,” he says.
As a result, Nokia is rapidly losing its leadership position in one of the world’s fastest-growing markets, wireless communications—a market that it helped invent. Its sales of low-end phones fell 16 percent in the second quarter of this year compared with the second quarter of 2010, and its sales of more expensive smart phones fell 34 percent for the same period. Nokia dominated the early days of the worldwide smart-phone market with phones based on the Symbian operating system, but Apple products and now Google’s Android phones have leapt ahead. “Symbian is now regarded as a dumb smart phone,” says Anderson. With its revenue dropping, Nokia’s market value has been cut in half over the past two years, wiping out more than $30 billion in shareholder wealth. That has put the future of this 150-year-old company in question.
Henry Tirri, Nokia’s new chief technology officer, may play a bigger role than anyone else in determining that future. Tirri, who previously led some 500 researchers as head of the Nokia Research Center, stepped into the job earlier this year after the company brought in a new chief executive, Stephen Elop, in hopes of figuring out a better strategy. Elop, the former head of Microsoft’s business division and the first non-Finn to head the company, outsourced all software development and support for Symbian to Accenture—transferring about 2,300 people to the consulting firm—and announced that Nokia would rely on Microsoft’s Windows Phone developers for its future smart-phone software. The company will lay off another 4,000 people next year, and in September it closed a Romanian factory and announced plans for 3,500 more layoffs.
Now Tirri is expected to “set the course for Nokia and our role in the mobile industry,” Elop said. And he doesn’t have a lot of time. As the CEO said in an interview that was webcast by Nokia, “We have to go faster and harder and more aggressively now than we’ve ever gone before.”
It is a daunting challenge for the 54-year-old Tirri, a Finnish machine-intelligence researcher who now lives in Silicon Valley. Tirri, who is also a professor at the University of Helsinki and the Helsinki University of Technology, says that as head of the research labs he “looked at future disruptions.” Now he also has to “nurture the Nokia technical community,” negotiate joint work with Microsoft, oversee the company’s intellectual-property strategy, and develop innovations for both the high-end smart-phone market and the business of making low-priced phones for Africa and Asia, the last places where the company still dominates.
Sniffing Out Opportunities
At Nokia headquarters in the Helsinki suburb of Espoo, Tirri, who jets frequently between Finland and California, has an office with a view of firs, birch trees, and an arm of the Baltic Sea. His soft-spoken demeanor is more that of an academic than a corporate executive. And Nokia remains committed to basic research, he says: “At times like this, clever companies maintain research. Research opens options.” Despite the layoffs in other parts of the company, he says, Nokia hasn’t made significant cuts in investments for its labs.
As head of research, Tirri says he views himself as a kind of venture capitalist who gets proposals from lab directors based on projects started by researchers. The lab directors are like angel investors who authorize preliminary work and then select the few ideas worthy of being assigned to research groups. Projects that will require teams of 30 or more have to hold the promise of creating businesses worth $1 billion or more for Nokia, which had $56 billion in revenue last year. Tirri says that even though business conditions have worsened for Nokia, and development—especially for software—has been cut, its research labs around the world are pushing ahead with work that may be just a few years away from yielding products.
Tirri once described his main research interest as “reasoning under uncertainty”—teaching machines to make decisions on the basis of probabilities. And he will surely need to reason under uncertainty as he negotiates the tumultuous cell-phone industry, which in a single 10-day period this past summer was upended by Google’s purchase of Motorola, Apple CEO Steve Jobs’s resignation, and Hewlett-Packard’s decision to abandon tablets and its WebOS. Tirri says that the Motorola and HP announcements made Nokia look smart for deciding to rely on Microsoft for operating-system development, because it is becoming clear that Windows will be one of only three or four main smart-phone operating systems for consumers and software developers to focus on.
Even with Microsoft producing the operating system for advanced phones, Tirri says, Nokia has plenty of opportunity to “innovate on top of the platform.” One main area of innovation is cognitive radio technology, which he believes could lead to phones with nearly unlimited bandwidth and lightning-fast response time. The idea is that even as more and more mobile devices compete for bandwidth, much of the radio spectrum remains underutilized. In many places, for example, analog television bandwidth is barely used because stations broadcast digital signals instead. In some countries, police and fire departments, and even taxi companies, have bandwidth reserved for their transmissions. Cognitive radio would make use of otherwise idle bandwidth by analyzing the environment, “dynamically asking whether space is available” on each slice of the spectrum, and deferring to a preferred user only if that user needed it at the moment, says Hannu Kauppinen, who took over from Tirri as research director. As part of a consortium of companies, Nokia has received permission from the U.K. government to use the “white space” in existing television frequencies for a trial in Cambridge, England, and Kauppinen is hopeful that broader use will be approved “if we can prove we don’t cause interference.”
Pushing the boundaries of radio research will be the only way to prevent future cellular traffic jams, Nokia believes. Kauppinen says the technologies the company’s researchers are investigating could “improve the speed of wireless communications up to 10 times.” Nokia researchers have already designed and produced an integrated circuit that can “sniff” channels to find unused spectrum. Regulatory issues block the way to putting this technology in action, however. “We still have to live in the regulatory era of 100 years ago, when one channel was allocated to one use,” he says.
Another major research area involves nanomaterials such as graphene, which Nokia researchers believe could enable the company to redesign and reshape handsets. Graphene is a super-strong, atom-thick form of carbon that is flexible, transparent, and highly conductive. Tapani Jokinen, Nokia’s head of design technology, is intrigued by its potential. He has been working on prototypes of flexible phones, using graphene for the circuitry, that could be bent, stretched, or wrapped around a forearm like a wristwatch. “It’s like a stretchable electronic skin,” he says.
Nokia has also developed technology that would use graphene in flexible haptic displays, which could deliver information through the sense of touch. One possibility is “haptic navigation”: a flexible phone handset would signal the person holding it to turn left or right by pressing against the hand in one direction or the other. Other potential uses of nanomaterials include an “electronic nose” in which silicon nanowires on a chip could identify particular molecules in the air. Incorporated into a handset, it might warn a user of allergens in restaurant food.
Tirri, meanwhile, sees potential for Nokia to create a major new business in location-based technology. The company owns the widely used Navteq mapping service, the principal alternative to Google Maps. Researchers believe that combining this with tagging technology could have a variety of useful applications.
Using low-power antennas placed on the ceiling, Nokia researchers can determine the location of an object equipped with a Bluetooth-enabled tag to an accuracy of 30 centimeters. The tags communicate with a cell phone via Bluetooth, and then the phone orients the user by means of a preloaded map. One way to use such a system would be in stores: a user would enter a shopping list into a cell phone equipped with near-field communication, which is already used for payment and other functions in some devices, and the screen would display a map showing where to find the desired products. The phone could also display special offers on products the shopper passed while navigating the store. In a home with a few locators installed, tags could be attached to keys, television remotes, eyeglasses, and other frequently misplaced objects.
Despite its struggles to compete with Apple and Google in the smart-phone market, Nokia still sells more phones than any other vendor. About 1.3 billion people around the planet own Nokia phones. Its ring tone may be the most-played musical theme ever. With its $56 billion in sales last year, it is one of the biggest technology companies in Europe, and its success is a source of national pride for many of Finland’s 5.4 million people. Indeed, Nokia, which started 150 years ago as a paper mill, helped invent the cell-phone business in 1991 by leading the development of GSM, the first mobile digital network technology. A year later, in 1992, it pioneered SMS text messaging. It built the first phone without an exterior antenna in 1998. In 2005, it sold its billionth phone, a Nokia 1100 in Nigeria. That model, introduced in 2003, may have been the best-selling consumer-electronics product ever, with about 250 million handsets sold.
Whether Nokia survives is arguably quite important, particularly in emerging markets, where it remains dominant. It is best placed to provide smart phones to many poor people who want to upgrade from less capable handsets. Its handsets support 80 languages, and it has been spectacularly successful in selling low-priced devices in emerging economies—markets that Apple isn’t close to cracking. Researchers in its Beijing labs developed a way for users to continuously write Chinese characters using a stylus, one character on top of the last, on small smart-phone touch screens; they found that the handheld devices could use various clues to more accurately recognize what the user was trying to write. “Two billion people don’t have data connections—only SMS,” Tirri says. That means touch-entry technologies for texting are crucial.
Nokia won’t challenge the iPhone with a jellyfish phone that can be wrapped around the wrist. But the company’s strong presence in many parts of the world gives it a unique opportunity to once again redefine how more than a billion people communicate.
William M. Bulkeley, a former technology writer for the Wall Street Journal, is a freelance writer based in Boston.
AI is here. Will you lead or follow?
Join us at EmTech Digital 2019.
How DARPA Took On the Twitter Bot Menace with One Hand Behind Its Back
When DARPA ran a competition to find Twitter bots designed to influence online debates, it inspired a new generation of anti-bot strategies.
How One Intelligent Machine Learned to Recognize Human Emotions
Nobody knew how to identify people’s emotional states by looking at their brain waves. Then a machine learning algorithm stepped in.
How an AI Algorithm Learned to Write Political Speeches
Political speeches are often written for politicians by trusted aides and confidantes. Could an AI algorithm do as well?