Downtown Tokyo is the venue for an experimental, non-Net, wireless network that puts site-specific details on your mobile device. Our reporter tries it out
There's no way to translate "ubiquitous computing" into Japanese. But that hasn't stopped Ken Sakamura from repeating (yubikitasu compyutingu) like a mantra whenever he's talking about the future of wireless communications. The University of Tokyo professor is his country's proselytizer-in-chief for a government-funded project to make cell phones and other mobile gizmos more useful.
How? By allowing them to tap into information hotspots where prepackaged audio, video, and text files are kept. That's the aim of Sakamura's Tokyo Ubiquitous Network Project. On Mar. 5, I borrowed a Ubiquitous Communicator, or portable media player, to test out the system for a couple of hours.
Sakamura's experiment might not sound too ambitious given that cell phones already do far more than voice calls and Internet surfing. In Japan, they're cutting-edge, doubling as credit cards and commuter passes, cameras, and e-mail terminals. Among the latest offerings are kids' models with built-in, satellite-based, global positioning system technology so parents can keep tabs on their offspring at all times.
Yet for all these advances, there's one thing that cell phones still are not, and that's location-aware. Have you ever stood before a landmark and tried using your cell phone's Net browser to learn more about it? I have, many times. On a visit to Asakusa's Sensoji temple, it would have helped to know that the temple, built in 645, is Tokyo's oldest, for instance. But Googling from a cell phone is often more trouble than it's worth. I end up fumbling over the keypad for a few minutes and then scrolling down through an endless list of sites before giving up. Why can't our phones instantly help us figure out what we're looking at?
Sakamura's network would do that, and without requiring Net connectivity. Handsets would simply pick up weak radio signals coming from a few meters away and process the data. That's a nifty feature because going online for information can jack up the phone bill.
Unlike traditional cellular networks that rely on antennas and towers, this system uses radio-frequency identification (RFID) chips, infrared sensors, and Wi-Fi units. For his field test in Tokyo's Ginza shopping district, Sakamura placed more than 1,200 chips, 270 infrared spotlights, and 16 Wi-Fi stations on lampposts and in flower beds, stores, and underground subway tunnels.
Only the project-issued portable media players work for now, though eventually cell phones might, too. The players are designed automatically to receive audio files offering street directions, store layout maps, and videos about the history of the district.
Doesn't Match HP's Offering
Imagine the convenience of carrying a device that tells you whether a store has a rare Bordeaux vintage in stock or a wheelchair access ramp. Or one that gives you directions while downloading digital photos or files that point out the historical landmarks in the neighborhood you're walking through. Chips could also be embedded in concrete to give the blind an electronic guidance system. "You don't have to push any buttons—it's automatic," says Sakamura, who heads the YRP Ubiquitous Networking Laboratory in Tokyo.
Hewlett Packard (HP) has a bright idea for a similar network it's developed: gaming. The tech giant has been toying with the concept of an open-source platform, called Mediascape, that lets programmers create games merging the real and virtual worlds (see BusinessWeek.com, 12/5/06, "Hewlett-Packard: The Unlikely Game Player"). Recent trials had gamers running around the Tower of London on a virtual scavenger hunt or walking the trails at Yosemite National Park. In five to seven years, HP engineers expect advances in wireless broadband hardware, RFID tags, and GPS to create a whole new online gaming experience.
Sakamura's project is also open-source, and he has some big tech names behind him. Microsoft (MSFT), IBM Japan, NTT DoCoMo (DCM), and Matsushita Electric Industrial (MC) all belong to a board overseeing the project. But the project didn't seem as advanced as HP's Mediascape, which works with the company's current lineup of iPAQ PDAs.
For now, the costs of Sakamura's network are prohibitive. RFID tags used for product tracking at factories and in supply-chain management generally cost between 1 and $5 per chip. Sakamura's chips are custom-made and their price tag can run upward of $100 apiece. A few high-powered infrared sensors cost several hundred dollars. Ideally, the tab for setting up these networks would be divvied up between local merchants and city hall.
The demonstration reinforced the fact that Sakamura has a lot of refining to do. One example: Cell phones nearby can clog up the airwaves and slow his network's functioning. "We're trying to develop a way to limit the interference," says Sakamura. Another problem is information overload. When my media player told me that the Nissan Ginza Gallery was the first car showroom in the area back in the 1970s or the stone lions seated at the entrance to Mitsukoshi were the department store's trademark, I stopped the playback. (Recently Sakamura's network got a test run at a zoo in Tokyo. That's a better use for this technology.)
And often the media player was simply too slow to catch up with my wanderings. As I was walking underground watching a video message from the Tokyo governor, the program froze and then switched to a different one featuring the Ginza subway station's chief, which also froze when I entered the field of another info hotspot. The future of wireless convergence will depend on making these gadgets easier to use.