“By aiming a hand-held device like this at what’s around you, you’ll learn what local services and businesses exist in the direction you’re pointing at,” explains Pradhan. “You’ll be able to preprogram your device to accept only websigns in those categories you’re interested in. You might want websigns indicating only local real estate for sale, for instance, or maybe just the nearest available parking spot.”
As Pradhan rotates from north to south, the list on the Jornada’s screen changes appropriately. Then, as he faces in the opposite direction to the one in which he began, more new names appear, such as Darbar Indian Cuisine … again? Quickly, Pradhan confers with Mehrban Jam, who’s overseen much of this project’s hardware development. Jam, a gentle and patient man, confirms Pradhan’s hypothesis about a glitch causing the Journada to misinterpret the GPS signals that define its position relative to the Indian restaurant two miles north.
HP Labs’ websigns project is supposed to “hyperlink” physical locations to Web data for ordinary users on the move — and do it by riding piggyback on existing infrastructures. Thus, a Palm or any other PDA can be modified to support websigns as easily as Pradhan’s Jornada; discussions have begun with Nokia about re-engineering ordinary cell phones. However, the technical problems for Pradhan, Jam and their team at HP Labs haven’t been insignificant.
They’ve solved their project’s essential difficulties nevertheless, and they may have accomplished a very good and useful trick. We’re often promised that in about two decades’ time, we’ll have a world of pervasive computing, with everybody and everything connected: Computational capability will be embedded throughout the environment; local networks will track users’ locations through wearable computers or wireless mobile tags; and users’ personal information — their virtual OS, files and multimedia applications — will migrate onto local systems, to be summoned by their voice commands as needed. It’s a pretty picture (if you’re not one of those folks terrified by the potential for loss of privacy or the threat of never being free from work), and not implausible. Except that nobody has figured out the first step for getting to this world of pervasive computing.
Or, at least, they hadn’t before now. Now Pradhan and Jam’s HP Labs team may have managed that first step with websigns.
“Large infrastructures for pervasive computing and computing systems physically embedded in the environment will be hard to justify,” Pradhan says. “What’s the business model? For websigns, on the other hand, the hardware requirements are minimal. For a positioning technology, GPS was the obvious choice and has started showing up in mobile phones anyway. For determining where users are pointing their handheld devices, the compass component is a magnetometer, which is cheap enough that it shouldn’t be a cost issue.”
So how do websigns work?
When a user first enters a general physical locality, websign client software on the user’s handheld device would access local websign servers and transmit her GPS coordinates. Then, as the user moves around, his or her mobile phone or PDA will display relevant websigns based on current location, the direction in which he or she might be aiming the device, and filtering and preference mechanisms attached to his or her personal profile.
While websign-enabled handheld devices would seemingly act as though they were sensing real electronic transmissions broadcast from specific physical structures or sites, they would actually be doing no such thing. Rather, they’d be filtering through the “websign cache” of an area where the user was – let’s assume it’s San Francisco’s Union Square, as in the illustration below – and displaying only those websigns that are in the user’s vicinity and have services relevant to her needs. websigns are, in essence, simply URLs assigned to local structures or sites.
A business could establish a websign very simply. Since there’d be no actual transmission from a business and its data would be stored in a database for use by a local websign server whenever users’ websign-enabled devices came into range, the business would initially just need to supply its own zip code via a Web-based interface designed by Pradhan’s HP team. Once the business designated this general location, an onscreen map would enable it to pinpoint its exact position. The system would convert this to latitude and longitude and assign a URL. And that’s all it would take.
“Websigns won’t be economically difficult to justify,” Pradhan says. “Especially for travel-related businesses.”
Pradhan says websigns began as a solution to the problem of how to extend an earlier project called beacons – physical boxes that broadcast URLs to mobile devices indoors – to an outdoors environment. Both projects are part of a larger HP Labs research initiative called “cooltown.” Though the HP Labs project hasn’t achieved the more widespread recognition of its counterpart at Xerox PARC, cooltown seems just as ambitious: a credible program to realize the vision of a world of pervasive computing through the Web as it exists today.
“We’ve other interesting projects in the skunk works stage,” Pradhan says. “I can’t say much, but there’s a watch we’re doing with Swatch which will be a mobile device. Also, as beacons provided a starting point for websigns, I’m working on a new project called gossip. With this, a user’s device could receive websigns for objects in a physical vicinity from other users’ devices – basically, a peer-based model for filtering websigns. Still, it’s early to be talking about this.”
