Talk to the Hand(held)
A new standard aims to voice-enable your Palm Pilot.
Editor’s note 8/06/02: Two corrections are noted at the end of this story.
When New York State banned using cell phone handsets while driving, motorists turned to hands-free phones-including those with voice-activated dialing. Unfortunately, there’s no voice-activated surfing to make a Palm Pilot as safe when checking directions at 65 miles per hour-at least not yet.
Speech-enabled handheld computers, personal digital assistants (PDAs), and other similar devices should enter the marketplace in the next year or two if more than 50 vendors such as Philips Electronics N.V., Microsoft Corp., and Intel Corp. are correct. Their killer technology isn’t a new processor, operating system, or application. Instead, it’s a new standard that can voice-enable existing Web applications. But first they need to convince developers to revise applications using Speech Application Language Tags (SALT), pieces of code programmers can use industry-wide to implement new features into Web-based applications.
Up until now, the specifications for such code have been pretty limited. The existing standard, called Voice extensible markup language (VoiceXML), was originally designed for speech-only applications running over a telephone, says Rob Kassel, SpeechWorks International, Inc.’s product manager for emerging technologies and a proponent of the SALT standard. It was perfect for helping customers dial up to ask about their bank balances. It didn’t take into consideration the need for a Web browser.1
SALT proponents say the new standards will enable input and feedback in various formats and interfaces, which they call multi-modal. Say you’re on a business trip walking down an unfamiliar city street with your PDA, for example. “You can say, show me my current location,’” says Kassel. A SALT-enabled PDA with a global positioning system (GPS), a wireless Internet link, and the right application would quickly accomplish a number of things. First, through natural language processing, it would know to retrieve your GPS coordinates and forward them to a mapping Web site such as Mapquest or Yahoo! Maps. The PDA’s browser would then display a map of your location. Next, ask the PDA where your next meeting is and hear it recite the conference room number over the PDA’s speaker or through an earpiece. You can either dictate a new meeting for your calendar or write it in with a stylus.
SALT supporters like Kassel say handheld computers are just one application: The technology will work equally well for cell phones or PCs. Albert Kooiman, director of business development at Philips Speech Processing, imagines a SALT set top box that makes TV surfing much easier. “You’re wiped out. You can sit through the scroll of listings for 98 channels or you could just say: Are there any movies going on?’” he says.
Snags with Tags
But SALT’s backers must overcome several hurdles before channel surfing becomes a form of conversation. For vendors like Philips to sell SALT-enabled hardware, there have to be SALT applications. And that invites the question: With no SALT-enabled PDAs available at present, why should Web developers re-write existing apps?
“There’s a chicken and an egg situation,” admits Kassel. Still, he feels that corporate developers will support SALT because the applications will work on today’s PCs. A SALT application that works fine over a wired network needs no modifications for the wireless SALT-enabled PDAs of the future. Kooiman estimates that a person with a basic understanding of Web-based development could prototype a SALT application in two to three days. Adding SALT to an existing application might take as little as an hour.
Recent steps by software vendors indicate that progress is imminent. Microsoft has integrated a speech software developer kit into Visual Studio .NET. “Speech development becomes drag and drop,” says James Mastan, director of marketing for .NET speech technologies at Microsoft.2
The platforms to use these applications are also coming into place. Philips has a SALT-enabled browser ready. Carnegie Mellon University is preparing an open source version. Microsoft expects to release a SALT-enabled version of Pocket IE sometime in the first half of 2003.
To mollify some critics that SALT has been developed without public input, the SALT Forum, an organizing body founded in October, 2001 and who completed the SALT 1.0 specification on July 15, is submitting the specification to a standards body.
Still, proponents have wildly different ideas of which SALT application will be most useful. The varied answers may indicate a standard with a broad appeal. Or it may mean that the industry has yet to find the single killer app that will get everyone talking to their PDAs.
- An earlier version of this article incorrectly identified a standard for voice-enabled telephone interfaces as XML. The standard is VoiceXML.
- An earlier version of this article incorrectly stated that Microsoft integrated a beta version of the speech SDK into Business Studio. According to a Microsoft spokesperson, the speech SDK is integrated into Microsoft’s Visual Studio .NET.
Couldn't make it to EmTech Next to meet experts in AI, Robotics and the Economy?Go behind the scenes and check out our video