Aug 1, 2001 12:00 PM,

Let Voice XML Speak

BY NIKKI SWARTZ, ASSISTANT DATA EDITOR

Voice XML 2.0 will be released this fall. Version 1.0 hasn't exactly been a hit. Even so, many believe voice XML, the programming language for interactive voice-response services, is integral to the future success of 3G applications, the wireless Internet and creating the revenue WAP enthusiasts have been hyping for years.

Speaking Voice XML

The VoiceXML Forum, created by Motorola (www.motorola.com), AT&T (www.att.com), Lucent Technologies (www.lucent.com) and IBM (www.ibm.com), completed version 1.0 of the standard in March 2000. Voice XML evolved from an older markup language, VoxML, originally developed by Motorola, which considered existing markup languages ill-suited for developing voice dialogues.

The voice XML language, based on the World Wide Web Consortium's (www.w3.org) XML standard, follows the syntactic rules of XML with semantics that support the creation of interactive speech applications.

Voice XML's architecture resembles traditional Web-based technologies, as well as wireless technologies such as WAP. Voice data is sent over the network to a voice XML gateway designed to translate voice into the voice XML language. Once translated, voice XML data is sent via HTTP to a content server. The gateway also can respond to the user from the voice XML page using text-to-speech and recorded voice.

For wireless devices to use voice XML, there must be a voice XML plug-in that works with a micro-browser, as well as a telephony-based connection to a voice XML gateway.

“At the heart of voice XML is a standard interface, and with that you can have portability of applications,” said Bill Ory, Motorola senior manager, voice XML business development. “If you write an application in voice XML, you should be able to port it to different platforms. As we evolve, it gives a standard framework of how people should code applications.”

Version 2.0, to be released this fall, is an evolution that will fill in and define the framework laid out by 1.0, adding more extensions for more capabilities. The VoiceXML Forum said adoption will simplify creation and delivery of Web-based, personalized interactive voice-response services; enable voice access to integrated call-center databases, information and services on Web sites and company intranets; and enable new voice-capable devices.

Today, most commercially deployed voice applications are not voice XML compliant, but voice XML 2.0 will spur greater adoption. Most carriers at least are looking at it as a viable technology. Many, including Sprint PCS, will launch voice XML-based applications soon.

“Sometimes we move more quickly than the standard. We designed Voice Command two years ago. Now as a standard evolves, we're going to evolve with it,” explained Kevin Packingham, Sprint PCS senior manager, product design and usability. “Part of that is just making sure we can deploy the quality of services we've already deployed. Voice XML will support what we're doing now; it's just going to take time to make sure the product is transitioned properly.”

Dennis Paschke, Sprint PCS vice president of marketing, product development, said voice XML adoption will help carriers expand their voice-content portfolios.

“We need a standards-based approach to ensure that all the developers out there are working to a common standard,” he said.

Talking Multimodal

Many industry insiders say the key to easier user interfacing is multi-modal capability, and voice XML is the foundation.

Multimodal interfacing, or “tap and talk” technology, allows different types of interactions, including keypad entry and speech input. Instead of being limited by today's WAP-only or voice-only sessions, multimodal interfacing will allow users to talk to their WAP phones and get information read to them while data pops up on the device. (See “Multimodal Headway”.)

“We view multimodal capability as a huge opportunity in the simultaneous voice and data environment,” said Paschke. “We get a much more interactive experience where processing on both ends can happen, rather than just requesting information and processing it one way, as dialing and command are now.”

Today's networks don't support voice and data channels being open simultaneously, but 3G will solve that problem. That capability, Jackson said, will improve the wireless Internet's stock.

“There are always going to be some things you'll prefer to do in a voice-only application, but there will be some things that you prefer to do in both WAP and voice,” he explained. “It'll just make the applications more usable, which will lead to people using the wireless Web more often.”

According to Daniel Miller, The Kelsey Group senior vice president (www.kelseygroup.com), carriers must anticipate not only multimodal but also mode-indifferent technology.

“We have a key enabling technology that does make the whole more valuable,” he said. “Voice XML, XML itself and the next generation of open instant messaging are the three big enablers that will simplify some of these complex things.”

Voicing Wireless Web Concerns

According to Jackson, the economy has slowed voice XML adoption.

“The focus we see out there in the market is on ‘how can I use voice/voice XML to replace call-center costs?’ rather than ‘how can I use this technology to make new revenue through advertising?’” he said. “Companies will start to invest in these types of cost-saving applications, but after the economy moves upward again, they'll start to realize other ways that the technology can be used to make money.”

Miller said voice XML applications would make it easier for Web-based applications to migrate to voice-oriented networks and drive minutes.

“The other opportunity is if the true voice portal gets deployed at the carrier level, then there's an opportunity to sell a premium tier of service built around robust, voice-based applications,” he said.

You don't need a 3G network to offer compelling voice XML-based applications, but voice XML may be getting lost in the rush to 3G.

“There's been such a huge focus on WAP that even voice was kind of pushed aside,” Ory said. “Voice XML is a here-and-now way to at least prototype some of these early (voice and multimodal) applications and see what the uptake is.”

Paschke agreed.

“With 3G, we'll be able to give a much richer experience through voice interaction, and so it's our challenge to develop those kinds of products that can leverage that technology advantage,” he said.

And voice XML may prove more profitable than WAP.

“People find it more natural to use voice. So there's greater potential,” Jackson said. “It could be even more popular if it was possible to offer multimodal applications. And we'll probably get there within the next 12 to 18 months.”

Multimodal Headway

Multimodal technology combines voice, text and graphics but has been limited to voice moving in one direction only while text or graphics travel the opposite way. Auvo Technologies (www.auvo.com) has developed a platform that provides true simultaneous integration of voice, text and graphics, enabling them to move together in either direction over packet-data networks.

“A lot of the promise around 3G networks has been based on simultaneous voice-and-data transmission down the same pipe,” said Grace Jenkins, Auvo CEO, in a press release. “We have the technology now to deliver on much of that promise well ahead of schedule — in a 2.5G environment.”

Auvo's technology allows multimodal access to information located anywhere on the Internet. A subscriber can use voice and/or text input to make menu selections, navigate and retrieve information. Auvo's technology leverages a direct IP connection and can be used with any 2.5G or 3G packet-based wireless network.

The technology doesn't require a voice-channel interface. Instead, handset software converts speech into data packets that can be sent instantly over the Internet to a host server or directly to other users. This allows for simultaneous use of data and speech over a data-capable network.

In August, Auvo and Vodafone's AirTel Móvil S.A. (www.airtel.es) will conduct the first live commercial network trial of the multimodal platform in Madrid. Using AirTel's GPRS network, the trial will demonstrate simultaneous, 2-way transmissions of multimodal applications.