When speech recognition technology (SRT) was first touted as the 'next big thing', people tapped into their sci-fi fantasies and envisioned a future of domestic bliss, where a quick word to the washing machine would see their clothes clean in a flash.
Then reality kicked in, with early SRT offerings proving to be inaccurate, expensive, and difficult to use. Examples such as Telstra's infamous directory assistance have also done little to endear the technology to users.
However, SRT seems about to emerge from its ugly duckling phase, with some big names backing its potential. Gordon Moore, of Moore's Law fame, believes that "speech recognition will change computing as we know it". Meanwhile, Bill Gates is also a big fan of SRT, having said that, "speech is not just the future of Windows, but the future of computing itself".
Researchers back these predictions, with The Kelsey Group estimating that the voice services market will be worth $US41 billion globally by 2005, while the Datacomm Research Company foresees more than two billion people using voice systems by then.
In the US, consumers are encountering speech interfaces with increasing regularity. However, it will take around two to four years before we see a similar occurrence locally, according to Peter Chidiac, Australia and New Zealand manager for SRT group SpeechWorks.
One area that will be revolutionised by the introduction of SRT is the call automation market. At present, the market is dominated by "touch tone" technologies; however, Chidiac predicts that this will change significantly over the coming years.
"It is expected by many research firms that by the year 2005 touch-tone will be regarded as a thing of the past, with most new shipments of call automation systems being with a speech interface rather than a touch-tone interface," he said.
At present, the finance, banking and tourism sectors are dominating the take-up of SRT, however Chidiac notes that transport, telecommunications and Web portal companies are also starting to make a lot of noise in the SRT space right now.
One of the challenges that SRT will have to face is shaking off the legacy of its earlier reputation. However, Chidiac is confident that this will not be a problem, as more and more systems are implemented correctly and return on investment is proven. Consumer acceptance of the technology will also ensure that SRT fulfils its potential, Chidiac said.
"Speech systems can eliminate the menu structure of touch-tone as a person can say what they want and therefore get straight to it," he said, adding that speech systems can also be given a "personality", because callers are interacting in a conversational manner.
After implementing a full, natural-language telephone banking system at Credit Union Australia, Chidiac noted that the system was so successful that it had customers actually thanking it before hanging up.
"The bottom line is that speech systems can speed up the call, offer a friendlier and easy-to-use interface (no longer do we have to look at the keypad on our phone) and therefore provide a high standard of customer service where call automation is offered," he said.