Time Is on Your Side: Wearables Spell Opportunity for Speech Interface Designers

Smart Looks for Smart Watches

Bill Meisel, president of TMA Associates and AVIOS executive director, was impressed by the look of Apple's smart watch. "It's much more attractive than other watches that we've seen. They've worked a lot on the user interface and tried to make the small screen more usable."

Apple watches retail from $349 up to $10,000. Of course, you don't have to spend $10,000 on a smart watch made of 18-karat gold with a sapphire crystal to get gorgeous design and voice recognition—in this case, Siri. Other smart watch providers, such as Samsung, LG, Motorola, and Pebble, are also coming out with eye-catching designs. Some manufacturers are on second- and third-generation products that no longer look clunky if not downright ugly. As more companies up their design game, it is becoming increasingly possible to have fashion without sacrificing functionality.

Still, all things considered, many people consider smart watches an expensive luxury item, not a must-have. "Prices for smart watches are high, and they will generally come down," Llamas says. "We'll see high-end smart watches that will buoy prices so that the decrease won't be like falling off the edge of a cliff. Compared to analog and digital watches, smart watches cost more, but price variations should be considered in light of the additional value that they bring, such as health and fitness, third-party applications, notifications from the smartphone, and more."

Under the Hood: Speech Recognition

Speech experts agree that natural language understanding (NLU) is the best form of speech technology, that is, speaking normally without sounding stilted or artificial and expecting that the solution can understand the intent of what's being said.

"It's late enough in the game that the end user has expectations that they can use words that they're comfortable with," says Dan Miller, founder and senior analyst at Opus Research. "That calls for natural language understanding. Implicit in natural language [are] the many ways somebody might say things."

Scholz echoes those thoughts. "Natural language is a primary opportunity. It takes the input from your speech and helps to identify actions and objects within what you just said," Scholz explains. "It's also extremely important to pay attention to context."

For instance, a speaker using natural language might say, "Turn the corner." If stated to a GPS system, it would be clear that "turn the corner" is meant as a navigational command. But if, for instance, you were talking about woodworking, "turn the corner" would mean something along the lines of filing an edge of a corner.

"Context becomes absolutely vital in facilitating understanding," Scholz says. "What this means for the designer is that a lot more effort has to be put into both preliminary natural language processing, identifying these actions and objects, and in maintenance of context."

Related to context is managing the focus of the specific conversation or dialogue. "The focus is something you have to recognize from the context and … that has to be altered as the conversation moves forward," Scholz says.

If you're using an application for online banking, initially, the focus might be identifying the name of the account. The focus might then change to how much money is in that account. As the conversation goes forward, the focus needs to change.

Natural speech and use of smart watches go hand in hand, Llamas says.

"Smart watch user interfaces are going to change to more natural [interactions] going forward," he says. "Nobody wants to peck, swipe, and gesture all day. What I would really like to do is lift my wrist to my mouth and say a command and it does it for me, none of this scrolling. But it's going to take time to get there."