Everything at your command: Voice-controlled computing

|     Thomas Schoerner     |

Berlin (dpa) – As powerful and necessary to planning everyday life as they’ve become, most smartphones and PCs still have one failing: Usually, data needs to be input with a keyboard, not via voice command.

By and large, voice control remains a dream. But there is movement: It’s standard in most smartphones and tablets. Even some game consoles and navigation devices and PC programs can take oral commands.

“Even TVs are available with voice control, but only the top-end models,” says Michael Schidlack of Bitkom, a German technology industry association.

There are even fields where voice control is becoming standard, like assistance devices for the disabled.

Apple’s iPhone and iPad are controlled with a voice recognition system called Siri. Google and Microsoft have introduced similar options.

The features apply to a lot of functions: Internet searches, managing diaries, setting alarms, opening apps, navigating, checking scores, accessing mail or contacts or just telephoning.

“Even if the operating system doesn’t offer the service, there are enough apps by now that have their own integrated voice recognition,” says Moritz Stueckler of German technology magazine t3n.

In some ways, Apple has the most advanced software.

“Siri can reserve a table in a restaurant, so long as the establishment accepts online reservations,” says Dennis Steimels of the German magazine PC Welt.

With Android, some manufacturers are imposing their own interfaces on top of the original system. That’s how Samsung’s Galaxy 3 came up with the voice recognition function S-Voice.

“You can even take pictures,” says Steimels.

Cloud computing systems are one of the reasons voice recognition has become so much better.

“Instead of conducting the relatively complex analysis of the language in the comparatively low-power device, it’s taking place in the much more powerful server system,” says Stueckler.

Of course, this has the drawback that the devices must be online for the voice recognition to work.

Voice recognition has also expanded to non-mobile devices. Windows users have been able to dictate texts and even simple commands for opening programs or copying and pasting documents since the Vista system was introduced. There are similar functions with Apple’s OS X.

Cars are also getting into the act, with ever more functions becoming voice-controlled. It’s not just about making it easier to handle the car, it’s also supposed to make driving safer.

“The driver doesn’t have to take his hands off the steering wheel and can focus on the road,” says Stueckler, while the computer focuses on making phone calls or picking new routes.

But today’s voice recognition is not 100-per-cent reliable, says Steimels.

“The software doesn’t always understand what’s said, or it misunderstands completely and suddenly wants to navigate you into a foreign country, even if we just wanted to go to the closest bakery.”

Background noises like loud-speaker announcements in public areas are hard to filter out and can distort the spoken order. The systems also have problems with long or complex sentences.

A lot of advances would be possible if the software could recognise its user’s voice.

“Apple and Google are going in this direction, whereby the program gets to know the user and reacts to his needs and preferences,” says Steimels.

Non-native English speakers having trouble with voice recognition might want to switch to English.

“At least in the mobile segment, the most work and resources is going into the analysis of the English language,” says Stueckler, noting the large market for English and the language’s relatively simple sentence structure.