The Rise and Rise of Voice Control

home automation voice control

Computer, locate Mr Spock

Whether it was Captain Kirk talking to the Enterprise’s computer, or Blake speaking to Zen and Orac; it’s not so long ago that the ability for electronic devices to understand the spoken word was purely in the realms of science fiction. Early attempts at dictation software often resulted in spending more time correcting mistaken words than it would have taken to just have typed it in the first place.

Big Data is the Game Changer

So how is it that in the last few years voice-activated devices like Alexa, Siri, Cortana and Google can suddenly understand us so much better? With millions of people all with their different accents, dialects and pronunciations; all correcting their devices when they misunderstand (usually by repeating the phrase) has allowed the servers that process the speech to become much more accurate at interpreting all the nuances of the different ways we speak. This volume of ‘training data’, combined with devices that now have microphone ‘arrays’ that can pick up sound wherever you are in the room—even in a relatively noisy environment—have seen the move from gimmick to a genuinely useful man-machine interface.

Initially the likes of Alexa were considered just information providers, offering (upon spoken request) news bulletins, weather forecasts and sports results. But interface them into a home control and automation platform such as Control4 or Savant Pro and they instantly become an extremely easy way of controlling your home. Turn on lights, change the heating temperature or choosing your favourite TV channel can all be achieved without leaving the sofa or reaching for a remote control; simply speak your command.

How will Voice Control improve Home Automation?

Currently you have to use specific words and phrases to achieve your desired tasks. Macro routines can be pre-programmed to string together a sequence of tasks triggered by a word or phrase but these generally have to have been considered in advance and pre-programmed into the control system. However natural language understanding is improving all the time. Soon you won’t have to restrict yourself just to the exact phrases that your device knows; just speak in the same way as you would to another human and the system will correctly interpret your intent.

Contextual Understanding

Rather than just issuing single commands, voice control systems are starting to have a ‘conversation’ ability; so they can ask follow-up questions to clarify exactly what you meant or to offer further options. It will recognise different voices and therefore know what you mean by “my bedroom” for example, rather than having to say “John’s bedroom”.

Similarly, you’ll no longer need to pre-program all the different potential sequences of tasks. Simply saying, “I want to watch the latest episode of Stranger Things in the cinema room in ten minutes” and the control system will work out all the steps necessary to complete the task. Whatever room you are currently in, the control system will know it needs to be sending commands to the cinema room equipment. It will know that ‘Stranger Things’ is a show on Netflix so it will turn on the necessary AV equipment and select the right sources. It will also recognise whose voice made the request and what was the most recent episode of the show they’ve watched in order to select the right one to view. If you’re the only person in the current room, the control system could ask (via the multi-room audio‘s ceiling speakers) if you’d like to turn off the lights and TV in the current room before you go to the cinema room.

All this probably isn’t that far away; the hardware is already here so it’s just a matter of future iterations of the control system software having ever more sophisticated artificial intelligence (AI). So if you’ve already got a Cyberhomes integrated control system, make sure you take advantage of our SMARTsupport plans so your Control4, Savant or Crestron control system can be continuously upgraded with the latest voice control functionality as it becomes available.

Posted by
on 25 Feb 2019
in
Back