The smart home market has flattered to deceive. Despite promises to simplify our lives and countless Kickstarter campaigns most homes remain decidedly un-smart.
That’s because many of these technologies still aren’t that smart themselves – they’re difficult to set up, often unreliable and can take up more time in troubleshooting than if you just turned up your thermostat yourself.
Adoption will remain low until these technologies become a seamless and unobtrusive part of our lives, which means making our interactions with them as natural as possible. That’s why ABI Research believes voice control will become “the bedrock for smart home applications”. In a recent report the business intelligence firm predicted that 120 million voice-enabled devices will ship annually by 2021.
The combination of speech recognition and natural language processing is a powerful one. Being able to say “It’s date night” and watching the lights dim, the gas fire light itself and Barry White come on the Hi-Fi rather than fiddling around on your smartphone has obvious appeal.
Amazon is undeniably leading the field with its cloud-based Alexa platform. The company released the API for the voice-control technology at the heart of its Echo wireless speaker last year and it has already seen widespread adoption – it now integrates with 1,000 different third-party services. Most notably from the smart home perspective this includes US home improvement giant Lowes’ Iris smart home system.
Google and Apple are hot on their heels though. The former is due to unveil its voice-controlled Google Home smart hub later this year, with all the weight of the company’s experience in natural language processing behind it. Apple also recently opened up its Siri digital assistant to developers, which combined with the forthcoming release of the company’s Home app in iOS 10 will allow users to control compatible smart home devices via voice through their phone.
While this shows the confidence in the technology it also highlights a major challenge. As ABI notes, a coherent multi-device smart home system needs a shared and standardised voice platform. At present the big players are focussed on creating their own ecosystem with little thought to interoperability, which presents a headache for smaller companies hoping to build smart home devices.
Whether and when to build voice control into devices will also depend largely on how sophisticated they are. As ABI’s research director Jonathan Collins notes, integrating voice control into smart home systems “breeds complexity” so for smaller players it may not be worth the hassle until the technology is well established.
Privacy concerns may also make many consumers wary about purchasing devices that listen in to what we say. Most solutions use “wake words” to tell a device that a command is about to be issued, but big technology companies’ poor track record when it comes to siphoning off user data are likely to give some pause for thought. Smart home devices are also notoriously insecure – a dedicated hacker could well use your voice-enabled fridge to monitor your conversations.
And as anyone who has used Siri, Google Now or Cortana will know – voice control is still far from error free. These errors could be far more off putting if they end up resetting your house’s heating system rather than just initiating the wrong Internet search.
But natural language processing is a fast moving field and these systems are getting smarter every day. As they improve it will be possible to issue complex commands in a single step that would that would require several minutes of scrolling between apps on traditional interfaces. That may be the magic bullet that finally makes smart homes smart and helps the sector live up to its expectations.