Your smart life is about to get even smarter with a new set of software development tools that will let coders include world-class speech recognition and natural language processing — the same stuff that powers Siri, Apple’s personal digital assistant — to thermostats, refrigerators, apps and, yes, even robots.
The folks at Nuance have created a new system, currently in beta, to allow any company to include code with language commands that are specific to their hardware or apps. It’s called Nuance Mix, and anyone can sign in and create their own speech-recognition code to work with their apps or connected devices.
“Any developer, big or small, can come in and define a custom set of use cases,” Nuance’s Kenn Harper told Cult of Mac during a demo of the SDK. “You’re going to start talking to everything at home and work — speech is about to get more ubiquitous.”
As Nuance’s mobile director of product management, Harper’s excited about the new crop of devices coming to our homes. We’re seeing more connected thermostats, fridges, home control devices, music streaming apps and speakers, robots, entertainment, virtual reality and wearables come to market, said Harper. Most of these devices won’t even have a screen, and they’ll all need a user interface.
“For the first time,” he said, “we might see voice becoming the user interface of choice.”
When you sign up for Nuance Mix beta as a developer, you get access to a robust set of coding tools that allow you to build a set of speech recognition with natural language parsing technology that applies to your own specific hardware and software requirements. A thermostat needs a vastly different set of commands than a refrigerator does, and a home robot needs something even more completely different.
It will, of course, need to be highly accurate. Siri’s famous for misunderstanding our spoken commands; you can’t have this with a personal care robot, thermostat or even wearable device.
“To get a high level of accuracy,” said Harper, “we’re creating the natural language piece, but also creating a fully customized speech model for voice recognition. We optimize speech and natural language together.”
Ultimately, what this all means for consumers is that we’ll talk more to our devices than ever before, just like we’re doing with our smartphones, which use Siri, Cortana or Google to process our requests. Imagine telling your fridge to text you whenever the temperature gets too warm, or asking your robot to dial 911 if your elderly grandmother slips and falls. The most natural way to ask for complex interactions is with our voice, and Nuance is the best-in-class speech recognition around. It makes sense they’ll power our talky future.
Developers can sign up for Nuance Mix’s beta program now and see how it all works.
I, for one, can’t wait to call out and say, “Hey, oven. Preheat yourself to 350 degrees and order my favorite take-and-bake.”