Google has released a fresh version of the Google Assistant SDK in order to allow hardware makers to install and empower the Google Assistnat in a wider range of hardware. This week's updated SDK tagrets companies large and small and ensures everyone will have a better experience with their smart devices.
To start, Google Assistant is available in more languages and in more countries. The voice-powered tool can be put to use in English (Australia, Canada, UK, and the US), French (Canada and France), German, and Japanese. This should allow hardware makers to target a broader range of markets around globe. Google says developers can now programmatically configure the API to handle these new markets.
Google Assistant can now control a wider range of hardware. Many of Google Assistant's finer points can be managed cia the Settings screen within Google Assisntant app on their smartphone. Not eveyone who has a Google Assistant-powered device will have access to the Google Assistant app. Moving forward, devices that take advantage of the latest SDK will let owners support the same level of customization -- including language, location, and device name -- directly on the device. For example, device location can be handled manually via a smartphone or automatically via the API using approximate longitude and latitude. The idea here is to make it simpler for people to obtain relevant results to queries such as "Ok Google, what's today's weather?"
The new SDK bakes in support for text-based queries. Sometimes it isn't appropriate or possible to speak voice-based queries. Google says this particular update was based directly on developer input, so it added alternate -- in this case, text-based -- input and output mechanisms. Moving foward, the voice response API includes the ability to handle typed-in text when needed.
Google says the improved SDK introduces the ability to build Actions directly into Assistant-enabled, SDK-based devives. This means developers will be able to specify what traits the device actually supports (after all, not all devices can support all traits of Assistant). For example, many first ask "how can I ask the Assistant to control my device?" The new SDK is able to take user instructions ("Ok Google, set the temperature to 78 degrees") and turn the into structured intents via cloud-based automated speech recognition. Developers will need to insert the client-side code for actually fulfilling those actions, but the SDK handles everything else.
Last, device management. Google says a new management API baked into the SDK will help developers register and manage their SDK-based hardware. The REST API makes it simple for devs to register, unregister, and view all their devices, as well as collate devices of the same type and catalog their individual characteristics.
A per usual, Google has a lot of new documentation to help developers navigate the additional features in the Google Assistant SDK. Those interested in creating commercial products are encouraged to reach out to Google directly.