Alexa Auto

Amazon Alexa Gets First Local Control in New Auto SDK

Amazon has rolled out the new Alexa Auto SDK 2.0 this week. The new version of the system is the first example of Alexa offering local control for the voice assistant without needing to connect to the cloud.

Alexa as Car Co-Pilot

Until now, connecting and conversing with Alexa in a car required a persistent internet connection, whether through the car or a smartphone. Now, certain Alexa skills and abilities can be accessed and used even if the car is in a tunnel or an area with poor internet service. This means music, navigation, phone calls, and other controls can be run through voice commands regardless of the car’s location.

To make this possible, the new Auto SDK integrates Alexa’s Automatic Speech Recognition (ASR) and Natural Language Understanding (NLU) into the car’s systems which previously was only accessible through the cloud.

“Since its release last summer, we’ve seen an incredible response from automakers interested in using the Alexa Auto SDK to embed Alexa directly into their vehicles,” Ned Curic, vice president of Alexa Auto at Amazon said in a statement. “At the same time, we know drivers often find themselves in places with poor or limited connectivity, and they still want to be able to access certain features and services through Alexa. This newest version of the Alexa Auto SDK delivers that capability for automakers, enabling their customers to take Alexa even more places on the go.”

Amazon has been working on making this shift for a while. Voicebot was first to report that offline Alexa was coming to its automative offering based on an interview at CES 2018. Reports then emerged last year that the company was working on ways to compress Alexa in large part because of the potential to integrate the assistant into automobiles. Amazon executives have said they don’t plan to limit the offline Alexa use to cars, but it’s a useful proving ground for the technology, and necessary for automotive use cases.

From the Cloud to the Edge

Moving audio processing away from the cloud offers plenty of benefits for voice tech developers beyond the automobile. Compressing and slimming down the size of the software can improve efficiency decrease latency while lowering the cost of development on the platform. There is also a strong privacy benefit as anything carried out locally by the software is not transmitted to cloud servers and thus is completely untouchable by other listeners.

Those benefits have been the source of success for several companies. For instance, Picovoice offers voice-to-text transcriptions without using the cloud, while Snips has created a whole voice platform to operate locally, and just this month partnered with NXP Semiconductors to incorporate their technology into NXP’s microprocessor kits. As Amazon, Google, and others work to integrate their voice assistants into cars, having the system run locally could be crucial to appealing to carmakers and consumers.


Amazon Alexa Machine Learning Team Develops New Approach to Offline Voice Assistant Usage

Snips Partners with NXP for New Voice Control Platform 

Picovoice Launches Speech-to-Text Platform ‘On the Edge’