Facebook Developing Speech Recognition for ‘Portal’ Device

Further evidence has emerged that Facebook is indeed working on its own smart speaker system, as well as an AI assistant to operate it.

The latest, and perhaps biggest, clue comes by way of code found in the Facebook and Messenger Android app by mobile researcher Jane Manchun Wong. The code enables a rudimentary speech recognition system, translating a user’s speech into text during use.

Intriguingly, it’s labelled, “Aloha Voice Testing”. About a year ago, Facebook was rumored to be working on a smart speaker/video chat device codenamed “Aloha”, so it looks like this speech recognition system could be designed to work with that piece of hardware – especially given that the code appears to feature interconnectivity with other devices via Wi-Fi or Bluetooth. But there is also an indication that the hardware device in question is now called “Portal”, with Aloha featuring a message reading, “Your mobile device is now connected to Portal.” So “Aloha” may now simply refer to the voice interaction software.

In developing its own voice interaction-based smart home system, Facebook is somewhat late to the game. Indeed, Bloomberg has reported that the company had initially planned to launch its system back in May, but delayed it when the Cambridge Analytica data mining scandal broke. But with the news cycle now having moved on, and voice interaction continuing to emerge as an increasingly central user interface in consumer devices, Facebook is evidently back at it, with a spokesperson now telling TechCrunch that the company will have some “hardware news” to share “in a few weeks”.

Sources: TechCrunch, The Verge, Bloomberg