A division of the company is using this motion-detecting technology to enable non-verbal interactions.
Creating socially intelligent devices is the goal of Google’s Advanced Technology and Products (ATAP) team. She seeks to rethink the way humans interact with computers. To do this, she uses Soli, a sensing technology that uses miniature radar to understand movement at different scales. According to the division of the company, it is able to detect both heartbeats and body movements.
“We use this motion detection to allow devices to understand the social context around them”, the team said. A device would thus be able to identify a person who approaches it or enters their personal space. It would be controlled by a wave of the hand or a turn of the head.
An understanding of the modes of interaction
To enable the nuances of body language, the ATAP team also uses machine learning algorithms. This would allow it to make more accurate predictions of how and when someone wants to engage with a device. Following the observations of these algorithms, she designed a set of “interaction primitives”. The first – “approach and leave” – allows a machine to understand the right time to strike up a conversation. She would also be able to understand when a person is just passing through and therefore does not seek to interact with them. Using the second, called “turning towards/moving away”, the team says they can create a more natural dialogue between people and devices. Finally, a machine would manage to understand when a person is likely to have his attention with the last so-called “peek”.
This isn’t the first time the ATAP division has incorporated Soli technology into a device. Last year, it used it to offer the “sleep detection” feature in the Nest Hub connected display. Using radar and a cough and snore detection algorithm, it tracks sleep without the need for user contact. The technology has also been integrated into Google’s Pixel 4 smartphone to allow you to pause songs or snooze alarms with a wave of your hand.