MAUI, United States - Bringing user information collected by watches, earbuds and AI glasses into a single device to run AI is emerging as a challenge for wearable connectivity. Users now use iPhones, Android phones and Windows PCs together, while also adding wearables from multiple brands. As users wear devices from different brands at the same time, the need is growing for industry standards that allow information sharing across devices beyond operating system boundaries.

Ziad Asghar (지아드 아스가르), Qualcomm senior vice president and head of XR, wearables and personal AI, said at a Snapdragon Summit Personal AI and Sound media Q&A on Sept. 23 local time that watches provide activity levels, heart rate and stress levels, earbuds provide additional health signals and glasses provide video information. One device among that group will gather the different inputs and deliver them to the device that runs the AI, he said.

The device that runs the AI depends on the nature of the question. Asghar explained that simple questions that glasses or a watch can handle are processed on that device, while AI runs where battery life and latency are most advantageous.

But whichever device runs AI, it must have the full context. To that end, Qualcomm is developing lower-level software that lets all devices share context without being affected by the operating system. Asghar said it needs to work across multiple platforms, and that requires something at a lower layer.

Qualcomm has created a framework to share device information and each device’s computing and AI capabilities, and named it the multi-device experience. Asghar said it effectively needs to become an industry standard to be done properly, and that more will be heard later about what it is pursuing in that direction.

Qualcomm is developing connectivity technology starting with audio devices. Andrew Reister (앤드루 레이스터), Qualcomm senior product manager, said Snapdragon Sound Elite 2nd generation inherits Qualcomm XPAN, which chooses between Bluetooth and Wi-Fi depending on the situation. That has lifted Wi-Fi data transfer speeds by more than three times and enabled support for the Wi-Fi 6E band. Asghar said that such transfer speeds and range are important for audio glasses that need to send video to the cloud, and that the audio glasses Meta launched that day are also Snapdragon-based.

Glasses use different platforms and connection methods depending on how they are used. For example, audio-only glasses without a camera use Sound Elite 2nd generation, while glasses that take high-resolution photos use Snapdragon AR1. Snap’s AR glasses, Spectacles, also use two Snapdragon platforms that split app and perception technologies, so they do not offload computing to other devices. When on the move, they currently connect to the cloud through a watch or smartphone modem. Reister said it is not trying to replace the smartphone, and is a feature for specific moments when it is difficult to take out a smartphone.

Qualcomm is also reviewing a plan to put 5G directly into AI glasses. Asghar said that if power efficiency improves, 5G connectivity can be supported even with an AI glasses battery, and that carrier partners are also showing strong interest. Adding 5G would make the glasses a fully independent device, he added.