Google

The single Networking patent filed this period focuses on how 5G handles uplink scheduling for latency-sensitive applications like wireless VR, AR, and cloud game Streaming. The approach allows a device to adjust its own uplink transmission resources in real time, on a per-traffic-period basis, within a pre-configured grant framework, rather than waiting for the network to issue dynamic scheduling commands. That distinction matters because dynamic grants introduce delays that undermine the low-latency experience these applications require, while purely static configured grants lack the flexibility to adapt to changing traffic demands. This filing describes a way to get the benefits of both.

The 2 AI & Machine Learning patents both center on improving the video game experience through machine learning, though they tackle different problems. One describes a system that monitors a player's in-game state in real time, detects where that player is struggling, and either locates an existing tutorial that addresses the issue or automatically generates one from the player's own captured session footage when no suitable content exists. The other takes a coaching approach rather than a tutorial approach, processing actual gameplay video and images to give players strategic guidance through a natural language conversation, where the player can ask follow-up questions and the system explains its reasoning step by step rather than simply outputting a recommended action. Together, the 2 filings describe complementary ways of using ML to close the gap between what players know and what the game requires of them.