Smooth voice control is one of the most important features of Motors' vehicles, and the company continues to put efforts into this area.

Xpeng has filed a patent related to vehicle control through voice, a method for natural language understanding of voice requests based on the current graphical user interface, such that the understanding of the voice request corresponds to the information in the graphical user interface.

It was filed on April 20, 2021, and disclosed on July 6, 2021.

The patent abstract shows that the vehicle can perform natural language understanding of a voice request based on information from a graphical user interface to enable a client to execute the command and display the result of the execution through the graphical user interface.

In short, if the patent is used for Xpeng's intelligent voice assistant, it is expected to improve its understanding of voice commands and present the results to the user in a better way on the interface.

Xpeng and are among the car companies leading the way in smart voice assistant functionality, with a previous report from the former stating that its smart voice assistant penetration rate exceeded 90 percent for more than 20 consecutive months.

Xpeng P7's in-car voice system allows for full-scene continuous dialogue, waking up once to give voice commands continuously, with up to 10 voice commands every 25 seconds.

It allows users to make semantic interruptions during voice conversations and enables invalid semantic filtering.

Xpeng P7 also has dual-tone zone locking capability, so that it will only execute the appropriate commands when the primary driver's voice wakes up, and not incorrectly execute commands related to other passenger locations.

These advanced capabilities make the Xpeng P7's voice in-car system smarter and more intelligent, and its conversations feel closer to those with real people.