Remote Driving Scene Reconstruction for Low-Bandwidth Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The existing remote driving systems face issues with network congestion and instability due to high bandwidth occupancy from real-time video transmission, leading to poor stability and efficiency in vehicle control.
Innovation Solution
A remote driving method and apparatus that utilizes pre-constructed global scene data to generate environment images based on local scene data, reducing the need for real-time video transmission and minimizing bandwidth occupancy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If real-time video transmission is used to transmit surrounding environment information from vehicle to remote driving simulator, then the remote driver can observe the environment, but network bandwidth occupancy increases causing congestion and instability
Solution Approach 1:
The patent extracts only the essential environmental information needed for remote driving from the complete video stream. Instead of transmitting all video data, the system identifies and transmits only critical elements such as road conditions, traffic signals, and relevant objects, thereby reducing bandwidth occupancy while maintaining control stability
Solution Approach 2:
The patent applies different transmission qualities to different parts of the environment based on their importance. Critical areas requiring precise control information receive higher quality transmission, while less important areas use lower quality or are omitted entirely, optimizing bandwidth usage while ensuring reliable control
2Loss of information
If multiple cameras are deployed to collect comprehensive video information, then the surrounding environment can be fully observed, but data transmission load and network congestion increase
Solution Approach 1:
The patent extracts only the most relevant environmental information from multiple camera sources. Instead of transmitting all camera feeds, the system identifies and transmits only critical data elements needed for safe remote operation, reducing transmission load while maintaining information completeness
Solution Approach 2:
The patent segments the environmental information from multiple cameras into distinct functional categories (road conditions, traffic signals, obstacles, etc.). This segmentation allows selective transmission of only necessary segments, improving control efficiency by reducing redundant data transmission
3Measurement precision
If high-resolution video images are transmitted in real-time, then the remote driver has clear visibility, but network bandwidth consumption increases leading to poor stability
Solution Approach 1:
The patent applies variable quality transmission where only critical regions of the environment are transmitted in high resolution, while other areas use lower resolution. This ensures clear visibility of important elements while reducing overall bandwidth consumption to maintain network stability
Data Source
AI summary
According to an aspect of the disclosure, a remote driving method, applied to a remote driving entity and includes displaying a first environment image corresponding to a target vehicle in response to a remote driving request, wherein the first environment image may include a first image of at least a part of a target environment corresponding to the target vehicle at a first location, and wherein the first environment image is generated based on first local scene data, corresponding to the first location, in pre-constructed global scene data of the target environment; and displaying a second environment image corresponding to the target vehicle in response to a vehicle driving operation of a driver on the target vehicle, the second environment image including a second image of at least a part of the target environment corresponding to the target vehicle at a current location.