HD Map Fusion for Low-Latency Autonomous Vehicle Navigation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current map resolution for autonomous driving is insufficient, requiring higher definition and real-time maps to support safe and reliable navigation, while existing V2V and V2X communication technologies face limitations in data rate and latency, making it challenging to integrate high-definition maps effectively across vehicles and infrastructure.
Innovation Solution
An apparatus and network configuration that combines sensor data with received high-definition maps to generate a fused map, enabling real-time transmission and processing of high-definition maps with millimeter-wave communication, supporting vehicle control commands and covering a wide area with low latency and high precision.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If current map resolution is used for autonomous driving, then device complexity is reduced, but navigation precision and safety are insufficient
Solution Approach 1:
The system segments the high-definition map into multiple frames that can be transmitted separately through V2V communication. Each frame contains a portion of the map data, allowing the system to achieve high resolution without overwhelming the communication bandwidth or processing capacity of individual vehicles.
Solution Approach 2:
The patent transitions from traditional 2D map representations to 3D point cloud data structures, enabling much higher resolution and depth information. This dimensional expansion allows autonomous vehicles to perceive the environment with centimeter-level precision while distributing the data burden across multiple transmission frames.
2Quantity of substance
If V2V communication is used to transmit HD maps, then map data can be shared between vehicles, but data rate is limited and cannot support huge data size
Solution Approach 1:
The HD map data is divided into multiple manageable frames for transmission. This segmentation allows the system to work within the bandwidth constraints of V2V communication while still conveying the complete high-definition map information through sequential frame transmission.
Solution Approach 2:
The system performs preliminary processing of the HD map data, converting it into a compressed point cloud format and organizing it into pre-segmented frames before transmission. This preparation reduces the real-time processing burden and optimizes the data for efficient V2V communication.
3Area of stationary object
If V2X communication through infrastructure is used, then map coverage can be extended, but latency increases beyond 10 ms requirement
Solution Approach 1:
The system merges V2V direct communication with V2X infrastructure-based communication. Vehicles can receive HD map frames directly from nearby vehicles with low latency, while also accessing broader coverage through infrastructure nodes, creating a hybrid architecture that achieves both speed and coverage.
Solution Approach 2:
The patent introduces intermediary caching mechanisms where roadside units and vehicles store previously received map frames. This allows vehicles to retrieve map data from nearby intermediaries rather than always communicating directly with the central infrastructure, reducing latency while maintaining extended coverage.
4Measurement precision
If HD map data size is increased to 1 Tbyte for one hour driving, then navigation precision is improved, but communication bandwidth requirement exceeds 2.2 Gbps
Solution Approach 1:
The 1 Tbyte HD map is segmented into multiple frames transmitted over time. This allows the system to achieve high navigation precision through cumulative reception of all frames while keeping the instantaneous data rate within the 2.2 Gbps capability of millimeter-wave communication systems.
Solution Approach 2:
The system transforms the HD map into a point cloud data representation, which achieves higher precision with more efficient compression. This parameter change in data structure allows maintaining navigation precision while reducing the communication bandwidth requirement.
Data Source
AI summary
An apparatus including a processor unit configured to provide a first high-definition map, a sensor unit configured for providing sensor data representing an environmental condition in a periphery of the apparatus and a receiver unit configured to receive data representing a second high-definition map. The processor unit is configured to fuse the second high-definition map and the sensor data so as to provide the first high-definition map. The apparatus further includes a transmitter unit configured for transmitting a result of fusing the second high-definition map and the sensor data, and includes a command generator unit configured to generate a command signal representing a vehicle control command for a vehicle carrying the apparatus based on the first high-definition map.