HD Map Layer Distribution for Low-Latency Autonomous Vehicle Updates
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing HD map systems are inefficient due to large data files that include unnecessary information, requiring multiple tile servers for low latency, and causing disruptions during map updates that can render cached data unsafe for use.
Innovation Solution
A distributed architecture using a network of servers, such as a content delivery network, to serve and modify HD map data efficiently, allowing for lightweight hosting and serving of map data, and enabling incremental updates to map components like tiles, layers, and segments.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional HD map systems include all sensor modalities and machine operations data, then mapping accuracy and localization precision are improved, but data file size increases and unnecessary information is transferred
Solution Approach 1:
The HD map data is segmented into multiple layers, each representing different sensor modalities (e.g., LiDAR, camera, radar) and machine operations (planning, control, localization). This segmentation allows the system to divide the comprehensive map into manageable, independently selectable components, enabling vehicles to download only the layers relevant to their specific sensors and functions rather than receiving entire large-scale map files.
Solution Approach 2:
The system extracts and separates specific map layers corresponding to particular sensor modalities and operational functions from the complete HD map. This extraction mechanism enables the server to provide customized map data packages tailored to individual vehicle requirements, removing unnecessary data elements while preserving the essential localization and navigation information needed for each vehicle's capabilities.
2Speed
If multiple tile servers are deployed globally to serve HD map data, then low latency is achieved, but system complexity and infrastructure costs increase
Solution Approach 1:
The patent implements a universal map data structure and serving mechanism that can handle multiple sensor modalities and functional requirements through a single, standardized interface. This universal approach allows the server infrastructure to serve diverse vehicle types (with different sensor configurations) without requiring specialized server instances, thereby reducing overall system complexity while maintaining efficient data delivery performance.
3Reliability
If HD map updates are applied to all tile servers, then map accuracy is maintained, but service disruptions occur and cached data becomes unsafe
Solution Approach 1:
The system performs preliminary validation and version tagging of map updates before deploying them to the serving infrastructure. By pre-processing updates and preparing them in advance with compatibility information, the system can safely serve updated map layers to vehicles without causing service disruptions. This preliminary action ensures that cached data remains safe and that transitions between map versions occur smoothly without requiring manual takeover or interrupting autonomous driving operations.
4Measurement precision
If complete HD map downloads are required for each vehicle, then comprehensive localization capability is achieved, but bandwidth consumption and update time increase
Solution Approach 1:
The HD map is segmented into functional layers corresponding to different sensor modalities and operational requirements. This segmentation enables vehicles to download only the specific layers needed for their localization and navigation tasks, dramatically reducing download time and bandwidth consumption while maintaining comprehensive localization capability for vehicles equipped with appropriate sensors.
Solution Approach 2:
The system implements partial downloading of map data based on vehicle-specific requirements. Rather than requiring complete map downloads, vehicles receive only the partial set of map layers necessary for their particular sensor configuration and operational needs, achieving sufficient localization accuracy with minimal data transfer and updated time.
Data Source
AI summary
In various examples, a network of servers, such as a content delivery network, is used to provide a lightweight approach to hosting and serving HD map data to vehicles. The lightweight approach may allow for modifying various map components, such as tiles, layers, and/or segments. Modifying may include adding, removing, and/or updating the various components. A request to modify a first version of a High definition (HD) map may be received. Map data may be recorded that represents a second version of the HD map. A second request associated with the HD map may be received from a vehicle. Based on this second request, second map data representative of at least a portion of a layer may be identified on at least one server of the network of servers. The second map data may then be transmitted to the vehicle by the network of servers.