Segmented HD Map Updates for Autonomous Vehicles
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional systems for generating high-definition (HD) maps update maps infrequently, leading to outdated maps that are less reliable for autonomous and semi-autonomous vehicles. Additionally, these systems require large amounts of time and computing resources due to batch rebuilds and centralized data processing.
Innovation Solution
The system segments maps into portions, updating each segment based on received data from machines navigating within the environment. This approach allows for continuous, segmented updates when sufficient data is available, reducing the time and resources needed for updates.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional systems perform batch rebuilds of HD maps using large amounts of sensor data, then the maps can be updated to reflect environmental changes, but the update process requires large amounts of time and computing resources
Solution Approach 1:
The patent divides the HD map into multiple segments or tiles that can be updated independently. Instead of performing a complete batch rebuild of the entire map, the system updates only the affected segments based on new sensor data, significantly reducing update time and computational resources while maintaining overall map accuracy.
2Reliability
If conventional systems centrally receive and store all sensor data for generating HD maps, then comprehensive map coverage is achieved, but large network bandwidth and computing resources are required
Solution Approach 1:
The patent extracts and processes only the necessary sensor data relevant to specific map segments rather than centrally collecting all sensor data. This selective data extraction approach maintains comprehensive map coverage while reducing the volume of data that needs to be transmitted and stored.
Solution Approach 2:
The system segments the map and associated sensor data into manageable portions that can be processed and updated independently. This segmentation allows the system to maintain comprehensive coverage without requiring centralized processing of all data, reducing network bandwidth and computing resource requirements.
3Ease of manufacture
If conventional systems update HD maps at fixed time intervals, then the update process is simple to implement, but the maps become outdated before the next update cycle
Solution Approach 1:
The patent implements dynamic update scheduling where the system transitions from fixed time intervals to event-driven updates. The system monitors environmental changes and triggers map updates when specific conditions are met (such as detecting new obstacles, road changes, or construction), ensuring maps remain current without requiring complex continuous processing.
Data Source
AI summary
In various examples, multiple stage map creations for autonomous and semi-autonomous systems and applications is described herein. For instance, systems and methods are described herein that update portions of a high-definition (HD) map in one or more stages in order to ensure that updates between different portions of the HD map are synchronized with respect to one another and/or ensure that updates to portions of the HD map are compatible with portions of the HD map that are not updated. In order to perform such processes, the systems and methods may determine to update one or more portions of the HD map. The systems and methods may use multiple states for the portion(s) when performing the updates, such as a first state (e.g., a locked state) when updating the portion(s) and a second state (e.g., an unlocked state) when not performing an update on the portion(s).