Autonomous Driving Map Verification for HD Road Boundary Changes
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous vehicles face challenges in maintaining accurate localization and safe driving when the HD map does not match the current road environment, leading to potential inaccuracies and risks.
Innovation Solution
The method involves generating a real-time map using vehicle sensors, comparing it with HD map data to detect discrepancies, and either discontinuing autonomous driving or providing user guidance when environmental changes are detected.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If autonomous driving relies on HD map matching for localization, then localization accuracy is improved, but the system becomes vulnerable to failures when map data does not match the actual road environment
Solution Approach 1:
The patent introduces a road environment recognition unit that acts as an intermediary between the localization system and the actual road environment. This unit independently recognizes road boundaries and lane lines from sensor data, providing a verification layer that mediates between HD map-based localization and real-world conditions, alerting when discrepancies occur
Solution Approach 2:
The patent implements a feedback mechanism where the road environment recognition results are continuously compared with HD map data, and discrepancies trigger alerts or autonomous driving discontinuation. This closed-loop feedback ensures the system responds to environmental changes and maintains reliability when map data becomes outdated
2Productivity
If autonomous driving system continues operation despite map discrepancies, then productivity is maintained, but safety risks increase
Solution Approach 1:
The patent applies beforehand cushioning by preparing multiple response strategies in advance for different levels of map discrepancy: continuing autonomous driving with monitoring, switching to manual driving, or discontinuating autonomous driving entirely. This pre-prepared approach allows safe degradation of functionality rather than abrupt failure when discrepancies are detected
3Measurement precision
If HD map data is frequently updated to match road environment changes, then localization accuracy is maintained, but cost and time requirements increase
Solution Approach 1:
The patent implements self-service by enabling the vehicle to independently recognize and verify road environment features using its own sensors, rather than relying entirely on external HD map updates. The vehicle autonomously detects road boundaries and lane lines, comparing them with map data to identify discrepancies, effectively performing its own environmental verification without requiring frequent costly map recalibrations
Data Source
AI summary
An autonomous driving control method receives a HD map and extracts a HD road boundary and a HD lane line based on the HD map. A real-time map displaying a road driving environment of a vehicle is generated using sensor information detected by a sensor mounted on the vehicle. A real-time road boundary and a real-time lane are extracted based on the generated real-time map. The HD road boundary is compared to and analyzed from the real-time road boundary, and the HD lane line is compared and analyzed to the real-time lane. It is determined whether the HD map has changed based on the analyzed result value.