Automated Valet Parking Position Failure Detection in Narrow Aisles
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Automated valet parking systems face challenges in accurately determining the position of autonomous vehicles due to errors from on-board sensors, which can lead to misalignment with parking aisles and increased risk of collisions with walls or other vehicles.
Innovation Solution
The system employs a dual-position acquisition method using both on-board sensors and facility sensors, with a position failure determination unit that sets adaptive thresholds based on aisle width, node intervals, and node types to accurately assess position errors and prevent collisions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If on-board sensors are used to acquire vehicle position, then the system can operate autonomously without external infrastructure, but position accuracy deteriorates due to sensor errors
Solution Approach 1:
The patent introduces facility sensors installed in the parking place as an intermediary component to detect vehicle position. These sensors act as a mediator between the autonomous vehicle and the parking place infrastructure, providing accurate position information that compensates for on-board sensor errors. The facility sensors include detection units positioned to detect vehicles in aisles and parking spaces, with a determination unit that calculates actual positions based on detection results.
2Device complexity
If uniform position thresholds are used for all directions, then the determination logic is simple, but position failure detection accuracy deteriorates due to different error characteristics in longitudinal and lateral directions
Solution Approach 1:
The patent applies different position thresholds for different spatial directions. Specifically, a first position threshold is used for longitudinal position differences (along the aisle) and a second position threshold is used for lateral position differences (across the aisle). This local differentiation allows the system to account for the different error characteristics and safety requirements in each direction, improving position failure detection accuracy without excessive complexity.
Data Source
AI summary
An automated valet parking system is configured to acquire a first vehicle position, and acquire a second vehicle position, and determine whether a position failure about the first vehicle position of an autonomous vehicle is. The first vehicle position includes a first longitudinal position and a first lateral position, the second vehicle position includes a second longitudinal position and a second lateral position, and a longitudinal threshold is smaller than a lateral threshold. The automated valet parking system is configured to determine that the position failure is in at least one of a case where a difference between the first longitudinal position and the second longitudinal position is equal to or larger than the longitudinal threshold, and a case where a difference between the first lateral position and the second lateral position is equal to or larger than the lateral threshold.