Automated Valet Parking Trajectory Planning for Entrance Visibility
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous driving vehicles may struggle to recognize the entrance of target parking spaces due to objects being outside the object recognition range of their external sensors, especially on wide runways, leading to potential parking errors.
Innovation Solution
An automated valet parking system that includes a parking lot management server which acquires traveling coordinates and generates trajectories for the vehicle to bring the entrance of the target parking space within the object recognition range of the vehicle's external sensors, ensuring accurate recognition and parking.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the autonomous driving vehicle uses a fixed object recognition range based on sensor specifications and vehicle position, then the system structure remains simple, but the object (entrance of target parking space) may be located outside the recognition range on wide runways
Solution Approach 1:
The patent applies dynamics by making the trajectory generation adaptive rather than fixed. The trajectory generation unit dynamically adjusts the vehicle's path based on real-time determination of whether the entrance is within the object recognition range. This allows the system to handle varying runway widths and sensor capabilities without requiring complex hardware modifications, resolving the contradiction between recognition reliability and system complexity.
Solution Approach 2:
The patent implements feedback by continuously determining whether the entrance of the target parking space is within the object recognition range and using this information to adjust the trajectory. The system monitors the recognition status and modifies the vehicle's path accordingly, ensuring the entrance remains within the recognizable range while maintaining a relatively simple overall system structure.
2Manufacturing precision
If the vehicle travels directly to the target parking space without adjusting trajectory, then the operation is simple and fast, but the entrance may not be recognizable leading to parking errors
Solution Approach 1:
The patent applies preliminary action by determining in advance whether the entrance will be within the object recognition range during the vehicle's approach. The trajectory is adjusted beforehand to ensure proper recognition, preventing parking errors before they occur. This proactive approach balances parking accuracy requirements with time efficiency by avoiding last-minute corrections.
3Reliability
If the object recognition range is expanded to cover wide runways, then object recognition is ensured, but the sensor requirements and system cost increase
Solution Approach 1:
The patent resolves this contradiction by shifting from a spatial solution (expanding sensor range) to a temporal/path-based solution (adjusting trajectory). Instead of increasing sensor capability to cover wide runways, the system adjusts the vehicle's approach path to ensure the entrance falls within the existing sensor's recognition range. This maintains sensor simplicity while ensuring reliable recognition through intelligent path planning.
Data Source
AI summary
An automated valet parking system includes a parking lot management server configured to: acquire information on a plurality of traveling coordinates arranged along a runway of a parking lot to an entrance of a target parking space based on a vehicle position of an autonomous driving vehicle, a position of the target parking space, and parking lot map information; generate a trajectory of the vehicle based on the information on the traveling coordinates and a result of detection from an in-vehicle external sensor; and determine whether the entrance of the target parking space is recognizable based on an object recognition range; and generate the trajectory that brings the vehicle closer to the target parking space so that the object for recognizing the entrance of the target parking space is included in the object recognition range when it is determined that the entrance of the target parking space is not recognizable.