Autonomous Vehicle Position Refinement Using Local and Remote Object Databases
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous driving systems face challenges in accurately determining vehicle position and navigating through environments with varying road conditions due to limitations in detecting and identifying off-road elevated objects, which can lead to inaccuracies in route refinement and safe operation.
Innovation Solution
A vehicle control system that utilizes a combination of local and remote databases to store and compare feature and position data of detected objects, with sensors like lidars and cameras, to identify expected objects and update vehicle position, incorporating user inputs for unidentified objects and filtering out on-road objects based on elevation thresholds.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the vehicle relies solely on local database for object identification, then the system complexity is reduced, but the measurement precision of vehicle position deteriorates due to limited reference data
Solution Approach 1:
The database system is segmented into local and remote components. The local database stores frequently accessed object reference data for quick comparison, while the remote database stores comprehensive reference data. This segmentation allows the system to achieve high measurement precision by accessing remote database when needed, while maintaining low device complexity through local database for routine operations.
Solution Approach 2:
The transceiver acts as an intermediary between the local and remote databases. When the identification unit cannot identify an object using local database, the transceiver communicates with the remote database to access additional object reference data, enabling the system to maintain high position accuracy without requiring the entire remote database to be present in the vehicle.
2Measurement precision
If the system compares all detected objects with comprehensive reference data, then the identification accuracy improves, but the processing time increases
Solution Approach 1:
The system performs preliminary comparison of detected objects with object reference data using the local database before accessing the remote database. This preliminary action filters out easily identifiable objects, reducing the number of objects that require time-consuming remote database queries and thereby reducing overall processing time while maintaining identification accuracy.
Solution Approach 2:
The system applies partial comparison action by first comparing objects with local database reference data, and only performing full comparison with remote database reference data when necessary. This partial action approach achieves sufficient identification accuracy for most cases while minimizing processing time by avoiding unnecessary comprehensive comparisons.
3Reliability
If the vehicle uses multiple sensors and databases to identify objects, then the reliability of autonomous operation improves, but the device complexity increases
Solution Approach 1:
The object reference data storage and comparison functionality is made universal by implementing both local and remote databases with the same data structure and comparison logic. The identification unit can operate with either local or remote database, or both, making the system reliable across different operating conditions while avoiding the need for separate specialized systems.
Solution Approach 2:
The system implements feedback by continuously comparing detected objects with reference data from local and remote databases, updating the vehicle position based on identification results, and using this updated position information to refine future object detections. This feedback loop enhances reliability by continuously verifying and correcting system state while maintaining a manageable architecture through iterative refinement.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach enhances the accuracy and safety of autonomous vehicle operation by continuously refining vehicle position using a multitude of reference points, reducing comparison requirements and ensuring precise navigation to desired locations.
Implementation Method 1
sensors like lidars and cameras
Implementation Method 2
detection unit can be configured to detect objects that are external to the vehicle
Data Source
AI summary
A vehicle can include a navigation unit configured to determine a position of the vehicle. A detection unit can be configured to recognize position and features of objects external to the vehicle. An identification unit can be configured to identify objects expected to be detected based on a determined position of the vehicle, by comparing the recognized features to feature data of object reference data stored in a local database. When the identification unit is unable to identify an object, the identification unit can compare recognized or stored object features to the additional feature data received from a remote database. A navigation unit can update the position of the vehicle based on position data of identified objects and more accurate and safe autonomous operation of a vehicle can be achieved.