Autonomous Vehicle Sensor Data Association
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous vehicles face challenges in accurately distinguishing between sensor data from stationary and dynamic objects due to incorrect associations, particularly when objects are close or one is thin, leading to misinterpretation of the driving environment and potential adverse vehicle performance.
Innovation Solution
The implementation of a method and system that uses radar and lidar sensors to minimize incorrect associations by determining the location of stationary objects based on radar data and preventing lidar data from being associated with radar data at that location, or by assigning unique identifiers to sensor data for each object type to ensure accurate data association across multiple sensor observations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If lidar data and radar data are associated based on spatial proximity, then sensor data fusion is simplified and processing speed is improved, but incorrect associations occur when dynamic objects pass near stationary objects
Solution Approach 1:
The system performs preliminary classification of radar detections as stationary or dynamic objects before associating lidar data. By pre-identifying stationary objects and their locations in prior frames, the system establishes a reference that prevents dynamic objects from being incorrectly associated with stationary object radar data, thus maintaining both processing speed and association accuracy
Solution Approach 2:
The system introduces an intermediary classification step that distinguishes between stationary and dynamic radar detections. This intermediary layer acts as a mediator between raw sensor data and final object associations, ensuring that lidar data is only associated with appropriate radar detections based on their motion characteristics rather than just spatial proximity
2Reliability
If the system processes multiple sensor observations to improve association accuracy, then data association reliability is improved, but computational complexity and processing time increase
Solution Approach 1:
The system segments the sensor processing pipeline into distinct stages: radar detection and classification, stationary object identification, and selective lidar-radar association. This segmentation allows each stage to focus on specific tasks with optimized algorithms, improving overall reliability without proportionally increasing complexity
Solution Approach 2:
The system applies full multi-frame processing and classification only to radar detections that require stationary object identification, rather than processing all sensor data with the same level of complexity. This partial application of complex processing reduces overall computational burden while maintaining association accuracy where it matters most
3Device complexity
If the system uses only radar data for object detection, then the sensor system is simpler and cost-effective, but the ability to distinguish stationary and dynamic objects accurately is reduced
Solution Approach 1:
The system merges radar and lidar sensor data to achieve accurate object classification. Radar provides detection and velocity information for both stationary and dynamic objects, while lidar provides precise spatial information. By combining these complementary data sources, the system achieves high measurement precision for distinguishing object types while maintaining a relatively simple sensor configuration
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 improves the safety and performance of autonomous vehicles by reducing false associations, avoiding sudden maneuvers, and preventing collisions with stationary objects, thereby enhancing overall vehicle operation and occupant comfort.
Implementation Method 1
The first sensor observation can include radar data for the dynamic object and the stationary object
Implementation Method 2
The second sensor observation can include radar data and lidar data for the dynamic object and for the stationary object
Data Source
AI summary
Minimizing incorrect associations of sensor data for an autonomous vehicle are described. A driving environment of the autonomous vehicle includes a stationary object and a dynamic object. Such objects can be detected by radar sensors and/or lidar sensors. In one example, a history of radar observation can be used to minimize incorrect sensor data associations. In such case, the location of a stationary object in the driving environment can be determined. When a dynamic object passes by the stationary object, lidar data of the dynamic object is prevented from being associated with radar data obtained substantially at the determined location of the stationary object. In another example, identifiers assigned to radar data can be used to minimize incorrect sensor data associations. In such case, lidar data of an object can be associated with radar data having a particular identifier.