Autonomous Vehicle Lane-Device Association Mapping
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current autonomous driving systems face challenges in efficiently and cost-effectively mapping intersections, particularly at complex intersections with multiple lanes or curves, where they struggle to determine which traffic control devices to obey without human annotation, which is time-consuming and expensive.
Innovation Solution
A system that includes a host vehicle equipped with sensors to detect traffic control devices and vehicle motion, a processor to identify and associate lanes with traffic control devices, and non-transient data storage for storing these associations, allowing autonomous vehicles to access and communicate this information for subsequent navigation, potentially using remote servers for data sharing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If human annotation is used to map intersections and associate traffic control devices with lanes, then mapping precision and reliability are improved, but time consumption and cost increase significantly
Solution Approach 1:
The system enables autonomous vehicles to automatically perform the mapping function themselves by detecting traffic control devices, identifying their positions, and associating them with lanes during normal operation. This self-service approach eliminates the need for external human annotation, thereby reducing time consumption while maintaining mapping precision through the vehicle's own sensor data and processing capabilities.
Solution Approach 2:
The patent introduces an automated processing system that acts as an intermediary between raw sensor data and the final mapping output. This intermediary system uses image processing algorithms and machine learning models to automatically detect traffic control devices, determine their types, and associate them with relevant lanes, replacing the human annotation process while maintaining or improving precision through consistent automated analysis.
2Measurement precision
If human annotation is used to map intersections and associate traffic control devices with lanes, then mapping precision and reliability are improved, but cost increases significantly
Solution Approach 1:
By enabling autonomous vehicles to perform self-annotation through automated detection and association systems, the patent eliminates the need for expensive human annotators. The system uses the vehicle's existing sensors, processors, and communication infrastructure to automatically create and update mapping data, significantly reducing costs while maintaining precision through automated image processing and data association algorithms.
Solution Approach 2:
The system creates digital copies of traffic control devices and their spatial relationships through automated image processing and data extraction. Instead of relying on human observers to manually record information, the system captures images, processes them through algorithms, and generates digital representations of the traffic control infrastructure, thereby reducing costs while preserving mapping precision through automated documentation.
3Productivity
If automated systems automatically detect and associate traffic control devices with lanes during normal operation, then productivity and cost-effectiveness are improved, but system complexity increases
Solution Approach 1:
The patent leverages the autonomous vehicle's existing multi-functional sensor suite and processing system to perform the additional mapping function. The same cameras, processors, and communication systems used for navigation and obstacle detection are also utilized for detecting traffic control devices and creating maps, thereby increasing productivity without requiring separate dedicated hardware systems, thus limiting the increase in overall system complexity.
Solution Approach 2:
The system merges the traffic control device detection and association functionality with the existing autonomous driving perception and navigation systems. By combining these functions into a unified processing pipeline that shares sensors, algorithms, and data structures, the patent achieves high mapping productivity while minimizing the increase in system complexity through integrated rather than separate system architectures.
Data Source
AI summary
A system for generating mapping data includes a host vehicle and at least one sensor coupled to the host vehicle. The sensor is capable of detecting a traffic control device and capable of detecting motion of a vehicle. The system additionally includes non-transient data storage and a processor. The processor is in communication with the sensor and the data storage. The processor is configured to, in response to the at least one sensor detecting a traffic control device and detecting motion of a vehicle during a drive cycle, identify a lane of traffic associated with the vehicle, associate the lane of traffic with the traffic control device, and store the association of the lane of traffic with the traffic control device in the data storage for subsequent access by an automated driving system.