Road Rule Mapping From Onboard Sensor Data Without HD Maps
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Pre-generated detailed maps for autonomous vehicles are labor-intensive to create and quickly become outdated due to changes in road infrastructure, making them unreliable for real-time navigation.
Innovation Solution
A system that uses locally captured sensor data, such as from cameras and Lidar sensors, to identify lane segments and observables, which are then processed by a neural network to determine driving rules in real-time, allowing for the generation of up-to-date 'rules of the road' maps without human intervention.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If pre-generated detailed maps are created using manual annotation, then map accuracy is improved, but the time and labor required for map creation increases significantly
Solution Approach 1:
The system enables autonomous vehicles to automatically generate and update their own navigation maps using onboard sensors (cameras, LiDAR, radar) without requiring manual annotation. The vehicle captures sensor data, processes it through neural networks to identify lane segments and observables, and determines driving rules autonomously, making the map creation process self-service rather than labor-intensive
Solution Approach 2:
The patent replaces the mechanical process of manual map annotation with an automated computational system. Neural networks and sensor fusion algorithms process sensor data to automatically identify road features and determine driving rules, substituting human labor with computational processing that occurs in real-time as the vehicle operates
2Reliability
If pre-generated detailed maps are updated manually to reflect road changes, then map reliability is improved, but the frequency and cost of updates increase
Solution Approach 1:
The system continuously updates map information as the autonomous vehicle operates on the road. Instead of periodic manual updates, the vehicle constantly captures sensor data, processes it to identify current road conditions and driving rules, and maintains up-to-date navigation information throughout its operation, ensuring continuous reliability without interruption
Solution Approach 2:
The autonomous vehicle automatically detects and adapts to road changes (construction, new signage, lane modifications) using its onboard sensors and neural networks. The system self-updates its navigation knowledge without requiring external manual intervention, maintaining reliability by independently responding to environmental changes as they occur
3Measurement precision
If manual annotation is used to create detailed maps, then driving rule accuracy is improved, but the complexity and cost of the mapping process increases
Solution Approach 1:
The patent replaces complex manual annotation processes with automated neural network processing. The system uses sensor fusion algorithms and deep learning models to automatically interpret sensor data, identify road features, and determine driving rules, reducing the complexity of human coordination and annotation while maintaining or improving accuracy through computational analysis
4Reliability
If pre-generated maps are used for navigation, then initial navigation performance is improved, but the maps become outdated quickly due to road modifications
Solution Approach 1:
The navigation map transitions from a static pre-generated dataset to a dynamic system that continuously updates. The autonomous vehicle's onboard sensors and processing systems continuously capture and process road information, allowing the navigation data to adapt in real-time to road modifications, construction, and environmental changes, extending the validity period indefinitely through continuous renewal
Data Source
AI summary
A rules of the road module and database determines rules using information about segments identified from locally captured image data and observables identified in the image data. The rules are generated without reliance on a high-definition map. Also, in some embodiments, the rules are determined in real-time, or near-real time by the rules of the road module during travel based on acquired images and sensor data about a path being travelled.