Dynamic Road Map Updates via Machine Learning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current advanced driver assistance systems (ADAS) and automated driving systems (ADS) do not effectively account for dynamic obstructions on roadways, such as accidents, potholes, or adverse weather conditions, which can cause road configurations to differ from pre-stored detailed maps.
Innovation Solution
A system and method that utilize a vehicle controller and sensors to gather input data about the environment, including abnormal traffic patterns, and use a machine learning algorithm, specifically a generative adversarial network (GAN), to generate an updated road map that includes dynamic changes in lane boundaries and road-surface conditions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If pre-stored detailed road maps are used for ADAS and ADS systems, then the systems can operate with predetermined road configuration information, but the systems cannot account for dynamic obstructions such as accidents, potholes, or adverse weather conditions that cause road configurations to differ from the stored maps
Solution Approach 1:
The patent implements a dynamic road map updating system that transforms static pre-stored road maps into adaptive structures. The system continuously receives sensor data from the vehicle and surrounding environment, processes this information through machine learning algorithms, and updates the road map configuration in real-time to reflect current road conditions, obstructions, and traffic patterns. This dynamic approach allows the system to maintain high reliability by adapting to changing road conditions while preserving the foundational structure of pre-stored maps.
Solution Approach 2:
The system establishes a feedback loop where sensor data from cameras, LIDAR, and other perception systems continuously monitors the actual road environment. This feedback is processed to detect discrepancies between the pre-stored road map and current conditions, triggering updates to the road map configuration. The feedback mechanism ensures that the system maintains accurate road configuration information while adapting to dynamic obstructions by constantly comparing expected versus actual road states.
2Adaptability or versatility
If real-time sensor data processing is implemented to detect abnormal traffic patterns and update road maps, then the system can account for dynamic obstructions, but the computational complexity and processing requirements increase
Solution Approach 1:
The system performs preliminary actions by pre-training machine learning models with extensive datasets of normal and abnormal traffic patterns before deployment. This pre-processing allows the models to quickly recognize and respond to anomalies in real-time without requiring complex computational resources during actual operation. The abnormal traffic pattern detection algorithms are pre-configured with knowledge of various obstruction types, enabling efficient real-time processing while maintaining high adaptability.
Solution Approach 2:
The patent applies local quality by focusing computational resources on specific areas of interest rather than processing entire road maps uniformly. The system identifies regions with abnormal traffic patterns or potential obstructions and concentrates processing power on these localized areas. This approach maintains high adaptability for detecting dynamic changes while reducing overall computational complexity by avoiding unnecessary processing in stable, normal traffic regions.
Data Source
AI summary
A system for updating a road map for a vehicle includes a plurality of vehicle sensors and a vehicle controller in electrical communication with the plurality of vehicle sensors. The vehicle controller is programmed to gather input data about an environment surrounding the vehicle using the plurality of vehicle sensors. The input data includes at least one abnormal traffic pattern indication. The vehicle controller is further programmed to generate an input label map based at least in part on the input data. The vehicle controller is further programmed to generate a vehicle output label map based at least in part on the input label map. The vehicle output label map is generated using a machine learning algorithm. The vehicle controller is further programmed to perform a first action based at least in part on the vehicle output label map.