Driving Surface Curvature Estimation Using Map and Perception Fusion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional systems for estimating road curvature rely on a single data source, leading to inaccurate and unreliable curvature predictions, especially when sensors are obscured or provide incomplete data.
Innovation Solution
A system that updates curvature estimates using a combination of map data and perception data, leveraging a Kalman filter to refine curvature predictions based on previous estimates, trajectory, and sensor measurements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If a single sensor modality is used to estimate road curvature, then the system complexity is reduced, but the measurement precision and reliability deteriorate when the sensor is obscured or provides incomplete data
Solution Approach 1:
The patent combines multiple sensor modalities (image sensor, LiDAR, radar) and map data to estimate road curvature. By merging these diverse data sources, the system achieves more reliable curvature estimation even when individual sensors are obscured, directly resolving the contradiction between system complexity and measurement precision.
Solution Approach 2:
The patent introduces map data as an intermediary source to complement real-time sensor data. When sensor data is incomplete or unreliable, map data serves as a mediator to provide curvature information, maintaining measurement precision without requiring complex real-time sensing in all conditions.
2Measurement precision
If multiple data sources are integrated to improve curvature estimation accuracy, then the measurement precision improves, but the device complexity increases
Solution Approach 1:
The patent implements a feedback mechanism where curvature estimates from multiple sources are continuously compared and refined. The system uses predicted curvature from map data and actual curvature from sensors to update and improve estimates, achieving high precision through iterative refinement rather than simply combining all possible sensors.
Solution Approach 2:
The patent changes the parameter of data fusion from simple concatenation to weighted integration based on data quality and reliability. By dynamically adjusting how different data sources contribute to the final estimate, the system achieves high precision without linearly increasing complexity.
3Ease of operation
If conventional single-source curvature estimation is used, then the ease of operation is maintained, but the reliability deteriorates in scenarios with obscured sensors or incomplete data
Solution Approach 1:
The patent enables the system to self-adjust by automatically selecting and weighting appropriate data sources based on their current reliability. The system monitors data quality from each sensor and map data, and autonomously determines the best combination, maintaining ease of operation while improving reliability through intelligent self-management.
Data Source
AI summary
In various examples, an estimated curvature associated with a driving surface may be updated or improved based on additional sources of information, such as map data and/or perception data. For instance, systems and methods are disclosed that may predict curvature (e.g., magnitudes of curvature) for one or more portions and/or points along a driving surface traversed by a machine. The predicted curvature may be determined based on one or more previous curvature predictions for the driving surface and based on a trajectory and/or a distance traveled by the machine subsequent to making those previous curvature predictions. In some instances, the predicted curvature may be updated based on one or more measured curvatures associated with the driving surface. These measured curvatures may be determined using map data associated with the driving surface and/or perception data generated from sensor data obtained using one or more sensors of the machine.