Vehicle Perception Data Mapping for Road Friction Training
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current technologies face challenges in providing high-quality training data for environment detection sensors to accurately estimate friction conditions, essential for both autonomous and non-autonomous vehicles, due to the scarcity or low quality of suitable training data.
Innovation Solution
A method that associates environment perception data with friction condition data using a second sensor's data, involving coordinate transformation and mapping to align the data, allowing for high-volume, high-quality data generation with low effort and high precision, which can be used as training data for machine learning units.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If environment perception sensors are trained to detect friction conditions, then the reliability of friction detection is improved, but the quality and availability of training data deteriorates (scarcity or low quality)
Solution Approach 1:
The patent uses a second sensor as an intermediary to measure friction conditions directly, and this sensor serves as a mediator to provide ground truth data for training the environment perception sensor. The second sensor's measurements bridge the gap between visual observations and actual friction conditions, enabling reliable training data generation without requiring scarce manual annotations or specialized friction measurement equipment.
2Measurement precision
If manual annotation of training data is performed to ensure high quality, then the precision of training data is improved, but the productivity of data generation deteriorates (comparatively low effort required by the method)
Solution Approach 1:
The system performs self-service by automatically generating high-quality training data through the coordinated operation of the environment perception sensor and the second sensor. The automated association process, including coordinate transformation and mapping, eliminates the need for manual annotation while maintaining high precision, thereby achieving both data quality and high productivity simultaneously.
Solution Approach 2:
The patent establishes predetermined coordinate systems for both sensors and pre-defines the association methodology before data collection begins. This preliminary setup enables automated real-time association of data points without manual intervention during the data generation process, ensuring both precision and high productivity.
3Measurement precision
If sensors are positioned to capture friction-relevant information, then the measurement precision of friction conditions is improved, but the device complexity increases (multiple sensors and coordinate transformations)
Solution Approach 1:
The second sensor serves multiple functions: it directly measures friction conditions, provides ground truth data for training, and enables the environment perception sensor to learn friction detection capabilities. This multi-functionality justifies the addition of the second sensor by maximizing its utility across multiple objectives.
Solution Approach 2:
The patent replaces complex manual friction measurement mechanisms with a sensor-based automated measurement system. The second sensor electronically captures friction conditions and enables automated data association through computational methods, substituting mechanical complexity with electronic and software-based solutions that are more precise and easier to manage.
Data Source
Figure 1~2
Figure 3~4
Figure 5~6
AI summary
The disclosure relates to a method for associating environment perception data (R) of a first vehicle (10) with data indicative of a friction condition (F). The method comprises receiving the environment perception data (R), position data, and data indicative of a friction condition (F) from a second sensor (20) being associated with the first vehicle (10). A first position (22) is transformed in a coordinate system being centered on the first sensor (14) using coordinate transformation based on the position data. Moreover, the first position (22) is mapped onto at least one corresponding element of the environment perception data (R) and the at least one element is associated with the data indicative of a friction condition (F). Furthermore, a computer program (32), a computer-readable storage medium (30), and data processing apparatus (24) are explained. Additionally, a system (12) for associating environment perception data (R) of a first vehicle (10) with data indicative of a friction condition (F), a vehicle (10), associated environment perception data (R) and a use of associated environment perception data (R) are described.