Vehicle Sensor Error Detection via V2V Environment Comparison
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing vehicle systems face issues with sensor calibration, damage, dirt, or failure, leading to incorrect object detection and inaccurate ego-localization, which hinders automated driving and assistance systems.
Innovation Solution
A method and system utilizing vehicle-to-vehicle communication to compare environment information from multiple sensors and vehicles, employing neural networks, probabilistic models, and fuzzy logic to detect sensor and localization errors, and deactivate affected systems.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If sensor-based environment perception systems are used for automated driving, then environment perception capability is improved, but sensor errors and localization errors occur leading to reduced reliability
Solution Approach 1:
The system implements a feedback mechanism where environment information from multiple sensors is continuously compared with V2X transmitted information. Discrepancies trigger error detection and notification processes, allowing the system to identify and respond to sensor or localization errors. This closed-loop feedback improves reliability by enabling real-time error detection without compromising measurement precision.
Solution Approach 2:
V2X communication serves as an intermediary layer between multiple vehicles' sensor systems. By comparing environment information through this intermediary channel, the system can detect errors in individual sensor systems without directly modifying the sensors themselves, thus maintaining measurement precision while improving overall reliability.
2Measurement precision
If multiple sensors are used for environment perception, then perception accuracy is improved, but system complexity increases
Solution Approach 1:
The system merges environment information from multiple sensors into a unified environment model that can be compared with V2X transmitted information. This consolidation approach maintains the benefits of multiple sensors for improved perception accuracy while reducing operational complexity by providing a single integrated comparison mechanism rather than requiring separate analysis for each sensor.
Solution Approach 2:
The environment perception system is designed with multi-functionality, where the same sensor system serves both primary navigation functions and error detection functions through V2X comparison. This universal approach allows the system to maintain high measurement precision while avoiding the additional complexity that would arise from dedicated separate error detection sensors.
3Productivity
If sensor errors are not detected, then system operation continues, but automated driving and assistance systems become unsafe
Solution Approach 1:
The system performs preliminary error detection by continuously comparing environment information with V2X transmitted information before automated driving operations proceed with potentially erroneous data. This preliminary check ensures that safety-critical decisions are not made based on erroneous sensor or localization data, maintaining both operational continuity and safety through proactive error identification.
Solution Approach 2:
The feedback mechanism notifies the vehicle when sensor or localization errors are detected, enabling immediate corrective action. This feedback loop ensures that automated driving operations can maintain productivity by quickly switching to backup systems or alerting the driver, thereby preserving safety without requiring complete system shutdowns.
Data Source
AI summary
A method for determining sensor errors and/or localization errors of a vehicle, including: recording a vehicle environment of an ego vehicle by means of at least one environment perception sensor of the ego vehicle; generating an environment representation based on the recording by the environment perception sensor; analyzing the environment representation in the ego vehicle by means of an analysis unit; determining environment information based on the analysis of the environment representation; entering the environment information in an environment model; determining at least one further vehicle in the environment representation from the at least one environment perception sensor of the ego vehicle; transmitting information from the at least one further vehicle to the ego vehicle by means of a vehicle-to-vehicle communication unit; comparing the transmitted information with the environment information from the ego vehicle; determining a sensor error and/or localization error based on the result of the comparison of the transmitted information and the environment information from the ego vehicle.