Driverless Vehicle Sensor Monitoring via Cross-Validation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current driverless vehicle technologies lack effective methods to monitor and address sensor abnormalities, which can compromise safety due to physical damage or hacker attacks, leading to potential malfunctions and safety risks.
Innovation Solution
A method and apparatus for monitoring sensors in driverless vehicles, involving real-time monitoring of physical and data transmission states, and cross-validation of output data using data from other sensors and high-precision maps to detect abnormalities and trigger alarms.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If sensor monitoring and cross-validation systems are implemented in driverless vehicles, then safety and reliability are improved, but device complexity increases
Solution Approach 1:
The system performs preliminary monitoring of sensor physical states and data transmission states before cross-validation of output data. This staged approach allows the system to detect abnormalities early in the process, preventing unnecessary complex validation operations and reducing overall system complexity while maintaining safety.
Solution Approach 2:
The monitoring system is divided into three independent modules: physical state monitoring, data transmission state monitoring, and output data cross-validation. Each module operates independently and can be processed separately, reducing the complexity of the overall system while ensuring comprehensive safety coverage.
2Measurement precision
If real-time monitoring of sensor physical state and data transmission state is performed, then abnormality detection capability is improved, but energy consumption increases
Solution Approach 1:
The system performs partial monitoring by focusing on key parameters such as physical state and data transmission state rather than continuously analyzing all sensor data in real-time. This selective monitoring approach maintains high abnormality detection capability while significantly reducing energy consumption compared to full real-time analysis.
Solution Approach 2:
The system performs preliminary checks on sensor physical states and data transmission states before conducting more energy-intensive cross-validation of output data. This staged approach ensures that energy-consuming validation operations are only performed when necessary, optimizing the balance between detection capability and energy consumption.
3Measurement precision
If cross-validation using multiple data sources is implemented, then data accuracy is improved, but processing time increases
Solution Approach 1:
The cross-validation process is segmented into distinct stages: physical state validation, data transmission validation, and output data validation. Each stage processes specific aspects of sensor data independently, allowing for optimized processing at each level and reducing overall processing time while maintaining data accuracy.
Solution Approach 2:
The system performs partial cross-validation by focusing on critical data aspects rather than validating every parameter exhaustively. The monitoring prioritizes essential validations (physical state, transmission state) and performs output data cross-validation selectively, achieving sufficient data accuracy without excessive processing time.
Data Source
AI summary
The present disclosure provides a method and apparatus of monitoring a sensor of a driverless vehicle, a device and a storage medium, wherein the method comprises: monitoring a physical state of a to-be-monitored sensor; monitoring a data transmission state of the to-be-monitored sensor; monitoring output data of the to-be-monitored sensor, and using predetermined data to perform cross-validation for the output data; when any monitoring result gets abnormal, determining the to-be-monitored sensor as getting abnormal, and giving an alarm. The solution of the present disclosure may be applied to improve safety of the driverless vehicle.