Autonomous Vehicle Perception Diagnostics via Geo-Source Comparison
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous vehicle perception systems face challenges in real-time diagnostics and verification of sensors and software outputs, which are crucial for reliable control and establishing ground truth regarding vehicle surroundings, due to the complexity of detecting and interpreting environmental data accurately.
Innovation Solution
A method for on-line diagnostic and prognostic assessment that involves detecting physical parameters of objects using sensors, communicating data to an electronic controller, comparing sensor data with geo-source model data, and generating a prognostic assessment of ground truth by weighing and grouping results from sensors and geo-source models to identify trustworthiness and detect faults in sensors and software.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple sensors and software systems are used to improve perception accuracy, then measurement precision is improved, but device complexity increases
Solution Approach 1:
The system implements on-line diagnostics that continuously monitor sensor outputs and software results, comparing them against expected ranges and patterns. This feedback mechanism enables real-time detection of deviations or faults in the complex multi-sensor perception system, maintaining measurement precision while managing complexity through automated monitoring
Solution Approach 2:
The perception system performs self-diagnosis by automatically assessing its own sensor data and software outputs without external intervention. The system generates its own diagnostic information and fault assessments, enabling it to self-monitor and self-evaluate its perception accuracy despite increasing complexity
2Reliability
If real-time diagnostics are implemented to improve reliability, then reliability is improved, but loss of time increases
Solution Approach 1:
The system pre-establishes expected ranges, patterns, and diagnostic criteria for sensor outputs and software results before actual operation. By having diagnostic thresholds and assessment rules pre-configured, the system can perform real-time reliability checks without time-consuming calculations during critical operation
Solution Approach 2:
The diagnostic system uses streamlined assessment methods that quickly evaluate sensor and software performance against pre-set criteria. Rather than performing exhaustive analysis, the system rapidly identifies faults by comparing current outputs against expected patterns, maintaining reliability while minimizing time loss
Data Source
AI summary
A method of on-line diagnostic and prognostic assessment of an autonomous vehicle perception system includes detecting, via a sensor, a physical parameter of an object external to the vehicle. The method also includes communicating data representing the physical parameter via the sensor to an electronic controller. The method additionally includes comparing the data from the sensor to data representing the physical parameter generated by a geo-source model. The method also includes comparing results generated by a perception software during analysis of the data from the sensor to labels representing the physical parameter from the geo-source model. Furthermore, the method includes generating a prognostic assessment of a ground truth for the physical parameter of the object using the comparisons of the sensor data to the geo-source model data and of the software results to the geo-source model labels. A system for on-line assessment of the vehicle perception system is also disclosed.