Vehicle Sensor Reliability Assessment Using External References
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Determining reliable sensor information for autonomous vehicles is challenging, especially in dynamic environments, as sensors can become decalibrated due to vibrations or harsh conditions, and existing fault detection algorithms may not suffice for systemic faults, necessitating recalibration without stopping the vehicle.
Innovation Solution
An autonomous vehicle system that utilizes external sensors, such as roadside units and other vehicles, to assess reliability factors like device type, age, distance, and integrity levels, dynamically calibrating internal sensors based on the most reliable external information, and implementing fault mitigation strategies if recalibration is insufficient.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If sensors are used to detect environmental parameters in autonomous vehicles, then measurement capability is improved, but reliability deteriorates due to decalibration from vibrations and harsh conditions
Solution Approach 1:
The patent introduces an intermediary system that receives sensor data from multiple sources (including external devices like roadside units and other vehicles) and determines reliability factors to identify the most trustworthy sensor values. This intermediary layer mediates between the unreliable individual sensors and the decision-making system, using reliability assessment algorithms to weigh and select the most accurate environmental parameter readings without requiring physical intervention or stopping the vehicle.
2Productivity
If dynamic calibration is performed without stopping the vehicle, then productivity is improved, but measurement precision may deteriorate due to ongoing vibrations and harsh conditions
Solution Approach 1:
The system performs preliminary actions by continuously assessing reliability factors and identifying the most reliable sensor values before making calibration adjustments. By pre-evaluating the trustworthiness of external and internal sensor data sources and selecting the optimal reference values in advance, the system can perform calibration computations without stopping the vehicle, maintaining productivity while minimizing the impact of ongoing harsh conditions on precision.
3Reliability
If multiple reliability factors are assessed from external and internal sources, then reliability determination is improved, but device complexity increases
Solution Approach 1:
The patent implements a universal reliability assessment mechanism that handles multiple data sources (external devices and internal sensors) through a single integrated process. The system uses a unified approach to evaluate reliability factors across different sensor types and sources, applying consistent algorithms to determine trustworthiness. This multi-functional design allows the same core mechanism to process diverse input sources without requiring separate complex subsystems for each data type, thereby improving reliability determination while controlling overall system complexity.
Data Source
AI summary
An environment perception method, at an ego vehicle, includes: receiving, from a device external to the ego vehicle, one or more first reliability factors associated with the device and with a first value of a parameter of an environment; determining, based on the one or more first reliability factors, a first reliability of the first value; determining, based on one or more second reliability factors associated with the ego vehicle and with a second value of the parameter of the environment provided by one or more sensors of the ego vehicle, a second reliability of the second value; and providing a reliability indication indicating which of the first value and the second value is more reliable based on which of the first reliability and the second reliability is a higher reliability.