Vehicle Path Processing Error Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous vehicles face challenges in accurately determining when to transition from autonomous to manual control due to deviations from predicted steerable paths, as existing systems lack robustness in detecting errors and path faults, leading to potential vehicle deviation from the travel path.
Innovation Solution
A system that identifies errors between predicted steerable and lane paths using different protocols, with a security measure to enhance data robustness, applies low-pass filters, and transitions control based on error thresholds and elapsed time to prevent path faults, ensuring safe operation by switching to semi-autonomous or manual mode when necessary.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the vehicle uses a predicted steerable path based on standard data collection protocols, then the system operates with standard functionality, but the system lacks robustness in detecting errors and path faults
Solution Approach 1:
The patent applies preliminary action by collecting lane path data using a second secure protocol in advance to establish a reference path before the vehicle operates. This pre-collected secure data is then used to verify the steerable path derived from first protocol data, enabling early detection of path faults without adding complexity to the real-time operation. The secure lane path serves as a pre-established benchmark for validating the operational path.
2Reliability
If the system transitions to manual mode immediately when error thresholds are exceeded, then safety is prioritized, but false alarms may occur due to transient errors
Solution Approach 1:
The patent applies dynamics by making the mode transition decision dynamic rather than static. Instead of immediately switching to manual mode when an error threshold is exceeded, the system continuously monitors the error between steerable and lane paths over time. The decision to transition is made dynamically based on whether the error persists beyond a threshold for a specified duration, allowing the system to adapt to transient errors while maintaining safety for persistent faults.
Solution Approach 2:
The patent implements feedback by continuously comparing the steerable path with the secure lane path and using this comparison feedback to determine mode transitions. The system monitors the error signal over time and uses this feedback loop to decide whether to maintain autonomous mode or switch to manual mode, ensuring that transitions are based on sustained errors rather than transient anomalies.
3Ease of operation
If the system uses only a predicted steerable path for autonomous operation, then the system is simpler to operate, but the vehicle may deviate from the intended travel path
Solution Approach 1:
The patent uses an intermediary approach by introducing a secure lane path derived from second protocol data as a mediator between the simple steerable path and the ground truth travel path. The lane path acts as an intermediate reference that validates whether the steerable path remains accurate, allowing the system to operate simply using steerable path data while periodically verifying accuracy against the intermediary lane path to detect deviations.
Data Source
AI summary
A computer includes a processor and a memory, the memory storing instructions executable by the processor to identify an error between a predicted steerable path of a vehicle based on data collected according to a first protocol and a predicted lane path based on data collected according to a second protocol and to identify a path fault when the error exceeds an error threshold for an elapsed time exceeding a time threshold.