Lane Change Gap Ranking With Rule-Based Early Filtering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for gap selection in autonomous vehicle lane changes are complex, making it difficult for humans to interpret scores and troubleshoot algorithms, and they often require full evaluation of all gaps, which is inefficient.
Innovation Solution
A gap evaluation system that uses a processor and memory to receive sensor data, generate profiles for potential gaps, apply gap rules with threshold values, and generate a prioritized list of gaps for evaluation, allowing for early termination of non-ideal gap evaluations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If a sum of weighted averages method is used to evaluate gaps, then many factors can be accounted for, but the score becomes difficult to interpret and troubleshoot
Solution Approach 1:
The gap evaluation process is segmented into multiple discrete stages: initial filtering based on critical thresholds, intermediate scoring of remaining gaps, and final selection. This segmentation allows each stage to focus on specific factors, making the overall process more interpretable while still considering multiple characteristics.
Solution Approach 2:
The patent applies preliminary filtering actions by establishing minimum threshold requirements for critical gap characteristics before full evaluation. Gaps that fail these preliminary checks are eliminated early, simplifying the interpretation of final scores since only gaps meeting basic requirements are fully evaluated and scored.
2Measurement precision
If all gaps are fully evaluated, then the best gap can be selected, but computing resources and time are wasted on unacceptable gaps
Solution Approach 1:
The system performs preliminary evaluations of gap characteristics against minimum thresholds before conducting full detailed evaluations. This preliminary action identifies and eliminates unacceptable gaps early in the process, preventing waste of computing resources on gaps that would ultimately be rejected.
Solution Approach 2:
The patent applies partial evaluation to gaps by assessing only critical characteristics initially, rather than evaluating all possible characteristics for every gap. This partial action approach suffices to identify unacceptable gaps, avoiding the excessive computation that would result from full evaluation of all gaps.
3Measurement precision
If full evaluation of all gaps is performed, then the best gap can be selected, but computing resources are consumed inefficiently
Solution Approach 1:
The system performs preliminary filtering based on critical gap characteristics before full evaluation. This preliminary action consumes minimal computing resources while eliminating clearly unacceptable gaps, thereby reducing overall computational energy requirements while maintaining accurate selection of viable gaps.
Solution Approach 2:
The patent applies partial evaluation to only those gaps that pass initial threshold checks, rather than performing complete evaluations on all gaps. This partial action approach maintains sufficient accuracy for safe lane change decisions while significantly reducing computing resource consumption.
Data Source
AI summary
A gap evaluation system for evaluating potential gaps for lane change operations of an autonomous vehicle, wherein the system is configured to receive, from the plurality of sensors, sensor data corresponding to a plurality of potential gaps, transmit sensor data to the at least one processor, generate a profile for each potential gap using the sensor data, retrieve a plurality of gap rules from the at least one memory, wherein each gap rule is directed to a characteristic of a potential gap and wherein each gap rule comprises a threshold value corresponding to the characteristic, evaluate each gap against the plurality of gap rules by comparing the profile for each potential gap to the threshold value of each gap rule, and generate a prioritized list of gaps based on the evaluation of the plurality of potential gaps.