Normalized Route Comparison for Autonomous Vehicles
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing systems face challenges in comparing and managing routes for autonomous vehicles (AVs) due to differences in vehicle capabilities, sensor data, and route data formats, leading to inefficiencies in service assignment and route optimization.
Innovation Solution
A route comparison engine generates normalized route data by selecting reference positions and traversing routes to create sets of corresponding normalized positions, allowing for qualitative and quantitative comparison of route similarity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If different AVs are routed based on their specific capabilities and sensor data, then route optimization and vehicle utilization improve, but route comparison complexity and data normalization requirements increase
Solution Approach 1:
The patent transforms route data from multiple different formats and coordinate systems into a unified normalized format with consistent parameters. This includes converting various coordinate representations to a standard coordinate system, normalizing route point densities, and standardizing vehicle capability descriptors. By changing the parameter representation across all inputs to a common framework, the system enables direct comparison while maintaining the ability to account for different vehicle capabilities and sensor configurations.
Solution Approach 2:
The patent introduces a route normalization engine as an intermediary component between route generation and route comparison. This normalization engine acts as a mediator that receives diverse route data from different vehicles with different capabilities, processes and standardizes the data into a common format, and then outputs normalized route data suitable for comparison. This intermediary layer isolates the complexity of data normalization from the comparison logic, allowing each to be optimized independently.
2Measurement precision
If route data is normalized to enable comparison, then route similarity assessment improves, but data processing time and computational resources increase
Solution Approach 1:
The patent extracts and separates the normalization process into distinct, independent steps that can be performed selectively. Rather than fully normalizing all route data regardless of need, the system extracts only the necessary normalization operations required for specific comparison tasks. This allows the system to perform minimal normalization when full precision is not required, reducing processing time while maintaining adequate measurement precision for the given context.
Solution Approach 2:
The patent divides the route normalization process into multiple sequential stages: coordinate system conversion, route point sampling and density normalization, capability descriptor standardization, and similarity metric computation. This segmentation allows the system to process different aspects of route data independently and at different levels of detail, enabling optimization of processing time by adjusting the depth of normalization applied at each stage based on requirements.
3Reliability
If comprehensive route comparison is performed to determine equivalent routes, then service assignment accuracy improves, but system computational load and processing overhead increase
Solution Approach 1:
The patent implements a multi-level comparison approach where the system performs partial comparisons first using key discriminative features, and only performs full comprehensive comparisons when needed. The comparison process can be stopped early if sufficient differentiation is achieved, or extended to full precision only when route equivalence cannot be determined with partial data. This allows the system to achieve adequate service assignment accuracy for most cases with reduced computational load, while maintaining the option for comprehensive comparison when required.
Data Source
AI summary
Various examples are directed to systems and methods for routing an autonomous vehicle. A comparison engine may access first route data describing a first route and second route data describing a second route. The comparison engine may generate first normalized route data using the first route data, where the first normalized route data comprises a first plurality of normalized positions along the first route. The comparison engine may also generate second normalized route data using the second route data, where the second normalized route data comprises a second plurality of normalized positions along the second route. The route comparison engine may compare a first position from a first plurality of normalized positions with a first corresponding position from the second plurality of normalized positions and may instruct an autonomous vehicle to begin executing an autonomous vehicle route based at least in part on the comparing.