Augmented Path Prediction for Low-Data Vehicle Navigation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous vehicles face challenges in navigating due to the vast amounts of data they need to process and store, including visual information, GPS data, and sensor data, which can limit their navigation capabilities and require significant updates to traditional mapping technologies.
Innovation Solution
The system employs cameras to provide navigation features by analyzing images in conjunction with GPS and sensor data to identify road segments and generate planned trajectories for autonomous vehicle navigation, using a processor to adjust navigational actuators and navigate relative to planned trajectories.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional mapping technology is used for autonomous vehicle navigation, then navigation accuracy can be maintained, but data storage and processing requirements become excessively large
Solution Approach 1:
The patent extracts only the essential navigational elements from traditional mapping data - specifically road segment geometry and lane information - while eliminating redundant data. This allows the system to maintain navigation accuracy by preserving critical path information while dramatically reducing the volume of map data that needs to be stored and processed.
Solution Approach 2:
The patent segments the road environment into discrete road segments with defined geometric properties and lane structures. By dividing the continuous road network into manageable segments with standardized attributes, the system can process and store only the necessary navigational information for each segment rather than handling entire map datasets.
2Reliability
If vast volumes of sensor data are collected and processed for autonomous navigation, then navigation reliability can be improved, but system complexity and processing time increase
Solution Approach 1:
The patent performs preliminary processing of sensor data by immediately associating raw sensor measurements with pre-defined road segment models. Instead of collecting and storing all sensor data for later analysis, the system pre-processes sensor inputs by matching them against known road segment geometries and lane configurations, reducing the complexity of subsequent navigation decisions.
Solution Approach 2:
The patent creates a universal road segment model that can represent various road types and configurations using a standardized framework. This multi-functional model can accommodate different sensor inputs, road geometries, and lane structures through a single unified representation, reducing system complexity by eliminating the need for separate processing pipelines for different road scenarios.
Data Source
AI summary
A navigation system may include at least one processor programmed to receive, from a camera, one or more images representative of an environment of a host vehicle; receive a first directional indicator representative of a first target trajectory direction for the host vehicle; and receive a first distance estimate associated with the first target trajectory direction. The processor may analyze the one or more images in view of the received first directional indicator and the received first distance estimate to identify in the one or more images a representation of a first road segment consistent with the first target trajectory direction and the first distance estimate; and generate a planned trajectory for the host vehicle, based at least in part on the first target trajectory direction and the first distance estimate. The planned trajectory may be made to extend along at least a portion of the first road segment. The processor may cause at least one adjustment of a navigational actuator of the host vehicle to cause the host vehicle to navigate relative to the planned trajectory.