Autonomous Vehicle Object Prediction With Priority-Based Processing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous vehicles face challenges in efficiently prioritizing and predicting the future states of objects in their environment, leading to delayed reaction times and potential safety risks due to the need to analyze all objects simultaneously before determining a motion plan.
Innovation Solution
A system that classifies objects as high-priority or low-priority based on their state data, allowing for the prediction of future states for high-priority objects first, using high-fidelity prediction systems for critical objects and low-fidelity systems for less critical ones, thereby reducing overall latency and enhancing safety.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the autonomous vehicle analyzes all objects simultaneously before determining a motion plan, then the analysis is comprehensive, but the reaction time is delayed
Solution Approach 1:
The patent segments the set of all perceived objects into multiple priority groups (first priority, second priority, third priority, etc.) based on their relevance to the autonomous vehicle. This segmentation allows the system to process objects in hierarchical order rather than simultaneously analyzing all objects, thereby reducing overall processing time while maintaining comprehensive analysis through multi-level priority assessment.
Solution Approach 2:
The patent performs preliminary classification of objects into priority groups before the actual motion plan determination. By pre-categorizing objects based on their state data and relevance to the vehicle, the system prepares the analysis structure in advance, enabling faster execution of the motion planning process without compromising the comprehensiveness of the analysis.
2Ease of operation
If the system processes all objects with equal computational resources, then the treatment is uniform, but the critical objects are not prioritized
Solution Approach 1:
The patent applies local quality by assigning different processing priorities and computational resource allocations to different objects based on their specific characteristics and relevance to the autonomous vehicle. High-priority objects receive more immediate and intensive processing, while lower-priority objects are processed with reduced resources, ensuring that critical safety-related objects are handled with appropriate urgency and attention.
3Measurement precision
If the autonomous vehicle uses high-fidelity prediction for all objects, then the prediction accuracy is maximized, but the computational load increases
Solution Approach 1:
The patent dynamically changes the fidelity parameter of prediction based on object priority. For high-priority objects, the system employs high-fidelity prediction models that provide maximum accuracy. For lower-priority objects, the system switches to lower-fidelity models that consume fewer computational resources. This parameter adjustment allows the system to optimize the balance between prediction accuracy and computational load according to the specific needs of each object.
Data Source
Figure 1
Figure 2~3
Figure 4
AI summary
Systems and methods for determining object prioritization and predicting future object locations for an autonomous vehicle are provided. A method can include obtaining, by a computing system comprising one or more processors, state data descriptive of at least a current or past state of a plurality of objects that are perceived by an autonomous vehicle. The method can further include determining, by the computing system, a priority classification for each object in the plurality of objects based at least in part on the respective state data for each object. The method can further include determining, by the computing system, an order at which the computing system determines a predicted future state for each object based at least in part on the priority classification for each object and determining, by the computing system, the predicted future state for each object based at least in part on the determined order.