Autonomous Driving Object Prioritization for Computational Load Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous vehicles face computational resource management challenges in classifying and prioritizing objects around them, leading to inefficient processing and increased heat generation due to the equal classification of all objects, regardless of their relevance to driving operations, especially in complex scenarios like traffic intersections.
Innovation Solution
Implementing a classification policy that categorizes objects based on their location and relevance to the vehicle's operation, such as 'track persistently' or 'track less frequently,' allowing for differential analysis frequencies to prioritize objects that significantly impact driving operations, using sensors like cameras, LiDAR, and Radars, and dynamically updating classifications based on new sensor data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If all objects are classified equally and analyzed with the same frequency, then comprehensive monitoring is achieved, but computational resources are wasted and heat generation increases
Solution Approach 1:
The patent applies local quality by differentiating the analysis frequency assigned to different objects based on their specific characteristics and relevance to the autonomous vehicle. High-priority objects (e.g., pedestrians, cyclists, vehicles in adjacent lanes) are analyzed at higher frequencies with more computational resources, while low-priority objects (e.g., distant stationary objects) are analyzed at lower frequencies, optimizing the distribution of computational resources across the monitoring system.
2Reliability
If all objects are analyzed at high frequency, then safety is improved, but device complexity and processing load increase
Solution Approach 1:
The patent segments the monitoring system into multiple priority levels, dividing objects into high-priority, medium-priority, and low-priority categories. This segmentation allows the system to apply different analysis frequencies and computational resources to different object groups, reducing the overall processing load while maintaining safety for critical objects through high-frequency analysis.
3Productivity
If differential analysis frequencies are used, then computational efficiency is improved, but classification accuracy requirements increase
Solution Approach 1:
The patent changes the parameter of analysis frequency based on object priority classification. By dynamically adjusting the analysis frequency parameter for different objects, the system achieves computational efficiency while maintaining classification accuracy through adaptive resource allocation - higher frequencies for critical objects and lower frequencies for less critical ones.
Data Source
AI summary
An autonomous vehicle can classify and prioritize agent of interest (AOI) objects located around the autonomous vehicle to manage computational resources. An example method performed by an autonomous vehicle includes determining, based on a location of the autonomous vehicle and based on a map, an area in which the autonomous vehicle is operated, determining, based on sensor data received from sensors located on or in the autonomous vehicle, attributes of objects located around the autonomous vehicle, where the attributes include information that describes a status of the objects located around the autonomous vehicle, selecting, based at least on the area, a classification policy that includes a plurality of rules that are associated with a plurality of classifications to classify the objects, and for each of the objects located around the autonomous vehicle: monitoring an object according to a classification of the object based on the classification policy.