Vehicle Driving Control Using Predicted Route Risk Zones
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing autonomous driving systems lack effective methods to predict and respond to potential dangers on curved roads or with small turning radii, particularly in real-time, by concentrating computing power on high-risk zones based on predicted vehicle routes.
Innovation Solution
A method and apparatus that estimate a vehicle's predicted route using odometry and location information, identify high-risk zones based on road map data, and generate driving control information through object detection, utilizing neural networks to process images from vehicle sensors, thereby enhancing driver assistance systems.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If the vehicle travels on sharply curved roads or with small turning radii, then the vehicle can navigate complex road geometries, but the system fails to accurately predict dangerous situations and respond in real-time
Solution Approach 1:
The system performs preliminary route prediction by estimating the vehicle's future trajectory based on current odometry and location data before actual dangerous situations occur. This allows the system to proactively identify high-risk zones ahead of time and prepare appropriate responses, rather than reacting only after dangers materialize.
Solution Approach 2:
The system segments the driving environment by identifying and focusing computational resources on specific high-risk zones along the predicted route, rather than uniformly processing all areas. This segmentation allows the system to maintain high reliability in danger detection for critical areas while managing computational complexity.
2Reliability
If the system processes object detection for the entire scene, then comprehensive monitoring is achieved, but computing power is wasted on low-risk areas
Solution Approach 1:
The system applies local quality by differentiating processing intensity across different spatial zones. High-risk zones identified through route prediction receive intensive object detection and monitoring resources, while low-risk areas receive minimal or no processing. This ensures comprehensive monitoring of critical areas without wasting computing power on safe zones.
Solution Approach 2:
The system performs partial action by conducting exhaustive object detection only in identified high-risk zones rather than across the entire scene. This selective approach provides sufficient monitoring coverage for dangerous areas while significantly reducing overall computational power consumption.
3Reliability
If the system responds to all detected objects, then safety coverage is maximized, but response time increases due to processing overhead
Solution Approach 1:
The system performs preliminary identification of high-risk zones based on route prediction before actual object detection and response generation. This pre-filtering step allows the system to prepare for potential dangers in advance, so when objects are detected in these pre-identified zones, the response time is minimized because the system is already in a state of alert and preparation.
Data Source
AI summary
Provided is a method and apparatus with driving control. A method includes estimating a predicted route of a vehicle based on odometry information of the vehicle and location information of the vehicle, based on road map information, determining a target zone corresponding to the predicted route, and generating driving control information of the vehicle based on an object detection result for the determined target zone.