Vehicle Collision Prevention via Trajectory Overlap Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current advanced driver assistance systems (ADAS) face challenges in effectively preventing collisions due to limitations in tracking and predicting the trajectories of nearby objects, especially under conditions of poor visibility or driver distraction, which can lead to accidents.
Innovation Solution
A method and apparatus that utilize sensors to track objects, determine their trajectories, verify collision levels by assessing trajectory overlap and probability of following paths, and perform collision prevention operations by changing vehicle paths or informing drivers of collision risks based on time to collision (TTC) and field of view analysis.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If ADAS systems use basic sensor tracking to monitor nearby objects, then the system complexity is low, but the collision prediction accuracy deteriorates under poor visibility or driver distraction conditions
Solution Approach 1:
The system segments the monitoring task by dividing objects into different categories (pedestrians, vehicles, cyclists) and applying object-specific trajectory prediction models to each category, improving prediction accuracy without proportionally increasing overall system complexity
Solution Approach 2:
The system performs preliminary trajectory prediction and collision risk assessment in advance by continuously tracking multiple future positions of detected objects, allowing the vehicle to prepare collision avoidance maneuvers before actual collision risk materializes
2Measurement precision
If the system determines trajectories based on moving path and road curvature, then the collision prediction accuracy improves, but the computational complexity increases
Solution Approach 1:
The system applies different levels of trajectory prediction complexity to different objects based on their characteristics - using simple linear extrapolation for stationary objects while applying complex curvature-based prediction only to moving objects on curved road sections
Solution Approach 2:
The system calculates full trajectory predictions only for objects that pose potential collision risk, while using simplified monitoring for objects clearly outside the hazard zone, avoiding unnecessary computational overhead
3Reliability
If the system performs comprehensive collision verification by checking trajectory overlap and following probability, then the collision detection reliability improves, but the processing time increases
Solution Approach 1:
The system performs comprehensive collision verification at specific periodic intervals rather than continuously, updating trajectory predictions and collision assessments at predetermined time frames to balance detection reliability with processing efficiency
Solution Approach 2:
The system uses the trajectory data already collected for basic monitoring purposes to simultaneously perform collision verification, eliminating the need for separate dedicated verification data collection and reducing overall processing time
4Reliability
If the system changes vehicle path based on collision prevention, then the collision avoidance effectiveness improves, but the deviation from original route increases
Solution Approach 1:
The system dynamically adjusts the degree of path deviation based on real-time collision risk assessment - applying minimal necessary deviation to avoid collisions while maintaining efficiency, rather than using fixed conservative avoidance margins
Solution Approach 2:
Instead of planning a completely alternative safe route when collision risk is detected, the system makes localized inverse adjustments to the original path - deviating only enough to clear the hazard while returning to the original route as quickly as possible
Data Source
AI summary
Disclosed are collision preventing apparatuses and methods for a vehicle. The collision preventing apparatus estimate whether objects located in a vicinity of the vehicle may collide with each other based on data collected from a sensor in the vehicle, determine a trajectory of each of the objects, verify whether the estimated trajectories intersect, determine that the objects may collide with each other in response to the estimated trajectories overlapping, inform each of the objects of a collision risk when the objects may collide with each other, and differently perform an operation of informing the objects of the collision risk based on a collision level of each of the objects.