Obstacle Avoidance Grouping for Multi-Object AES Trajectories
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing autonomous emergency steering (AES) systems in motor vehicles fail to activate in hazardous situations due to reliance on independent obstacle detection, leading to inactivation in the presence of multiple obstacles or when objects straddle traffic lanes, and do not account for the vehicle's surroundings as a unified space.
Innovation Solution
A method that groups detected objects based on proximity criteria, computes combined position and dynamics, and activates the AES system to determine an avoidance trajectory, ignoring traffic lane boundaries and considering the surroundings as a single space.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple objects are detected and processed independently, then each object can be analyzed individually, but the AES function becomes unavailable due to excessive number of obstacles
Solution Approach 1:
The patent combines multiple detected objects into a single group when they are located close to each other. Instead of processing each object independently, the system merges nearby objects into one collective target, reducing the computational burden and enabling the AES function to remain available in situations with multiple obstacles.
2Measurement precision
If traffic lane detection is performed to determine AES activation, then obstacle location can be assessed, but objects straddling lanes reduce the available safety margin
Solution Approach 1:
The patent extracts the AES activation decision from the traffic lane context. Instead of requiring objects to be clearly within a single lane to activate AES, the system evaluates obstacles independently of lane boundaries, allowing activation even when objects straddle multiple lanes by considering the overall spatial distribution.
3Measurement precision
If AES considers each object in isolation, then the most hazardous object can be identified, but the unified space concept is lost
Solution Approach 1:
The patent merges nearby objects into groups to evaluate them collectively while still identifying the most hazardous group. This approach maintains hazard assessment accuracy by considering the combined position and dynamics of grouped objects, while simultaneously enabling unified space reasoning that adapts to complex scenarios like pelotons of cyclists.
Data Source
AI summary
A method for object avoidance by a motor vehicle includes: detecting objects located in the surroundings of the motor vehicle; acquiring data characterizing the position and/or dynamics of each object, then, if several objects have been detected; verifying whether at least one criterion of proximity between at least two of the detected objects is met and, if so; combining the two objects into one group; calculating the data characterizing the position and/or dynamics of the group; and activating a system for obstacle avoidance and/or determining an avoidance trajectory, according to the data characterizing the position and/or dynamics of the group.