Autonomous Vehicle Trajectory Planning for Road User Reactions
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current autonomous vehicle planning systems fail to effectively incorporate the behavior predictions of other road users, leading to potential conflicts and inefficient decision-making, as they do not account for how the autonomous vehicle's actions may influence the behaviors of other agents in real-time scenarios.
Innovation Solution
The method involves dividing planned trajectories into sub-portions and analyzing each for conflicts with behavior predictions of other road users, annotating the trajectory with reactions based on predefined scenarios, and using these annotated trajectories to control the vehicle, thereby ensuring safer interactions and reducing computational complexity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If autonomous vehicle planning systems incorporate behavior predictions of other road users, then safety and interaction quality improve, but computational complexity increases
Solution Approach 1:
The patent segments the planning problem by dividing trajectories into sub-portions and processing conflicts incrementally. The system divides the road user behavior predictions into discrete conflict scenarios, allowing the vehicle to evaluate and resolve conflicts in manageable segments rather than processing all possible interactions simultaneously, thus reducing computational complexity while maintaining safety.
Solution Approach 2:
The system performs preliminary actions by pre-defining conflict scenarios and reaction types before actual conflict resolution is needed. By establishing a framework of predefined conflict categories and corresponding reaction strategies in advance, the system reduces real-time computational burden while ensuring comprehensive safety coverage when conflicts arise.
2Reliability
If the system analyzes all trajectory sub-portions for conflicts, then interaction safety improves, but processing time increases
Solution Approach 1:
The patent segments the trajectory analysis into sub-portions that are processed incrementally. By dividing the complete trajectory into smaller segments and evaluating conflicts in each segment sequentially, the system achieves thorough interaction safety verification without requiring to process the entire trajectory simultaneously, thus reducing processing time while maintaining comprehensive safety checks.
3Manufacturing precision
If the system considers multiple road user reactions, then trajectory accuracy improves, but decision-making efficiency decreases
Solution Approach 1:
The patent applies local quality by assigning specific reaction types to specific conflict scenarios based on their characteristics. Instead of evaluating all possible reactions for all scenarios, the system identifies the locally appropriate reaction type for each conflict situation (e.g., yielding, maintaining course, avoiding), thereby achieving high trajectory accuracy while maintaining decision-making efficiency through targeted rather than exhaustive analysis.
Data Source
AI summary
Aspects of the disclosure provide a method of processing trajectories. For instance, a planned trajectory may be received. A behavior prediction of the other road user may be received. For a sub portion of a plurality of sub portions of the planned trajectory, whether the sub portion conflicts with the behavior prediction of the other road user may be determined. When the sub portion is determined to conflict, whether the sub portion and the behavior prediction of the other road user meet at least one of a plurality of sets of preconditions may be determined. When the sub portion and the behavior prediction of the other road user are determined to meet at least one of the plurality of sets of preconditions, the planned trajectory may be annotated with a reaction for the other road user. The annotated planned trajectory may be used to control the autonomous vehicle.