Autonomous Vehicle Trajectory Adjustment via Predictive Model Updates
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous vehicles face challenges in making effective decisions when encountering obstacles or objects that do not behave as predicted, leading to ineffective or dangerous navigation due to the inability to adjust their decision-making processes in real-time.
Innovation Solution
A system that uses sensors and processors to select and adjust vehicle trajectories based on predicted changes in other objects' trajectories, updating a model with actual changes to improve future navigation and minimize interactions with other objects.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If the vehicle uses a fixed decision-making process based on pre-programmed rules, then the system complexity is low and reliability is maintained, but the adaptability to unexpected object behaviors is insufficient
Solution Approach 1:
The system implements feedback by continuously monitoring actual object behaviors and comparing them with predicted trajectories. When deviations are detected, the feedback loop triggers model updates that adjust the decision-making process for future interactions, enabling the system to adapt to unexpected behaviors while maintaining structured complexity through defined feedback mechanisms.
Solution Approach 2:
The system performs preliminary actions by pre-calculating multiple potential trajectories and preparing adjustment strategies before interactions occur. The decision-making system pre-programs various response scenarios and object behavior models, allowing rapid adaptation when unexpected behaviors are encountered without requiring complex real-time deliberation.
2Reliability
If the vehicle adjusts trajectories frequently to account for actual object behaviors, then the safety and adaptability improve, but the loss of time for computation and navigation increases
Solution Approach 1:
The system applies partial adjustment by selectively modifying trajectories only when actual object behaviors deviate from predictions beyond threshold levels. Instead of continuously adjusting all trajectories, the system performs targeted corrections only when necessary, maintaining safety while minimizing computation time lost to unnecessary adjustments.
Solution Approach 2:
The system prepares multiple pre-calculated trajectory options and adjustment strategies in advance. When object behavior deviations are detected, the system can quickly select from pre-prepared solutions rather than computing new trajectories from scratch, reducing the time loss associated with frequent adjustments while maintaining high safety standards.
3Measurement precision
If the vehicle interacts more with other objects to learn actual behaviors, then the model accuracy improves, but the risk of collisions and harmful interactions increases
Solution Approach 1:
The system applies preliminary anti-action by predicting potential harmful interactions before they occur and adjusting trajectories proactively to prevent collisions. The decision-making process identifies objects that may not behave as predicted and pre-calculates avoidance maneuvers, allowing the vehicle to learn from actual behaviors while maintaining safety through preventive trajectory adjustments.
Solution Approach 2:
The system uses the decision-making process and trajectory adjustment mechanisms as intermediaries between learning objectives and physical interactions. Rather than directly interacting with objects to gather data, the system uses sensors to observe actual behaviors from a distance and updates models computationally, eliminating the need for dangerous physical experimentation while maintaining model accuracy.
Data Source
AI summary
Provided herein is a system of a vehicle that comprises one or more sensors, one or more processors, and memory storing instructions that, when executed by the one or more processors, causes the system to perform: selecting a trajectory along a route of the vehicle; predicting a trajectory of another object along the route; adjusting the selected trajectory based on a predicted change, in response to adjusting the selected trajectory, to the predicted trajectory of the another object, the predicted change to the predicted trajectory of the another object being stored in a model; determining an actual change, in response to adjusting the selected trajectory, to a trajectory of the another object, in response to an interaction between the vehicle and the another object; updating the model based on the determined actual change to the trajectory of the another object; and selecting a future trajectory based on the updated model.