Vehicle Action Planning Using Occlusion-Aware Cost Simulation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional autonomous and semi-autonomous vehicle planning systems often choose the most conservative action, which may not be the safest and can lead to unnecessary traffic delays and unnatural yielding situations, failing to consider the totality of circumstances such as safety, comfort, and operational rules.
Innovation Solution
A cost-based action determination system that dynamically evaluates actions based on safety, comfort, progress, operational rules, and occlusion costs, using simulations to determine the optimal action by calculating and comparing the costs associated with each potential path or maneuver, allowing the vehicle to safely proceed through intersections without collisions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the vehicle chooses the most conservative action, then safety is improved, but traffic delays increase and passenger comfort deteriorates
Solution Approach 1:
The system changes the decision-making parameter from a fixed conservative rule to a dynamic cost function that evaluates multiple factors (safety cost, comfort cost, progress cost, operational rules cost). This allows the vehicle to select actions that optimize overall performance rather than always choosing the safest option, thereby reducing unnecessary traffic delays while maintaining adequate safety levels.
Solution Approach 2:
The planning system transitions from static conservative rules to dynamic cost-based evaluation that adapts to real-time environmental conditions. By simulating different actions and calculating their associated costs dynamically, the system can adjust its behavior based on current safety requirements, traffic conditions, and passenger comfort considerations.
2Reliability
If the vehicle chooses the most conservative action, then safety is improved, but passenger comfort deteriorates
Solution Approach 1:
The system incorporates comfort cost as a parameter in the action evaluation function, alongside safety cost. This allows the system to balance safety requirements with passenger comfort expectations, selecting actions that provide adequate safety while avoiding unnecessarily aggressive or uncomfortable maneuvers that would occur with purely conservative decision-making.
3Reliability
If the vehicle chooses the most conservative action, then safety is improved, but operational efficiency deteriorates
Solution Approach 1:
The system introduces progress cost as an evaluation parameter that considers travel time, distance to destination, and route efficiency. By balancing safety cost with progress cost, the system can select actions that maintain adequate safety while improving operational efficiency, such as proceeding through intersections when safe rather than always yielding conservatively.
4Productivity
If the vehicle evaluates multiple actions based on total cost, then operational efficiency is improved, but system complexity increases
Solution Approach 1:
The total cost function is segmented into distinct components: safety cost, comfort cost, progress cost, and operational rules cost. Each component can be calculated and evaluated independently, then combined to determine the total cost. This segmentation makes the complex evaluation process more manageable and allows for targeted optimization of individual cost components.
Solution Approach 2:
The system performs preliminary simulations of candidate actions before selecting the final action. By pre-evaluating multiple potential actions and their associated costs, the system can make more informed decisions without requiring complex real-time calculations during critical moments, thereby managing system complexity while maintaining operational efficiency.
Data Source
AI summary
A vehicle computing system may implement techniques to determine an action for a vehicle to perform based on a cost associated therewith. The cost may be based on a detected object (e.g., another vehicle, bicyclist, pedestrian, etc.) operating in the environment and/or a possible object associated with an occluded region (e.g., a blocked area in which an object may be located). The vehicle computing system may determine two or more actions the vehicle could take with respect to the detected object and/or the occluded region and may generate a simulation associated with each action. The vehicle computing system may run the simulation associated with each action to determine a safety cost, a progress cost, a comfort cost, an operational rules cost, and/or an occlusion cost associated with each action. The vehicle computing system may select the action for the vehicle to perform based on an optimal cost being associated therewith.