Vehicle Responsibility Allocation for Autonomous Navigation Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous and semi-autonomous vehicles face challenges in navigating safely due to overly cautious behaviors caused by worst-case assumptions and static assumptions about human driver behaviors, leading to unnecessary maneuvers and inefficient operations.
Innovation Solution
A neural network is trained using real-world data to determine responsibility allocations between vehicles and other objects, allowing machines to adjust their controls based on actual motion patterns and account for unreasonable maneuvers, thereby improving navigation efficiency and safety.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If worst-case assumptions are used to model other driver behaviors, then safety is improved, but the autonomous vehicle behaves overly cautiously and makes unnecessary maneuvers
Solution Approach 1:
The system dynamically adjusts the responsibility allocation for the autonomous vehicle based on real-time assessment of other drivers' behaviors and circumstances. Instead of using fixed worst-case assumptions, the responsibility level is continuously updated to reflect actual conditions, allowing the vehicle to transition between cautious and efficient behaviors as appropriate.
Solution Approach 2:
The system changes the parameter of responsibility allocation from a static worst-case value to a dynamic value determined by machine learning models. These models analyze multiple factors including driver behavior patterns, environmental conditions, and vehicle states to adjust the responsibility parameter, enabling the vehicle to operate efficiently while maintaining safety.
2Device complexity
If static assumptions are used to model other driver behaviors, then system complexity is reduced, but the system cannot account for unreasonable or unexpected maneuvers
Solution Approach 1:
The system replaces simple static assumption models with machine learning-based behavioral models that can dynamically predict other drivers' actions. These models process sensor data and historical behavior patterns to anticipate unreasonable or unexpected maneuvers, significantly improving safety prediction capabilities while maintaining computational efficiency through optimized model architectures.
Data Source
AI summary
In various examples, learning responsibility allocations for machine interactions is described herein. Systems and methods are disclosed that train a neural network(s) to generate outputs indicating estimated levels of responsibilities associated with interactions between vehicles or machines and other objects (e.g., other vehicles, machines, pedestrians, animals, etc.). In some examples, the neural network(s) is trained using real-world data, such as data representing scenes depicting actual interactions between vehicles and objects and/or parameters (e.g., velocities, positions, directions, etc.) associated with the interactions. Then, in practice, a vehicle (e.g., an autonomous vehicle, a semi-autonomous vehicle, etc.) may use the neural network(s) to generate an output indicating a proposed or estimated level of responsibility associated with an interaction between the vehicle and an object. The vehicle may then use the output to determine one or more controls for the vehicle to use when navigating.