Autonomous Agent Conditional Policies for Complex Driving Decisions

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing autonomous vehicle systems struggle to navigate complex scenarios while mimicking human behavior and minimizing disruption to other road users, lacking a reliable method for optimizing decision-making.

Innovation Solution

Implementing multi-step conditional policies with trigger conditions that enable forward simulations to predict and optimize vehicle behavior, reducing computational load and improving predictive accuracy, allowing the vehicle to handle complex maneuvers like right-of-way conventions.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If conventional autonomous vehicle systems are configured to drive cautiously and minimize risk, then safety is improved, but the vehicle causes excessive disruption to other drivers and cannot effectively handle complex scenarios

Engineering Contradiction:

ImprovesafetyVSAvoiddisruption to other drivers

Core Design Contradiction:

ReliabilityVSEase of operation

Solution Approach 1:

The system dynamically adjusts the autonomous vehicle's behavior based on the detected scenario type. For common scenarios, the vehicle operates cautiously to minimize disruption, while for complex scenarios, it switches to more assertive behavior patterns that mimic human drivers, thereby resolving the contradiction between safety and ease of operation

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

The system changes operational parameters such as speed, acceleration, and following distance based on the scenario classification. In complex scenarios, the vehicle relaxes safety margins to match human driving behavior, while in routine scenarios, it maintains conservative parameters to reduce disruption to traffic flow

Inventive Principle:

Principle #35Parameter changes

2Measurement precision

If multi-step conditional policies with trigger conditions are implemented, then predictive accuracy and handling of complex maneuvers are improved, but computational load increases

Engineering Contradiction:

Improvepredictive accuracyVSAvoidcomputational load

Core Design Contradiction:

Measurement precisionVSUse of energy by moving object

Solution Approach 1:

The decision-making process is segmented into multiple evaluation stages: first evaluating simpler single-step policies, then progressively evaluating multi-step policies with trigger conditions only when needed. This hierarchical segmentation reduces average computational load while maintaining high predictive accuracy for complex maneuvers

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The system applies multi-step conditional policies selectively rather than universally. For routine driving situations, simpler policies suffice, but for complex maneuvers like intersection navigation or emergency situations, the full multi-step evaluation is activated, optimizing the balance between accuracy and computational resources

Inventive Principle:

Principle #16Partial or excessive action

Data Source

PatentUS12559132B2Method and system for conditional operation of an autonomous agent

Publication Date: 2026.02.24 MAY MOBILITY INC

AI summary

A method for conditional operation of an autonomous agent includes: collecting a set of inputs; processing the set of inputs; determining a set of policies for the agent; evaluating the set of policies; and operating the ego agent. A system for conditional operation of an autonomous agent includes a set of computing subsystems (equivalently referred to herein as a set of computers) and/or processing subsystems (equivalently referred to herein as a set of processors), which function to implement any or all of the processes of the method.