Intersection Obstacle Intent Prediction for Preempt-or-Yield Decisions

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Autonomous driving systems face challenges in accurately predicting the intent of moving obstacles, leading to delays in decision-making and increased risk of collisions due to delayed response and imperfect perception of motion status and future behavior.

Innovation Solution

A method and apparatus for determining the intent of a target obstacle by analyzing its historical motion status to predict whether it will preempt or yield at an intersection, using probability distributions and limit values to make advance decisions, thereby improving safety and accuracy.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If the autonomous driving system relies on real-time detection and prediction of moving obstacles, then the system can respond to current situations, but there is a delay in signal tracking and decision-making when facing interaction scenarios

Engineering Contradiction:
ImprovesafetyVSAvoidresponse delay
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

The system performs preliminary classification of target obstacles into preempting and yielding types based on historical motion status before the autonomous vehicle reaches the intersection. This advance classification allows the decision-making system to prepare appropriate responses in advance, eliminating delays when the vehicle actually needs to react at the intersection point.

Inventive Principle:
Principle #10Preliminary action

2Measurement precision

If the system uses traditional detection and prediction methods for moving obstacles, then it can track current motion status, but it cannot accurately identify the future behavior intent of target obstacles

Engineering Contradiction:
Improvemotion status detectionVSAvoidintent information
Core Design Contradiction:
Measurement precisionVSLoss of information

Solution Approach 1:

The system classifies target obstacles into preempting or yielding types in advance based on their historical motion status patterns. By analyzing historical data before the interaction scenario occurs, the system extracts intent information that would otherwise be lost, enabling accurate prediction of future behavior rather than just tracking current motion status.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent introduces an intermediary classification mechanism that acts as a bridge between raw motion status detection and future behavior prediction. This intermediary layer analyzes historical motion patterns to infer intent, converting basic motion data into meaningful intent information about whether the target will preempt or yield.

Inventive Principle:
Principle #24Intermediary (Mediator)

3Measurement precision

If the autonomous driving system waits until the obstacle is close to make decisions, then it can ensure accurate detection, but it cannot make advance decisions for safe maneuvering

Engineering Contradiction:
Improvedetection accuracyVSAvoiddecision-making time
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The system performs intent classification in advance when the target obstacle is still far away, using historical motion status data that is available earlier in the interaction timeline. This preliminary classification enables the vehicle to make advance decisions about safe maneuvering, rather than waiting until the obstacle is close when detection accuracy is highest but decision time is insufficient.

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentEP4691874A1Method and apparatus for determining target intention
Publication Date: 2026.02.11 YINWANG INTELLIGENT TECHNOLOGIES CO LTD
  • EP4691874A1 patent drawingFigure 1
  • EP4691874A1 patent drawingFigure 2
  • EP4691874A1 patent drawingFigure 3(a)~3(b)

AI summary

A method and an apparatus for determining an intent of a target are provided. The method may be applied to the field of intelligent driving. The method includes: determining, based on a historical motion status of a target obstacle, a probability distribution of a motion status in which the target obstacle preempts to pass through an intersection point and a probability distribution of a motion status in which the target obstacle yields to pass through the intersection point; and determining, in advance based on a current motion status of the target obstacle, the probability distribution of the motion status in which the target obstacle preempts to pass through the intersection point, and the probability distribution of the motion status in which the target obstacle yields to pass through the intersection point, whether the target obstacle preempts or yields to pass through the intersection point. Embodiments of this application can be applied to an intelligent vehicle or an electric vehicle. Before the intelligent vehicle or the electric vehicle intersects with the target obstacle, an intent of preempting or yielding of the obstacle can be determined in advance, so that a correct decision and plan can be made in advance, thereby greatly improving safety of intelligent driving.