Autonomous Vehicle Behavior Prediction and Control

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Autonomous vehicles face challenges in predicting the behavior of other vehicles in their environment and determining the likelihood of these behaviors, which can impact safe navigation and control.

Innovation Solution

A method and system that use a computer system to determine the current state of a vehicle and its environment, predict the behavior of other vehicles, calculate a confidence level for these predictions, and adjust vehicle control accordingly, incorporating sensor data and potentially server-based analysis.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If autonomous vehicles use sensor data and computer systems to predict behavior of other vehicles, then safety and collision avoidance improve, but system complexity and computational requirements increase

Engineering Contradiction:
ImprovesafetyVSAvoidsystem complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The system segments behavior prediction into multiple discrete components: detecting current states of other vehicles, generating multiple possible trajectory segments, evaluating confidence levels for each segment, and selecting optimal paths. This modular approach manages complexity while improving safety through comprehensive analysis.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The system performs preliminary actions by pre-calculating multiple possible trajectories and confidence levels before actual navigation decisions are required. This advance preparation reduces real-time computational burden while maintaining high safety standards through thorough预先 analysis of potential vehicle behaviors.

Inventive Principle:
Principle #10Preliminary action

2Measurement precision

If the system calculates confidence levels for predicted behaviors of other vehicles, then navigation accuracy improves, but computational time and processing requirements increase

Engineering Contradiction:
Improveprediction accuracyVSAvoidcomputational time
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The system applies partial action by calculating confidence levels selectively for the most relevant trajectories and other vehicles, rather than exhaustively analyzing all possible scenarios. This approach maintains high prediction accuracy for critical decisions while reducing unnecessary computational time spent on low-probability events.

Inventive Principle:
Principle #16Partial or excessive action

Solution Approach 2:

The system uses feedback mechanisms where confidence level calculations from previous time steps inform and optimize current predictions. This iterative approach improves measurement precision over time while reducing computational time by leveraging learned patterns from historical data to prioritize current analysis.

Inventive Principle:
Principle #23Feedback

Data Source

PatentEP2825435B1Modifying behavior of autonomous vehicle based on predicted behavior of other vehicles
Publication Date: 2019.12.04 WAYMO LLC
  • EP2825435B1 patent drawingFigure 1
  • EP2825435B1 patent drawingFigure 2
  • EP2825435B1 patent drawingFigure 3A

AI summary

A vehicle configured to operate in an autonomous mode could determine a current state of the vehicle and the current state of the environment of the vehicle. The environment of the vehicle includes at least one other vehicle. A predicted behavior of the at least one other vehicle could be determined based on the current state of the vehicle and the current state of the environment of the vehicle. A confidence level could also be determined based on the predicted behavior, the current state of the vehicle, and the current state of the environment of the vehicle. In some embodiments, the confidence level may be related to the likelihood of the at least one other vehicle to perform the predicted behavior. The vehicle in the autonomous mode could be controlled based on the predicted behavior, the confidence level, and the current state of the vehicle and its environment.