Probabilistic Object Tracking for Autonomous Vehicles

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Autonomous vehicles face challenges in safely navigating through transportation networks due to the inability to accurately detect and predict the trajectories of static and dynamic objects, which can lead to inefficient and unsafe operation.

Innovation Solution

A method and system for world objects tracking and prediction, where an autonomous vehicle's world model module receives sensor data, associates it with objects, determines hypotheses about their intentions, calculates likelihoods, and predicts their positions, enabling the vehicle to plan a safe trajectory.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If autonomous vehicles use basic object detection methods, then the system complexity is low, but the tracking and prediction accuracy is insufficient leading to unsafe operation

Engineering Contradiction:
Improvetracking and prediction accuracyVSAvoidsystem complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The system segments the tracking and prediction task into multiple independent modules: object detection module, hypothesis generation module, likelihood calculation module, and trajectory prediction module. Each module handles a specific aspect of the problem, allowing complex functionality to be achieved through coordinated simple components rather than a single complex system.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The system dynamically generates and evaluates multiple hypotheses about object intentions and trajectories. Instead of using a static detection approach, the system continuously updates hypothesis likelihoods based on new sensor data and object behavior patterns, allowing the tracking system to adapt to changing conditions while maintaining manageable complexity through probabilistic reasoning.

Inventive Principle:
Principle #15Dynamics

2Reliability

If autonomous vehicles implement comprehensive object tracking and prediction, then navigation safety is improved, but the computational processing time increases

Engineering Contradiction:
Improvenavigation safetyVSAvoidcomputational processing time
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

The system performs preliminary actions by pre-generating multiple possible trajectory hypotheses for each detected object based on its current state and behavior patterns. This allows the system to have prediction candidates ready before they are needed for collision avoidance decisions, reducing real-time computational burden while maintaining safety through pre-computed likelihood assessments.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The system changes parameters by dynamically adjusting hypothesis likelihoods based on observed object behavior and contextual factors. Instead of computing all possible trajectories with equal weight, the system modifies probability parameters to focus computational resources on the most likely scenarios, achieving safe navigation with reduced processing time by eliminating low-probability hypotheses early in the decision pipeline.

Inventive Principle:
Principle #35Parameter changes

Data Source

PatentEP3803828B1Probabilistic object tracking and prediction framework
Publication Date: 2022.03.30 NISSAN NORTH AMERICA INC
  • EP3803828B1 patent drawingFigure 1
  • EP3803828B1 patent drawingFigure 2
  • EP3803828B1 patent drawingFigure 3

AI summary

World objects tracking and prediction by an autonomous vehicle is disclosed. A method includes receiving, from sensors of the AV, a first observation data; associating the first observation data with a first world object; determining hypotheses for the first world object, wherein a hypothesis corresponds to an intention of the first world object; determining a respective hypothesis likelihood of each of the hypotheses indicating a likelihood that the first world object follows the intention; determining, for at least one hypothesis of the hypotheses, a respective state, wherein the respective state comprises predicted positions of the first world object; and in response to a query, providing a hypothesis of the hypotheses based on the respective hypothesis likelihood of each of the hypotheses.