Covariance Voxel Map Alignment for Low-Overlap Sensor Fusion

Solution Overview

Problem

Integrating sensor data into map data for autonomous vehicles is time-consuming and resource-intensive, particularly when merging voxel spaces with little overlap, which can require significant computational resources and struggle with initial convergence.

Innovation Solution

A system that applies a search algorithm to localize voxel spaces using covariances, ensuring sufficient overlap before merging, and utilizes residual values and principal component analysis to align and merge voxel spaces efficiently.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If sensor data is integrated into map data by merging voxel spaces, then the completeness of environmental representation is improved, but the processing time and computational resources increase significantly

Engineering Contradiction:

Improvecompleteness of environmental representationVSAvoidprocessing time

Core Design Contradiction:

ReliabilityVSLoss of time

Solution Approach 1:

The system performs preliminary alignment of sensor data with map data before merging voxel spaces. By establishing the correct spatial relationship between sensor data and map data in advance, the system avoids time-consuming realignment operations during the merging process, thus reducing overall processing time while maintaining complete environmental representation.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system changes the parameter of spatial transformation parameters (translation and rotation) to optimize the alignment between sensor data and map data. By iteratively adjusting these parameters to minimize residual values, the system achieves accurate alignment faster, reducing processing time while ensuring complete and accurate environmental representation.

Inventive Principle:

Principle #35Parameter changes

2Reliability

If sensor data is integrated into map data by merging voxel spaces, then the completeness of environmental representation is improved, but the computational resources increase significantly

Engineering Contradiction:

Improvecompleteness of environmental representationVSAvoidcomputational resources

Core Design Contradiction:

ReliabilityVSUse of energy by moving object

Solution Approach 1:

The system performs preliminary alignment of sensor data with map data before merging voxel spaces. By establishing the correct spatial relationship between sensor data and map data in advance, the system avoids time-consuming realignment operations during the merging process, thus reducing overall processing time while maintaining complete environmental representation.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system changes the parameter of spatial transformation parameters (translation and rotation) to optimize the alignment between sensor data and map data. By iteratively adjusting these parameters to minimize residual values, the system achieves accurate alignment faster, reducing processing time while ensuring complete and accurate environmental representation.

Inventive Principle:

Principle #35Parameter changes

3Area of stationary object

If voxel spaces with little overlap are merged, then the coverage of environmental representation is improved, but the difficulty of alignment and convergence increases

Engineering Contradiction:

Improvecoverage of environmental representationVSAvoiddifficulty of alignment and convergence

Core Design Contradiction:

Area of stationary objectVSDifficulty of detecting and measuring

Solution Approach 1:

The system changes the parameter of spatial transformation parameters (translation and rotation) to optimize the alignment between sensor data and map data. By iteratively adjusting these parameters to minimize residual values, the system achieves accurate alignment faster, reducing processing time while ensuring complete and accurate environmental representation.

Inventive Principle:

Principle #35Parameter changes

Solution Approach 2:

The system replaces traditional mechanical alignment methods with a covariance-based alignment approach. By using covariance matrices to quantify the spatial relationship and uncertainty between voxel spaces, the system can effectively align even voxel spaces with little overlap, reducing the difficulty of alignment and convergence while maintaining comprehensive environmental coverage.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Data Source

PatentUS12475580B1System for aligning sensor data with maps comprising covariances

Publication Date: 2025.11.18 ZOOX INC

AI summary

Techniques for integrating sensor data into a scene or map based on statistical data of captured environmental data are discussed herein. The data may be stored as a multi-resolution voxel space and the techniques may comprise first applying a pre-alignment or localization technique prior to fully integrating the sensor data.