Map Data Fusion via Spatial Kalman Filtering

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing map data systems for autonomous vehicles face challenges such as calibration issues, bias, and noise in crowdsourced maps, lack of lane line attributes in telemetry-based maps, and high costs and outdatedness in aerial and high-definition maps, which hinder their accuracy and reliability.

Innovation Solution

A system that fuses two or more versions of map data using spatial Kalman filtering, where central computers process road network data, map data points, and ground truth data to estimate state vectors representing lane markings, thereby creating more accurate and precise fused map data.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If crowdsourced map data is used, then real-time information and lane marking construction are provided, but calibration issues and GPS bias/noise reduce accuracy

Engineering Contradiction:

Improvereal-time map update speedVSAvoidmap data accuracy

Core Design Contradiction:

ProductivityVSMeasurement precision

Solution Approach 1:

The patent combines multiple map data sources (crowdsourced map data, aerial map data, and high-definition map data) into a unified fused map data product. By merging these diverse sources with different characteristics, the system achieves both real-time update capability from crowdsourced data and high accuracy from survey vehicles, resolving the contradiction between productivity and measurement precision.

Inventive Principle:

Principle #5Merging (Combining)

2Ease of operation

If telemetry-based map data is used, then high confidence level and ease of procurement are achieved, but lane line attributes are not provided

Engineering Contradiction:

Improveease of map data procurementVSAvoidlane line attribute information

Core Design Contradiction:

Ease of operationVSLoss of information

Solution Approach 1:

The system merges telemetry-based map data (which provides high confidence and ease of procurement) with aerial and high-definition map data (which provide lane line attributes). This combination allows the fused map data to inherit the reliability of telemetry data while supplementing it with the attribute information from other sources, resolving the information loss problem.

Inventive Principle:

Principle #5Merging (Combining)

3Measurement precision

If aerial map data is used, then high precision is provided, but lane line attributes are inconsistent and creation is time-consuming

Engineering Contradiction:

Improvemap data precisionVSAvoidmap data creation time

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

The patent combines aerial map data (high precision but time-consuming) with crowdsourced map data (fast updates but lower precision) and high-definition map data (complete attributes but expensive and slow to update). The fusion process leverages the precision of aerial data while using crowdsourced data for timely updates, resolving the time loss issue.

Inventive Principle:

Principle #5Merging (Combining)

4Measurement precision

If high-definition map data is used, then high precision and lane line attributes are provided, but update time is excessive

Engineering Contradiction:

Improvemap data precisionVSAvoidmap data update speed

Core Design Contradiction:

Measurement precisionVSProductivity

Solution Approach 1:

The system merges high-definition map data (high precision and complete attributes but slow to update) with crowdsourced map data (fast updates). The fusion allows the system to maintain the high precision and attribute completeness of high-definition data while incorporating the rapid update capability of crowdsourced data, resolving the productivity contradiction.

Inventive Principle:

Principle #5Merging (Combining)

5Adaptability or versatility

If cascaded layered architecture is used, then alternative map versions are provided, but integration of different map versions is not achieved

Engineering Contradiction:

Improvemap data source flexibilityVSAvoidintegrated map information

Core Design Contradiction:

Adaptability or versatilityVSLoss of information

Solution Approach 1:

The patent implements a fusion process that actively integrates multiple map data versions using a discrete random curve model and Kalman filtering, rather than merely providing alternative versions. This merging approach combines the strengths of different sources into a unified fused map data product, achieving both adaptability and information integration.

Inventive Principle:

Principle #5Merging (Combining)

Data Source

PatentUS20250102320A1System for fusing two or more versions of map data based on spatial kalman filtering

Publication Date: 2025.03.27 GM GLOBAL TECHNOLOGY OPERATIONS LLC

AI summary

A system for fusing two or more versions of map data together includes one or more central computers that receive road network data representing a road network for a predefined geofenced area. The central computers receive road network data that includes a discrete random curve that represents lane markings. The discrete random curve includes a plurality of state vectors that are each defined by a respective location and tangent angle. The central computers estimate the position for the state vectors of the discrete random curve based on a signed distance and the tangent angle by minimizing a spatial Kalman filter cost function and execute a Kalman smoothening function to estimate the position and the tangent angle for the state vectors that are part of the discrete random curve, where the state vectors each represent a map point of the fused map data.