GNSS-Based 3D Environment Model Generation for Autonomous Vehicles

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Current 3D environment models for autonomous vehicles are either unreliable or expensive to acquire, especially for larger areas, due to varying accuracy and incomplete data in open street maps, and existing methods for generating more accurate models from GNSS measurements are not cost-effective or freely available.

Innovation Solution

A method that uses GNSS measurements to classify data based on line of sight and generate wall objects, allowing for the creation of three-dimensional environment models, particularly for autonomous vehicles, by receiving and processing measurement data from a global satellite navigation system, classifying it with respect to direct or indirect lines of sight, and generating wall models using optimization algorithms.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If open street maps are used for 3D environment models, then the models are freely available, but the accuracy and completeness are insufficient

Engineering Contradiction:

Improveaccuracy of 3D environment modelVSAvoidcost of acquiring 3D environment model

Core Design Contradiction:

ReliabilityVSEase of manufacture

Solution Approach 1:

The system uses freely available GNSS measurement data from standard receivers to automatically generate 3D environment models, making the process self-sufficient without requiring expensive external data sources or professional surveying equipment

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

The patent replaces expensive mechanical surveying systems (laser scanners, aerial cameras) with GNSS-based electronic measurement and classification systems, achieving comparable or superior results through signal processing and machine learning algorithms

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

2Measurement precision

If expensive accurate 3D models are acquired from aerial images or laser gauges, then the accuracy is improved, but the cost increases greatly

Engineering Contradiction:

Improveaccuracy of wall element positioningVSAvoidcost of acquiring 3D environment model

Core Design Contradiction:

Measurement precisionVSEase of manufacture

Solution Approach 1:

The system uses inexpensive GNSS receivers and freely available satellite navigation data to create accurate 3D models, replacing expensive specialized surveying equipment with affordable consumer-grade technology

Inventive Principle:

Principle #27Cheap short-living objects (Disposable)

Solution Approach 2:

The patent transforms GNSS measurement parameters (signal strength, time of flight, satellite positions) into spatial information about wall elements through classification algorithms, extracting geometric data from non-geometric measurement parameters

Inventive Principle:

Principle #35Parameter changes

3Reliability

If GNSS measurement data is used directly without classification, then the processing is simple, but the reliability of line of sight determination is insufficient

Engineering Contradiction:

Improveline of sight classification accuracyVSAvoidcomplexity of data processing

Core Design Contradiction:

ReliabilityVSDevice complexity

Solution Approach 1:

The patent segments GNSS measurement data into distinct categories (line of sight, reflected, blocked) based on signal characteristics, enabling targeted processing for each type and improving overall classification reliability

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The system uses feedback from signal strength measurements, satellite geometry, and environmental context to iteratively refine line of sight classifications, improving accuracy through multiple passes of validation and correction

Inventive Principle:

Principle #23Feedback

Data Source

PatentUS20230324563A1Method for Creating an Environment Model

Publication Date: 2023.10.12 ROBERT BOSCH GMBH

AI summary

A method for creating an environment model includes receiving measurement data from a satellite navigation system, classifying the measurement data with respect to a line of sight, and generating wall objects.