Geomagnetic Passive Navigation with AI-Enhanced Map Resolution
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing navigation systems, particularly those relying on Global Navigation Satellite Systems (GNSS), are vulnerable to signal obstructions, drop-outs, erroneous data, jamming, and spoofing, which can lead to lost or inaccurate navigation, especially in airborne and maritime systems.
Innovation Solution
The use of geomagnetic-aided passive navigation, which involves generating high-resolution geomagnetic maps using machine learning techniques such as Generative Adversarial Networks (GANs) to enhance spatial resolution and provide independent navigation capabilities.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of stationary object
If satellite-based geomagnetic surveys are used to create navigation maps, then coverage area is large, but spatial resolution is insufficient to capture fine localized anomalies
Solution Approach 1:
The patent segments the survey process into two distinct components: satellite-based surveys for broad area coverage and aerial-based surveys for high-resolution detailed mapping. This segmentation allows each method to be optimized for its specific purpose, with satellite data providing extensive geographic coverage and aerial data providing the fine spatial resolution needed for navigation anomalies.
Solution Approach 2:
The patent merges satellite-based geomagnetic map data with aerial-based survey data to create a composite navigation map. The system integrates these different data sources, combining the broad coverage of satellite surveys with the high spatial resolution of aerial surveys to produce a unified map that has both extensive area coverage and fine detail for navigation.
2Measurement precision
If aerial surveys are performed to obtain high spatial resolution, then navigation accuracy improves, but cost and complexity increase significantly
Solution Approach 1:
The patent segments the survey methodology into satellite-based and aerial-based approaches, allowing the system to use satellite data for general navigation needs and only deploy expensive aerial surveys in specific areas where high-resolution anomaly detection is critical for navigation safety.
Solution Approach 2:
The patent creates a multi-functional navigation map that serves multiple purposes: it provides both broad-area navigation guidance from satellite data and high-precision anomaly detection from aerial data, making the system versatile for different navigation requirements without requiring separate systems.
3Reliability
If passive geomagnetic navigation is used, then immunity to jamming and spoofing improves, but navigation accuracy may be insufficient without high-resolution maps
Solution Approach 1:
The patent merges satellite-based and aerial-based geomagnetic survey data into a unified navigation map that provides both the reliability of passive geomagnetic navigation and the precision needed for accurate position determination. This combination enables the system to maintain attack immunity while achieving sufficient navigation accuracy.
Solution Approach 2:
The patent creates a composite navigation map that integrates data from multiple survey sources (satellite and aerial), analogous to composite materials that combine different substances to achieve properties that neither substance alone could provide. The resulting map has both the reliability of passive navigation and the precision of high-resolution mapping.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach provides greater immunity to outages and attacks by offering a passive navigation method that is independent of weather and time-of-day, with enhanced spatial resolution improving navigation accuracy and reliability.
Implementation Method 1
a machine learning approach can be used such as to synthesize geomagnetic maps having enhanced resolution versus lower resolution survey data. For example, a Generative Adversarial Network (GAN) neural network topology can form a generator neural network.
Implementation Method 2
Localization techniques using the Earth's magnetic field can provide an alternative or augmentation to other navigation approaches
Implementation Method 3
an indicium of a position of a vehicle on an artificially-generated geomagnetic map can be used along with other sensor data to provide an enhanced position estimate (or more generally, a state variable estimate) using a particle filtering technique
Data Source
AI summary
A machine learning approach can be used such as to synthesize geomagnetic maps having enhanced resolution versus lower resolution survey data. An on-board magnetometer can be used to measure a local magnetic field intensity, and a measured magnetic field intensity can be compared to the enhanced-resolution geomagnetic map. An indicium of a position of a vehicle on the enhanced-resolution geomagnetic map can be used, along with other sensor data, to provide an enhanced position estimate (or more generally, a state variable estimate) using a particle filtering technique supported by a deep reinforcement learning approach. Such a “geomagnetic-aided navigation” approach can be robust and need not rely on other navigational aids such as Global Navigation Satellite System (GNSS) or terrestrial transmitters.