AI Sequence Positioning from Earth Signals to Correct Navigation Drift
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
GNSS signals are susceptible to jamming and spoofing, and inertial measurement units suffer from geographic position drift over long periods, while navigation filters using single magnetic measurements can generate erroneous geographic positions due to similar magnetic anomaly field strengths across adjacent locations.
Innovation Solution
Utilize a sequence of Earth characteristic values, such as magnetic anomaly, gravity, or terrain height, measured over a path, to train an artificial intelligence (AI) to determine the vehicle's estimated spatial position, eliminating the need for expensive trajectory databases and improving positional accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a single magnetic anomaly measurement is used to determine geographic location, then the determination process is simple and fast, but the accuracy is poor because magnetic field strength may be substantially the same for two or more adjacent geographic locations
Solution Approach 1:
The patent transitions from using a single magnetic anomaly measurement (one-dimensional data point) to using a sequence of magnetic anomaly measurements over time (temporal dimension added). This allows the system to track changes in magnetic field characteristics as the vehicle moves, enabling more accurate location determination even when individual measurements are ambiguous due to similar field strengths at adjacent locations.
2Reliability
If inertial measurement unit is used to determine geographic position, then GNSS jamming and spoofing are avoided, but geographic position drift occurs over long periods of time
Solution Approach 1:
The patent implements a feedback mechanism where magnetic anomaly measurements are continuously taken and compared against a database of expected magnetic field patterns at known locations. This feedback loop allows the system to correct inertial drift by periodically recalibrating the position estimate based on magnetic landmark recognition, maintaining both reliability under jamming and long-term accuracy.
3Productivity
If navigation filter uses single magnetic measurement to determine geographic location, then processing is computationally simple, but erroneous geographic positions are generated due to similar magnetic anomaly field strengths across adjacent locations
Solution Approach 1:
The patent combines multiple magnetic anomaly measurements taken at different times into a single integrated analysis. By merging this sequence of measurements with the vehicle's known motion trajectory from the inertial measurement unit, the system creates a more robust location estimate that leverages both the temporal sequence of magnetic data and the kinematic constraints of vehicle motion.
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
The AI-based method enhances positional accuracy by leveraging a sequence of Earth characteristics, reducing errors associated with single magnetic measurements and drift, even in the absence of reliable GNSS signals.
Implementation Method 1
a magnetometer, which is configured to obtain by measurement a magnetic anomaly that varies with geographic location on or about a surface of the Earth
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A sequence, of values of a characteristic of the Earth, is received. Each value is obtained from a measurement, that varies with spatial position with respect to a reference. Using the sequence of values, an estimated spatial position with respect to the reference, is determined, at a time, with an artificial intelligence (Al), e.g., a generative AI (for example, an AI sequence model). The time may be a past or a current time.