AI-Based Magnetometer Calibration for Mobile Orientation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Magnetometer measurements in mobile devices are prone to significant errors due to external and internal magnetic interferences, which hinder precise orientation determination, especially in navigation applications, and are exacerbated in low-cost magnetometers.
Innovation Solution
A method that utilizes a mobile support equipped with a magnetometer and movement/position sensors, such as accelerometers and GNSS/GPS receivers, to infer corrected magnetometric measurements through an artificial intelligence algorithm trained on simultaneous data, allowing for the detection and elimination of disturbances and potential recalibration of the magnetometer.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If magnetometer measurements are used directly in mobile devices, then the device can determine orientation, but measurement errors attain several degrees due to magnetic interferences
Solution Approach 1:
The patent introduces an intermediary processing system that includes: (1) collecting magnetometric measurements from the magnetometer, (2) collecting movement and position measurements from accelerometers, gyrometers, and GNSS receivers, (3) processing these measurements through algorithms to detect magnetic interferences, and (4) correcting the magnetometric measurements based on the detected interferences and complementary data from other sensors. This intermediary processing chain acts as a mediator between the raw magnetometer measurements and the final orientation determination, filtering out harmful magnetic interferences.
Solution Approach 2:
The patent merges data from multiple sensors (magnetometer, accelerometers, gyrometers, GNSS receivers) into a unified processing system. By combining magnetometric measurements with movement and position measurements from other sensors, the system creates a more robust orientation determination that is less susceptible to magnetic interferences alone. The fusion of multiple data sources allows the system to cross-validate and correct measurements.
2Ease of manufacture
If low-cost magnetometers are used in mobile terminals, then device cost is reduced, but measurement errors are exacerbated due to weaker sensor performance and greater susceptibility to disturbances
Solution Approach 1:
The patent implements feedback mechanisms where the processing system continuously monitors magnetometric measurements alongside movement and position data from other sensors. When magnetic interferences are detected (through analysis of measurement patterns, sudden deviations, or inconsistency with expected motion dynamics), the system automatically applies corrections. This feedback loop enables low-cost magnetometers to achieve acceptable precision by dynamically compensating for their inherent weaknesses.
Solution Approach 2:
The system performs self-calibration and self-correction by using its own multi-sensor data to identify and remedy measurement errors. The processing unit analyzes the combined sensor data to detect when the magnetometer is affected by interferences and automatically corrects its readings without external intervention, enabling low-cost sensors to function adequately.
3Measurement precision
If magnetometer measurements are corrected using multiple sensors and processing algorithms, then measurement precision is improved, but device complexity increases
Solution Approach 1:
The patent makes the processing system multi-functional by designing it to simultaneously: (1) collect and process magnetometric measurements for orientation, (2) collect and process movement measurements for motion detection, (3) collect and process position measurements for location tracking, (4) detect magnetic interferences, and (5) correct magnetometric measurements. This universal processing architecture handles multiple sensor types and multiple objectives through a single integrated system, reducing the need for separate dedicated processing chains for each function.
4Ease of operation
If local disturbances (magnets, metallic masses, electromagnetic sources) are present, then magnetometric measurements are distorted, but these disturbances cannot be easily identified or eliminated
Solution Approach 1:
The patent performs preliminary detection and characterization of magnetic disturbances by continuously analyzing magnetometric measurements alongside movement and position data before final orientation calculation. The system identifies patterns indicative of local disturbances (such as sudden magnetic field changes inconsistent with motion, or persistent deviations) and flags them for correction. This preliminary action allows the system to prepare correction strategies in advance rather than reacting after errors have already degraded orientation accuracy.
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 significantly reduces measurement errors by correlating the orientation of the mobile device with respect to the magnetic field and its movements, enabling more accurate orientation determination and calibration, thereby enhancing the precision of magnetometric readings.
Implementation Method 1
The measurements of the magnetometer can serve to indicate the direction of the magnetic North, or, after correction for the magnetic declination, the direction of the geographic North
Implementation Method 2
the processing unit (36) comprises one or more microprocessors and is capable of inferring the corrected magnetometric measurements (28) via an artificial intelligence algorithm (24)
Data Source
AI summary
An aspect of the invention relates to a method for correcting magnetometric measurements (10) made by a magnetometer mounted on a mobile support. The mobile support carries one or more movement and/or position sensors fixedly mounted with respect to the magnetometer. The method comprises:obtaining magnetometric measurements (12) and movement and/or position measurements (14, 16) made simultaneously by the magnetometers, respectively the movement and/or position sensors during a time interval;inferring, by a processing unit, corrected magnetometric measurements (28) on the basis of the magnetometric measurements (12) and the movement and/or position measurements (14, 16) made simultaneously during the time interval and in which method the processing unit (36) comprises one or more microprocessors and is capable of inferring the corrected magnetometric measurements (28) via an artificial intelligence algorithm (24), said algorithm being trained by means of training data, to find a correction for the magnetometric measurements according to the log of magnetometric measurements as well as movement and/or position measurements recorded during the time interval.