Multibody Motion Tracking With ML-Informed Physics Models
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing IMU-based motion measurement systems for multibody systems are complex, require knowledge of initial conditions, and need multiple IMUs, making them difficult for untrained users and limiting their use to controlled environments.
Innovation Solution
A method and system combining inertial measurement units with machine learning-informed physics models and numerical optimization to calculate joint positions, velocities, and accelerations, using a minimal set of IMUs, without requiring initial conditions or a calibrated environment.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a full set of IMUs is attached to each body segment, then measurement precision is improved, but device complexity increases
Solution Approach 1:
The system segments the measurement function by separating the inertial sensing (performed by minimal IMUs) from the computational processing (performed by the machine learning model). This allows accurate motion measurement to be achieved without requiring IMUs on every body segment, as the machine learning model infers the motion of uninstrumented segments from the limited IMU data.
2Measurement precision
If filtering algorithms are used to convert IMU measurements, then measurement precision is improved, but device complexity increases
Solution Approach 1:
The patent replaces traditional mechanical filtering algorithms (such as Kalman filters or extended Kalman filters) with a machine learning model. This substitution maintains the ability to accurately estimate joint positions, velocities, and accelerations from noisy IMU data while simplifying the computational approach and reducing the need for complex mathematical modeling and initial condition specifications.
3Measurement precision
If initial conditions are required for IMU orientations and positions, then measurement precision is improved, but ease of operation deteriorates
Solution Approach 1:
The machine learning model performs self-calibration by learning the relationship between IMU measurements and joint kinematics directly from data, eliminating the need for users to manually specify initial conditions for IMU orientations, positions, and velocities. The system automatically adapts to the specific multibody system being measured, making it easy for untrained users to operate without requiring knowledge of kinematic models or initial state specifications.
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
Enables accurate motion measurement of multibody systems in any environment, reducing complexity and cost, and allowing untrained users to easily measure motion with high accuracy.
Implementation Method 1
inertial measurement units (IMUs) to measure the motion of a multibody system
Implementation Method 2
inputting the measured IMU data time histories into a machine learning model to calculate machine learning model predictions of joint positions, velocities, and accelerations
Implementation Method 3
determining optimal values for the plurality of joint position design variables by optimizing the guess for the plurality of joint position design variables using a cost function
Data Source
AI summary
A computer-implemented method for calculating multibody motion data includes obtaining measured IMU data time histories with a plurality of inertial measurement units and using computer-based algorithm incorporating a machine-learning informed kinetic model to generate the multibody motion data. A system for generating motion data from a multibody system includes include a plurality of inertial measurement units, each configured to be placeable on a corresponding segment of the multibody system and a computer system comprising a processor configured to implement the computer-based algorithm.