Digital twin for predicting performance outcomes
Inventors
NICOLELLA, Daniel P. • Saylor, Kase J. • Libardoni, Mark J.
Assignees
Southwest Research Institute SwRI
Publication Number
US-11721437-B2
Publication Date
2023-08-08
Expiration Date
2039-06-02
Interested in licensing this patent?
MTEC can help explore whether this patent might be available for licensing for your application.
Abstract
A method of generating a digital twin and of using the digital twin to predict activity of an animate subject. The digital twin is generated from at least system model data and movement data. The digital twin can be activated to simulate a specified activity that the subject is performing or will perform. If desired, the subject can be instructed to perform the same activity while wearing at least one wearable sensor, which is applied to the digital twin. Using artificial intelligence techniques, the activity simulation predicts one or more physical outcomes from the activity.
Core Innovation
The invention is a method for generating a digital twin from system model data and movement data to simulate and predict the activity of an animate subject, particularly humans. This digital twin incorporates individualized neuromusculoskeletal modeling, non-invasive metabolic state monitoring, medical imaging, physics-based models, and state sensing data to create a comprehensive digital representation of the subject. Artificial intelligence techniques process data from the digital twin to predict one or more physical outcomes from specified activities.
The digital twin can be activated to simulate specific activities, which allows prediction and feedback about human performance, risk of injury, or other outcomes based on simulations and real-time sensor data. By combining data from wearables, biomarkers, medical scans, and motion capture, the method provides a multivariate analysis of an individual's current and future physical state—contrasting their data against established baselines derived from control groups to spot deviation and risk.
The problem solved by this invention is the inadequacy of current human performance monitoring, which typically relies on a limited number of preselected variables analyzed in a univariate manner. Such conventional methods fail to capture the complex and interrelated traits that contribute to human performance, often missing subtle or suboptimal patterns that may lead to performance failure or injury, especially under stress or in challenging environments. This invention addresses these limitations by offering a comprehensive, integrative system for individualized performance prediction and optimization.
Claims Coverage
There is one independent claim in the patent, and it describes a comprehensive method comprising several inventive features.
Generation of individual digital twins using internal imaging and movement data
A digital model is generated for each subject in a control group using internal imaging data, representing at least the subject's internal musculoskeletal system. Movement data is collected from each subject using a motion capture system, and these data are combined to generate a dynamic digital twin for each subject capable of simulating physical activity through muscle activation.
Performance benchmarking and identification of nominal performance using digital twins
The digital twins are activated to simulate a specified physical activity. The results from these simulations are analyzed to identify which subjects successfully complete the activity, thereby producing a subset of nominal performance digital twins and establishing nominal performance data for that activity.
Real-time subject monitoring and comparison against nominal digital twin performance
A real-time subject performs the same physical activity while wearing a wearable sensor. Actual activity data from the real-time subject are collected and compared to the nominal performance data from the control group's digital twins to determine if the real-time subject's data exceed specified thresholds.
Collectively, the inventive features establish a method for generating digital twins from imaging and movement data, establishing nominal performance metrics, and using these to monitor and evaluate the real-time performance and risk assessment of animate subjects.
Stated Advantages
Helps achieve optimal human performance by enabling control over the interactions between individual traits and processes.
Facilitates the customization of training, nutrition, rest, and recovery protocols to optimize performance and reduce injuries.
Detects subtle changes and suboptimal patterns in individual characteristics, allowing for early intervention before injury or performance failure occurs.
Provides sensitive multivariate monitoring and assessment of subject performance, supporting operational readiness and minimizing risk.
Documented Applications
Prediction and optimization of athletic performance.
Assessment and enhancement of tactical military performance.
Prediction of passive activities, such as monitoring for illness and disease.
Providing feedback for modifying training, nutrition, rest, and recovery protocols.
Interested in licensing this patent?