Machine-Learned Movement Intent Detection for Mobility Augmentation

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Conventional mobility aids such as crutches, canes, wheelchairs, exoskeletons, and functional electrical stimulation (FES) are not personalized to the user and fail to optimize mobility augmentation based on available information, limiting their effectiveness for individuals with disabilities.

Innovation Solution

A mobility augmentation system using machine learning to monitor muscle electroactivity, kinematics, and kinetics to predict intended movements, applying personalized actuation strategies and optimizing them through real-time feedback loops.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If conventional mobility aids (crutches, canes, wheelchairs, exoskeletons, FES) are used, then mobility assistance is provided, but the aids are not personalized to the user and fail to optimize mobility augmentation

Engineering Contradiction:

Improvepersonalization to userVSAvoideffectiveness of mobility augmentation

Core Design Contradiction:

Adaptability or versatilityVSReliability

Solution Approach 1:

The system performs preliminary actions by collecting and analyzing motor intent data (EMG, IMU, foot pressure) to predict intended movements before the user actually executes them. This allows the mobility augmentation system to prepare and apply actuation signals in advance, enabling personalized and timely assistance that adapts to the user's specific movement intentions rather than reacting generically to observed movement.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system implements continuous feedback loops where monitored movement data is compared against target movement data, and the machine learning model is re-trained based on this feedback. This closed-loop approach allows the system to learn from the user's actual movements and optimize actuation strategies over time, achieving personalization and improved effectiveness through iterative adaptation.

Inventive Principle:

Principle #23Feedback

2Measurement precision

If machine learning models are trained on generalized population data, then mobility prediction capability is established, but the system lacks personalization to individual users

Engineering Contradiction:

Improvemovement prediction accuracyVSAvoidcustomization to individual user

Core Design Contradiction:

Measurement precisionVSAdaptability or versatility

Solution Approach 1:

The system performs preliminary training on generalized population data to establish baseline movement prediction capabilities before deployment. This preliminary action provides a solid foundation that can later be personalized through feedback from individual user data, allowing the system to quickly adapt to specific users while leveraging population-level patterns.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The machine learning model transitions from a static generalized model to a dynamic personalized model through continuous re-training with user-specific feedback data. This dynamic adaptation allows the system to evolve its prediction accuracy for each individual user over time, balancing the benefits of population-level generalization with user-specific customization.

Inventive Principle:

Principle #15Dynamics

3Measurement precision

If the system monitors multiple data types (EMG, IMU, foot pressure) to predict movement intent, then prediction accuracy improves, but system complexity increases

Engineering Contradiction:

Improvemovement intent prediction accuracyVSAvoidsensor and processing complexity

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The system merges multiple sensor data types (EMG from muscle electroactivity, IMU from inertial measurements, and foot pressure signals) into a unified motor intent prediction framework. By combining these complementary data sources, the system achieves higher prediction accuracy for intended movements while managing complexity through integrated processing that leverages the strengths of each sensor modality.

Inventive Principle:

Principle #5Merging (Combining)

Data Source

PatentUS20250275858A1Machine-learned movement determination based on intent identification

Publication Date: 2025.09.04 CIONIC INC

AI summary

A mobility augmentation system monitors data representative of a user's motor intent and augments the user's mobility based on the monitored motor intent data. A machine-learned model is trained to identify an intended movement based on the monitored motor intent data. The machine-learned model may be trained based on generalized or specific motor intent data (e.g., user-specific motor intent data). A machine-learned model initially trained on generalized motor intent data may be re-trained on user-specific motor intent data such that the machine-learned model is optimized to the movements of the user. The system uses the machine-learned model to identify a difference between the user's monitored movement and target movement signals. Based on the identified difference, the system determines actuation signals to augment the user's movement. The actuation signals determined can be an adjustment to a currently applied actuation such that the system optimizes the actuation strategy during application.