Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Estimating Human Muscular Fatigue in Dynamic Collaborative Robotic Tasks with Learning-Based Models

Feras Kiki, Pouya P. Niaz, Alireza Madani, Cagatay Basdogan

TL;DR

This work tackles estimating human muscular fatigue in dynamic, repetitive pHRI tasks using arm sEMG signals. It frames fatigue progression as a regression problem to predict the fraction of cycles to fatigue (FCF) with subject-specific regression models (Random Forest, XGBoost, Linear Regression) and a CNN trained on EMG spectrograms. Results show CNN achieving the lowest RMSE (~$20.8\%$) while tree-based models approach that performance, and cross-task validation indicates robustness to movement direction and muscle recruitment, enabling fatigue monitoring without task-specific retraining. The findings support fatigue-aware shared autonomy in pHRI and provide a public EMG fatigue dataset to facilitate future research.

Abstract

Assessing human muscle fatigue is critical for optimizing performance and safety in physical human-robot interaction(pHRI). This work presents a data-driven framework to estimate fatigue in dynamic, cyclic pHRI using arm-mounted surface electromyography(sEMG). Subject-specific machine-learning regression models(Random Forest, XGBoost, and Linear Regression predict the fraction of cycles to fatigue(FCF) from three frequency-domain and one time-domain EMG features, and are benchmarked against a convolutional neural network(CNN) that ingests spectrograms of filtered EMG. Framing fatigue estimation as regression (rather than classification) captures continuous progression toward fatigue, supporting earlier detection, timely intervention, and adaptive robot control. In experiments with ten participants, a collaborative robot under admittance control guided repetitive lateral (left-right) end-effector motions until muscular fatigue. Average FCF RMSE across participants was 20.8+/-4.3% for the CNN, 23.3+/-3.8% for Random Forest, 24.8+/-4.5% for XGBoost, and 26.9+/-6.1% for Linear Regression. To probe cross-task generalization, one participant additionally performed unseen vertical (up-down) and circular repetitions; models trained only on lateral data were tested directly and largely retained accuracy, indicating robustness to changes in movement direction, arm kinematics, and muscle recruitment, while Linear Regression deteriorated. Overall, the study shows that both feature-based ML and spectrogram-based DL can estimate remaining work capacity during repetitive pHRI, with the CNN delivering the lowest error and the tree-based models close behind. The reported transfer to new motion patterns suggests potential for practical fatigue monitoring without retraining for every task, improving operator protection and enabling fatigue-aware shared autonomy, for safer fatigue-adaptive pHRI control.

Estimating Human Muscular Fatigue in Dynamic Collaborative Robotic Tasks with Learning-Based Models

TL;DR

This work tackles estimating human muscular fatigue in dynamic, repetitive pHRI tasks using arm sEMG signals. It frames fatigue progression as a regression problem to predict the fraction of cycles to fatigue (FCF) with subject-specific regression models (Random Forest, XGBoost, Linear Regression) and a CNN trained on EMG spectrograms. Results show CNN achieving the lowest RMSE (~) while tree-based models approach that performance, and cross-task validation indicates robustness to movement direction and muscle recruitment, enabling fatigue monitoring without task-specific retraining. The findings support fatigue-aware shared autonomy in pHRI and provide a public EMG fatigue dataset to facilitate future research.

Abstract

Assessing human muscle fatigue is critical for optimizing performance and safety in physical human-robot interaction(pHRI). This work presents a data-driven framework to estimate fatigue in dynamic, cyclic pHRI using arm-mounted surface electromyography(sEMG). Subject-specific machine-learning regression models(Random Forest, XGBoost, and Linear Regression predict the fraction of cycles to fatigue(FCF) from three frequency-domain and one time-domain EMG features, and are benchmarked against a convolutional neural network(CNN) that ingests spectrograms of filtered EMG. Framing fatigue estimation as regression (rather than classification) captures continuous progression toward fatigue, supporting earlier detection, timely intervention, and adaptive robot control. In experiments with ten participants, a collaborative robot under admittance control guided repetitive lateral (left-right) end-effector motions until muscular fatigue. Average FCF RMSE across participants was 20.8+/-4.3% for the CNN, 23.3+/-3.8% for Random Forest, 24.8+/-4.5% for XGBoost, and 26.9+/-6.1% for Linear Regression. To probe cross-task generalization, one participant additionally performed unseen vertical (up-down) and circular repetitions; models trained only on lateral data were tested directly and largely retained accuracy, indicating robustness to changes in movement direction, arm kinematics, and muscle recruitment, while Linear Regression deteriorated. Overall, the study shows that both feature-based ML and spectrogram-based DL can estimate remaining work capacity during repetitive pHRI, with the CNN delivering the lowest error and the tree-based models close behind. The reported transfer to new motion patterns suggests potential for practical fatigue monitoring without retraining for every task, improving operator protection and enabling fatigue-aware shared autonomy, for safer fatigue-adaptive pHRI control.
Paper Structure (20 sections, 1 equation, 6 figures)

This paper contains 20 sections, 1 equation, 6 figures.

Table of Contents

  1. Introduction
  2. Related Literature
  3. Approach
  4. Hardware Setup
  5. Closed Loop Control Architecture
  6. Experiments
  7. Training Experiments
  8. Participants
  9. Subjective Rating of Muscle Fatigue
  10. Machine Learning
  11. Data Processing
  12. Modeling
  13. Deep Learning
  14. Cross-Task Generalization
  15. Results
  16. ...and 5 more sections

Figures (6)

  • Figure 1: Our approach to estimating human muscular fatigue in cyclic pHRI tasks employs learning-based regression models that predict the fraction of cycles to fatigue (FCF) from sEMG measurements.
  • Figure 2: The participants of our pHRI experiments were asked to perform cyclic and continuous movements along the X-axis. The end-effector position (first row), human force (second row), and surface EMG signals recorded from the LD (third row) and PD (fourth row) muscles until the onset of fatigue for a representative participant. The vertical dotted lines in the plots are used to separate the cycles.
  • Figure 3: Relative change in the maximum values of MNF, MDF, TP, and RMS amplitude with respect to the initial cycle plotted against the fraction of cycle to fatigue (FCF). Each row represents one feature, while the left and right columns are for the data collected from the LD and PD muscles. In each plot, a dot represents a feature for a cycle, and the data of each participant is represented by a unique color. The dashed lines in each plot represent the individual linear fits for each participant, while the solid line depicts the linear fit for the average across all participants to illustrate the general trend.
  • Figure 4: a. Subjective level of fatigue perceived by the participants (SRF) was evaluated using Borg CR10 score. b. The fraction of cycles to fatigue (FCF) versus SRF. Each dot in the plot represents an SRF value of a participant recorded for a cycle. Data for each participant is represented by a unique color in the plot. The dashed lines represent the individual linear fits for each participant, while the solid line depicts the linear fit for the average across all participants to illustrate the general trend.
  • Figure 5: Bar plots illustrating the mean RMSE values of each participant for the primary task, averaged across all trials, for the different regression models.
  • ...and 1 more figures