Experimental Study of Fully Passive, Fully Active, and Active-Passive Upper-Limb Exoskeleton Efficiency: An Assessment of Lifting Tasks

Ali Nasr, Clark R. Dickerson, John McPhee

TL;DR

This study experimentally compares the efficiency of three types of shoulder exoskeletons (fully passive, fully active, and active-passive) during lifting tasks. The active-passive exoskeleton significantly outperformed the others, allowing participants to lift for nearly twice as long before fatigue set in, highlighting its potential for industrial applications.

What Problem Does This Paper Address?

Musculoskeletal diseases, particularly in the upper extremities like the shoulders, are a significant issue for industrial workers. Robotic exoskeletons offer a promising solution to augment physical abilities and mitigate these risks. However, the optimal design and assistance strategy for these devices remain an active area of research. This paper specifically addresses the need to evaluate the efficiency of different exoskeleton assistance modes—fully passive (FP), fully active (FA), and a hybrid active-passive (AP)—during a common industrial task: lifting. Furthermore, it investigates how sex differences influence the effectiveness of these exoskeletons, an aspect often overlooked in previous studies.

Methods

The researchers conducted an experimental study with 20 healthy participants (10 male, 10 female) performing a weight-lifting task with a 5 lb kettlebell. They tested four exoskeleton configurations: inactive (IE) as a baseline, fully passive (FP), fully active (FA), and active-passive (AP).

Data was collected using a Delsys Trigno wireless system, which captured surface electromyography (sEMG) from six key muscle sites and inertial measurement unit (IMU) data for kinematics. A custom-adapted Motorized EVO upper limb exoskeleton was used, featuring a passive spring mechanism and an active BLDC motor.

The study evaluated exoskeleton efficiency using 12 criteria across three categories:

  1. sEMG Data: Measures included mean absolute value (MAV) for muscle activation, median frequency drop for fatigue, and accumulated muscle activations.
  2. Kinematic Data and Inverse Dynamic Simulation: This involved using a musculoskeletal model (MuscleNET) to estimate joint torque, power, and computational fatigue from recorded kinematics and sEMG.
  3. Subjective Feedback: A survey assessed participants' self-reported fatigue levels, areas of fatigue, and comfort with the exoskeleton.

Statistical analyses were performed using JMP software, with a focus on comparing the four exoskeleton modes and analyzing the influence of sex.

Datasets

The study utilized data collected from the 20 participants during the experimental lifting tasks. This included:

  • Surface electromyography (sEMG) signals from six muscle sites.
  • Kinematic data (Euler angles, angular acceleration) from IMUs.
  • Motor angle data from the exoskeleton's BLDC motor.
  • Participant survey responses regarding fatigue and comfort.

No pre-existing datasets were used for training or evaluation; the data was generated specifically for this study.

Key Insights and Ideas

  • Active-Passive Dominance: The active-passive (AP) exoskeleton emerged as the most efficient, significantly outperforming both fully passive (FP) and fully active (FA) modes. This suggests that a synergistic combination of passive and active assistance is highly effective for lifting tasks.
  • Sex-Specific Differences in Fatigue: Female participants generally experienced more pronounced fatigue than their male counterparts during heavy lifting, underscoring the need for sex-specific ergonomic considerations in industrial settings.
  • Anatomical and Biomechanical Variations: Females in the study tended to be shorter with shorter arm lengths, requiring less joint torque but surprisingly exhibiting faster lifting paces and higher shoulder adduction/abduction power. This highlights the importance of tailoring exoskeleton joint stiffness and control algorithms for different body types.
  • Fit Matters for Females: Several female participants expressed concerns about the fit of the exoskeleton belt, suggesting that current designs may not adequately accommodate female anatomy.

Key Results

  • Fatigue Reduction: Participants using the AP exoskeleton could hold the weighted object for nearly twice as long before fatigue occurred compared to the FP exoskeleton.
  • Efficiency Ranking: The most efficient exoskeleton types, from most to least effective, were ranked as AP, FA, FP, and IE (inactive).
  • Torque and Power: Females generally required less joint torque than males, but more shoulder adduction/abduction power, likely due to their faster lifting pace.
  • Metabolic Energy Expenditure: Females exhibited commensurately lower muscle metabolic energy expenditure (MMEE) compared to males, consistent with their generally lower body mass and stature.

Limitations and Future Work

  • Sample Size for Subjective Feedback: While the overall participant cohort was substantial for quantitative analysis, the study acknowledges that the sample size might have been too small for comprehensive subjective emotional comparisons.
  • Exoskeleton Specificity: The findings are specific to the particular AP shoulder exoskeleton model tested. A broader evaluation across various AP models could provide a more comprehensive assessment.
  • Velocity Variations: The study focused on lifting tasks. Future work could explore experiments involving different velocity settings to provide a more holistic understanding of performance.
  • Sex-Specific Controllers: The authors strongly recommend future developments focus on proposing sex-specific controllers, accommodating anthropometric and joint demand differences, and improving the kinematic fitting of wearable devices for females. This includes optimizing exoskeleton joint stiffness and tailoring control algorithms for rapid object manipulation.

Summary updated

Abstract

Recently, robotic exoskeletons are gaining attention for assisting industrial workers. The exoskeleton power source ranges from fully passive (FP) to fully active (FA), or a mixture of both. The objective of this experimental study was to assess the efficiency of a new active–passive (AP) shoulder exoskeleton using statistical analyses of 11 quantitative measures from surface electromyography (sEMG) and kinematic data and a user survey for weight lifting tasks. Two groups of females and males lifted heavy kettlebells, while a shoulder exoskeleton helped them in modes of fully passive (FP), fully active (FA), and active–passive (AP). The AP exoskeleton outperformed the FP and FA exoskeletons because the participants could hold the weighted object for nearly twice as long before fatigue occurred. Future developments should concentrate on developing sex-specific controllers as well as on better-fitting wearable devices for women.
BibTeX
@inproceedings{nasr2024experimental,
  author = {Ali Nasr and Clark R. Dickerson and John McPhee},
  title = {{Experimental Study of Fully Passive, Fully Active, and Active-Passive Upper-Limb Exoskeleton Efficiency: An Assessment of Lifting Tasks}},
  booktitle = {Sensors},
  year = {2024},
  url = {https://openreview.net/forum?id=lWTzaOsrGo},
}
Figure