Quantification of Postures for Low-Height Object Manipulation Conducted by Manual Material Handlers in a Retail Environment

Exploring lumbar and lower limb kinematics and kinetics for evidence that lifting technique is associated with LBP

Purpose

To investigate if lumbar and lower limb kinematics or kinetics are different between groups with and without a history of LBP during lifting. Secondly, to investigate relationships between biomechanical variables and pain ramp during repeated lifting.

Methods

21 LBP and 20 noLBP participants completed a 100-lift task, where lumbar and lower limb kinematics and kinetics were measured during lifting, with a simultaneous report of LBP intensity every 10 lifts. Lifts were performed in a laboratory setting, limiting ecological validity.

Results

The LBP group used a different lifting technique to the noLBP group at the beginning of the task (slower and more squat-like). Kinetic differences at the beginning included less peak lumbar external anterior shear force and greater peak knee power demonstrated by the LBP group. However, at the end of the task, both groups lifted with a much more similar technique that could be classified as more stoop-like and faster. Peak knee power remained greater in the LBP group throughout and was the only kinetic difference between groups at the end of the lifting task. While both groups lifted using a more comparable technique at the end, the LBP group still demonstrated a tendency to perform a slower and more squat-like lift throughout the task. Only one of 21 variables (pelvic tilt at box lift-off), was associated with pain ramp in the LBP group. Conclusions: Workers with a history of LBP, lift with a style that is slower and more squat-like than workers without any history of LBP. Common assumptions that LBP is associated with lumbar kinematics or kinetics such as greater lumbar flexion or greater forces were not observed in this study, raising questions about the current paradigm around ‘safe lifting’.

Design, modeling, and demonstration of a new dual-mode back-assist exosuit with extension mechanism

Occupational exoskeletons and exosuits have been shown to reduce muscle demands and fatigue for physical tasks relevant to a variety of industries (e.g. logistics, construction, manufacturing, military, healthcare). However, adoption of these devices into the workforce has been slowed by practical factors related to comfort, form-factor, weight, and not interfering with movement or posture. We previously introduced an un-motorized, low-profile, dual-mode exosuit comprised of textile and elastic materials to address these adoption barriers. Here we build upon this prior work by introducing an extension mechanism that increases the moment arm of the exosuit while in engaged mode, then collapses in disengaged mode to retain key benefits related to being lightweight, low-profile, and unobstructive. Here we demonstrate both analytically and empirically how this extensible exosuit concept can (i) reduce device-to-body forces (which can improve comfort for some users and situations), or (ii) increase the magnitude of torque assistance about the low back (which may be valuable for heavy-lifting jobs) without increasing shoulder or leg forces relative to the prior form-fitting exosuit. We also introduce a novel mode-switching mechanism, as well as a human-exosuit biomechanical model to elucidate how individual design parameters affect exosuit assistance torque and device-to-body forces. The proof-of-concept prototype, case study, and modeling work provide a foundation for understanding and implementing extensible exosuits for a broad range of applications. We envision promising opportunities to apply this new dual-mode extensible exosuit concept to assist heavy-lifting, to further enhance user comfort, and to address the unique needs of last-mile delivery workers.

Motion Inference Using Sparse Inertial Sensors, Self-Supervised Learning, and a New Dataset of Unscripted Human Motion

In recent years, wearable sensors have become common, with possible applications in biomechanical monitoring, sports and fitness training, rehabilitation, assistive devices, or human-computer interaction. Our goal was to achieve accurate kinematics estimates using a small number of sensors. To accomplish this, we introduced a new dataset (the Virginia Tech Natural Motion Dataset) of full-body human motion capture using XSens MVN Link that contains more than 40 h of unscripted daily life motion in the open world. Using this dataset, we conducted self-supervised machine learning to do kinematics inference: we predicted the complete kinematics of the upper body or full body using a reduced set of sensors (3 or 4 for the upper body, 5 or 6 for the full body). We used several sequence-to-sequence (Seq2Seq) and Transformer models for motion inference. We compared the results using four different machine learning models and four different configurations of sensor placements. Our models produced mean angular errors of 10–15 degrees for both the upper body and full body, as well as worst-case errors of less than 30 degrees. The dataset and our machine learning code are freely available.