@article{wang_xie_lu_su_jung_xu_2022, title={A mobile platform-based app to assist undergraduate learning of human kinematics in biomechanics courses}, volume={142}, ISSN={["1873-2380"]}, DOI={10.1016/j.jbiomech.2022.111243}, abstractNote={Whole-body biomechanics examines different physical characteristics of the human body movement by applying principles of Newtonian mechanics. Therefore, undergraduate biomechanics courses are highly demanding in mathematics and physics. While the inclusion of laboratory experiences can augment student comprehension of biomechanics concepts, the cost and the required expertise associated with experiment equipment can be a burden of offering laboratory sessions. In this study, we developed a mobile app to facilitate learning human kinematics in biomechanics curriculums. First, a mobile-based computer-vision algorithm that is based on Convolutional pose machine (CPM), MobileNet V2, and TensorFlow Lite framework is adopted to reconstruct 2D human poses from the images collected by a mobile device camera. Key joint locations are then applied to the human kinematics variable estimator for human kinematics analysis. Simultaneously, students can view various kinematics data for a selected joint or body segment in real-time through the user interface of the mobile device. The proposed app can serve as a potential instructional tool to assist in conducting human motion experiments in biomechanics courses.}, journal={JOURNAL OF BIOMECHANICS}, author={Wang, Hanwen and Xie, Ziyang and Lu, Lu and Su, Bingyi and Jung, Sehee and Xu, Xu}, year={2022}, month={Sep} }
 @misc{lu_xie_wang_li_xu_2022, title={Mental stress and safety awareness during human-robot collaboration - Review}, volume={105}, ISSN={["1872-9126"]}, url={https://doi.org/10.1016/j.apergo.2022.103832}, DOI={10.1016/j.apergo.2022.103832}, abstractNote={Human-robot collaboration (HRC) is an emerging research area that has gained tremendous attention in both academia and industry. Yet, the feature that humans and robots sharing the workplace has led to safety concerns. In particular, the mental stress or safety awareness of human teammates during HRC remains unclear but is also of great importance to workplace safety. In this manuscript, we reviewed twenty-five studies for understanding the relationships between HRC and workers' mental stress or safety awareness. Specifically, we aimed to understand: (1) robot-related factors that may affect human workers' mental stress or safety awareness, (2) a number of measurements that could be used to evaluate workers' mental stress in HRC, and (3) various methods for measuring safety awareness that had been adopted or could be applied in HRC. According to our literature review, robot-related factors including robot characteristics, social touching and trajectory have relationships with workers' mental stress or safety awareness. For the measurement of mental stress and safety awareness, each method mentioned has its validity and rationality. Additionally, a discussion related to the potential co-robot actions to lower mental stress or improve safety awareness as well as future implications were provided.}, journal={APPLIED ERGONOMICS}, publisher={Elsevier BV}, author={Lu, Lu and Xie, Ziyang and Wang, Hanwen and Li, Li and Xu, Xu}, year={2022}, month={Nov} }
 @article{wang_xie_lu_li_xu_2021, title={A computer-vision method to estimate joint angles and L5/S1 moments during lifting tasks through a single camera}, volume={129}, ISSN={["1873-2380"]}, DOI={10.1016/j.jbiomech.2021.110860}, abstractNote={Weight lifting is a risk factor of work-related low-back musculoskeletal disorders (MSD). From the ergonomics perspective, it is important to measure workers' body motion during a lifting task and estimate low-back joint moments to ensure the low-back biomechanical loadings are within the failure tolerance. With the recent development of advanced deep neural networks, an increasing number of computer vision algorithms have been presented to estimate 3D human poses through videos. In this study, we first performed a 3D pose estimation of lifting tasks using a single RGB camera and VideoPose3D, an open-source library with a fully convolutional model. Joint angle trajectories and L5/S1 joint moment were then calculated following a top-down inverse dynamic biomechanical model. To evaluate the accuracy of the computer-vision-based angular trajectories and L5/S1 joint moments, we conducted an experiment in which participants performed a variety of lifting tasks. The body motions of the participants were concurrently captured by an RGB camera and a laboratory-grade motion tracking system. The body joint angles and L5/S1 joint moments obtained from the camera were compared with those obtained from the motion tracking system. The results showed a strong correlation (r > 0.9, RMSE < 10°) between the two methods for shoulder flexion, trunk flexion, trunk rotation, and elbow flexion. The computer-vision-based method also yielded a good estimate for the total L5/S1 moment and the L5/S1 moment in the sagittal plane (r > 0.9, RMSE < 20 N·m). This study showed computer vision could facilitate safety practitioners to quickly identify the jobs with high MSD risks through field survey videos.}, journal={JOURNAL OF BIOMECHANICS}, author={Wang, Hanwen and Xie, Ziyang and Lu, Lu and Li, Li and Xu, Xu}, year={2021}, month={Dec} }
 @article{li_prabhu_xie_wang_lu_xu_2021, title={Lifting Posture Prediction With Generative Models for Improving Occupational Safety}, volume={51}, ISSN={["2168-2305"]}, url={https://doi.org/10.1109/THMS.2021.3102511}, DOI={10.1109/THMS.2021.3102511}, abstractNote={Lifting tasks have been identified to be highly associated with work-related low back pain. Posture prediction can be used for simulating workers’ posture of lifting tasks and thus facilitate the prevention of low back pain (LBP). This study adopts two generative models, conditional variational encoder and conditional generative adversarial network, to predict lifting postures. A regular feed-forward neural network (FNN) developed upon previous studies is also investigated for comparison purposes. Ground-truth lifting posture data collected by a motion capture system is used for training and testing the models. The models are trained with datasets of different size and loss functions, and the results are compared. The conditional variational autoencoder and the regular FNN achieved comparable top performance in lifting posture prediction in terms of accuracy and posture validity. Both generative models are able to partially capture the variability of constrained postures. Overall, the results prove that using a generative model is able to predict postures with reasonable accuracy and validity (RMSE of coordinates = 0.049 m; RMSE of joint angles = 19.58 <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$^\circ$</tex-math></inline-formula> ). The predicted postures can support biomechanical analysis and ergonomics assessment of a lifting task to reduce the risk of low back injuries.}, number={5}, journal={IEEE TRANSACTIONS ON HUMAN-MACHINE SYSTEMS}, publisher={Institute of Electrical and Electronics Engineers (IEEE)}, author={Li, Li and Prabhu, Saiesh and Xie, Ziyang and Wang, Hanwen and Lu, Lu and Xu, Xu}, year={2021}, month={Oct}, pages={494–503} }