Background: The COVID-19 pandemic is affecting the mental and emotional well-being of patients, family members, and health care workers. Patients in the isolation ward may have psychological problems due to long-term hospitalization, the development of the epidemic, and the inability to see their families. A medical assistive robot (MAR), acting as an intermediary of communication, can be deployed to address these mental pressures. Objective: CareDo, a MAR with telepresence and teleoperation functions, was developed in this work for remote health care. The aim of this study was to investigate its practical performance in the isolation ward during the pandemic. Methods: Two systems were integrated into the CareDo robot. For the telepresence system, a web real-time communications solution is used for the multiuser chat system and a convolutional neural network is used for expression recognition. For the teleoperation system, an incremental motion mapping method is used for operating the robot remotely. A clinical trial of this system was conducted at First Affiliated Hospital, Zhejiang University. Results: During the clinical trials, tasks such as video chatting, emotion detection, and medical supplies delivery were performed via the CareDo robot. Seven voice commands were set for performing system wakeup, video chatting, and system exiting. Durations from 1 to 3 seconds of common commands were set to improve voice command detection. The facial expression was recorded 152 times for a patient in 1 day for the psychological intervention. The recognition accuracy reached 95% and 92.8% for happy and neutral expressions, respectively. Conclusions: Patients and health care workers can use this MAR in the isolation ward for telehealth care during the COVID-19 pandemic. This can be a useful approach to break the chains of virus transmission and can also be an effective way to conduct remote psychological intervention.

The rising attention to human-centricity, a core value of Industry 5.0, makes researchers propose various methods to place humans in the center of production, such as the Human Digital Twin (HDT) and gait identity recognition. However, previous HDT applications in the Industry 5.0 era have less consideration of identity recognition, the vital way to link physical signals to personal archives in cyber systems. Herein GaitFormer, a Two-Stream Transformer gait recognition method is proposed for application with the wearable inertial measurement unit (IMU) motion capture system. GaitFormer consists of Temporal and Channel Modules enclosing Gaussian Range Encoding, multi-head Auto-Correlation mechanism, and multi-scale as convolutional neural network (CNN) layers. It outperforms state-of-the-art architectures with extensive experiments over publicly available IMU-based datasets respectively: whuGait and OU-ISIR. Then the study intends to combine the motion capture system and gait identity recognition algorithm to propose an application of personal monitoring in the Human-Cyber-Physical-System (HCPS), which uses GaitFormer to recognize the identity of the user with motion capture system and show their corresponding customized HDT. This work may contribute to making a connection between gait identity recognition and HDT, constructing personal working profiles in the environment of Industry 5.0.

The fourth revolution of healthcare technologies, i.e., Healthcare 4.0, is putting robotics into human-dominated environments. In such a context, one of the main challenges is to develop human-centered robotics technologies that enable safe and reliable human-robot interaction toward human-robot symbiosis. Herein, robot skin is developed to endow healthcare robots with on-body proximity perception so as to fulfill the promise of safe and reliable robotic systems alongside humans. The sensing performance of the robot skin is evaluated by extensive experiments, providing important guidance on its effective implementation into a specific robot platform. Results show that the developed robot skin has a detection range of 0-50 mm, a maximum sensitivity of 0.7 pF/mm, a minimum resolution of 0.05 mm, a repeatability error of 6.6%, a hysteresis error of 7.1%, and bending durability of 2000 cycles. The robot skin is further customized and scaled up to form a large-area sensing system on the exterior of robot arms to support functional safety, which is experimentally validated by approaching distance monitoring and reactive collision avoidance. During the validation, the sensing feedback of the robot skin and the motion of the host robot are visualized remotely in the robot digital twin in a real-time manner via a cloud server. The cloud-based monitoring interface bridges the gap between local healthcare robots and remote professionals, illustrating promising applications where professionals monitor the robot state and intervene in challenging situations to provide instant support for emergent safety issues in human-robot interaction.

With the rapid technology development pushing the introduction of the fifth industrial revolution, Industry 5.0, robots are getting rid of fences and sharing the workspace with humans. In such a context, ensuring the safety of humans and robots is a critical demand. One of the effective methods for this is proximity sensing, among which capacitive sensors are widely used to detect the proximity of humans. However, the capacitive sensor cannot get accurate distance information since the capacitance varies with the characteristics of obstacles. This work develops a capacitive robot skin, seamlessly integrated into the proposed smooth velocity transition framework to deal with this challenge. The robot skin is customized to cover a large area on the exterior of a 6-DOF robot arm. A hierarchical proximity perception approach is used to grade the sensing state. Based on this, distance reduction and collision avoidance velocity generation methods are used to reach a smooth and quick decay of the velocity. The control strategy is applied in a pick-and-place scenery for verification. Compared to the traditional threshold trigger method, the proposed smooth velocity transition framework can greatly reduce the absolute value of the local maximum acceleration, which can enable a flexible and natural human-robot interaction while ensuring human safety.

Balance assessment is crucial for health monitoring and rehabilitation evaluation of neurological diseases like Parkinson’s disease (PD) and stroke. The Berg Balance Scale (BBS) is a widely used clinical tool for balance evaluation. However, its dependence on trained therapists for subjective, time-consuming assessments limits its scalability. Current researchers have proposed several automated assessment systems. However, they suffer from difficulty in use in clinical settings and the need for feature engineering. The rapid advancement of wearable inertial measurement units (IMUs) provides an objective tool for motion analysis that is suitable for use in clinical environments. Thus, to address the limitations of manual scoring and complexities of capturing gait features, we proposed an automated BBS assessment system using an attention-based deep learning algorithm with IMU data, integrating convolutional neural networks (CNNs) for spatial feature extraction, bidirectional long short-term memory (Bi-LSTM) networks for temporal modeling, and attention mechanisms to emphasize informative features. Validated with 20 healthy subjects (young and elderly) and 20 patients (PD and stroke), the system achieved a mean absolute error (MAE) of 1.1627 and root mean squared error (RMSE) of 1.5333. Requiring only 5 min of walking data, this approach provided an efficient, objective solution for balance assessment to assist healthcare physicians as well as patients in their own health monitoring. The key limitations included: a limited generalizability to severely impaired patients who were unable to walk independently, and the inability to predict the score of individual tasks.

The total focusing method (TFM) has been increasingly applied to weld inspection given its high image quality and defect sensitivity. Oblique incidence is widely used to steer the beam effectively, considering the defect orientation and structural complexity of welded structures. However, the conventional TFM based on the delay-and-sum (DAS) principle is time-consuming, especially for oblique incidence. In this paper, a fast full-matrix imaging algorithm in the Fourier domain is proposed to accelerate the TFM under the condition of oblique incidence. The algorithm adopts the Chebyshev polynomials of the second kind to directly expand the Fourier extrapolator with lateral sound velocity variation. The direct expansion maintains image accuracy and resolution in wide-angle situations, covering both small and large angles, making it highly suitable for weld inspection. Simulations prove that the third-order Chebyshev expansion is required to achieve image accuracy equivalent to the TFM with wide-angle incidence. Experiments verify the algorithm’s performance for weld flaws using the proposed method with the transverse wave and the full-skip mode. The depth deviation is within 0.53 mm, and the sizing error is below 15%. The imaging efficiency is improved by a factor of up to 8 compared to conventional TFM. We conclude that the proposed method is applicable to high-speed weld inspection with various oblique incidence angles.