Publications

Purpose This paper aims to present an aerial manipulator that can measure the crack depth of concrete structures. Currently, the detection of crack depth in concrete structures depends mainly on manual operations with nondestructive testing (NDT) instruments. There still lacks automatic equipment to release the workers at height from the highly dangerous and time-consuming NDT detection tasks. Design/methodology/approach The proposed aerial manipulator consists of a hexacopter, a two degrees of freedom (DOF) manipulator and a custom-designed ultrasonic crack detector based on the ultrasonic flat-measured principles. A four-phase control strategy is proposed for the aerial manipulator to complete the concrete crack depth measuring task. Findings The experimental results show that the proposed control strategy is effective and the prototype of the aerial manipulator succeeds in contact detection of the cracks in the concrete specimens. The field test results further verify that the proposed aerial manipulator is capable of contact inspection of bridge piers and other similar concrete structures. Originality/value The proposed aerial manipulator provides an unmanned aerial vehicle (UAV) based solution for NDT detection of concrete crack depth. It serves as a novel tool for the inspection personnel working at height to get the interior damage conditions of concrete structures safely and quickly.

This article proposes a four-dimensional brain mapping method, which can represent the continuous process of a person's fatigue state in the form of image frames in a space-time range. This work couples 3-D-convolutional neural networks and long-short-term memory networks to form a cognitive detection model of brain fatigue dynamics, which can simulate the continuous process of a person's brain fatigue dynamics and accurately identify different cognitive fatigue states. Our approach can be applied to any type of brain fatigue detection.

A wall-climbing robot is an electromechanical device capable of autonomous or semi-autonomous movement on intricate vertical surfaces (e.g., walls, glass facades, pipelines, ceilings, etc.), typically incorporating sensing and adaptive control systems to enhance task performance. It is designed to perform tasks such as inspection, cleaning, maintenance, and rescue while maintaining stable adhesion to the surface. Its applications span various sectors, including industrial maintenance, marine engineering, and aerospace manufacturing. This paper provides a systematic review of the physical principles and scalability of various attachment methods used in wall-climbing robots, with a focus on the applicability and limitations of different attachment mechanisms in relation to robot size and structural design. For specific attachment methods, the design and compatibility of motion and attachment mechanisms are analyzed to offer design guidance for wall-climbing robots tailored to different operational tasks. Additionally, this paper reviews localization and path planning methods for wall-climbing robots, comparing graph search, sampling-based, and feedback-based algorithms to guide strategy selection across varying environments and tasks. Finally, this paper outlines future development trends in wall-climbing robots, including the diversification of locomotion mechanisms, hybridization of attachment systems, and advancements in intelligent localization and path planning. This work provides a comprehensive theoretical foundation and practical reference for the design and application of wall-climbing robots.

Emotion recognition realizing human inner perception has a very important application prospect in human-computer interaction. In order to improve the accuracy of emotion recognition, a novel method combining fused nonlinear features and team-collaboration identification strategy was proposed for emotion recognition using physiological signals. Four nonlinear features, namely approximate entropy (ApEn), sample entropy (SaEn), fuzzy entropy (FuEn) and wavelet packet entropy (WpEn) are employed to reflect emotional states deeply with each type of physiological signal. Then the features of different physiological signals are fused to represent the emotional states from multiple perspectives. Each classifier has its own advantages and disadvantages. In order to make full use of the advantages of other classifiers and avoid the limitation of single classifier, the team-collaboration model is built and the team-collaboration decision-making mechanism is designed according to the proposed team-collaboration identification strategy which is based on the fusion of support vector machine (SVM), decision tree (DT) and extreme learning machine (ELM). Through analysis, SVM is selected as the main classifier with DT and ELM as auxiliary classifiers. According to the designed decision-making mechanism, the proposed team-collaboration identification strategy can effectively employ different classification methods to make decision based on the characteristics of the samples through SVM classification. For samples which are easy to be identified by SVM, SVM directly determines the identification results, whereas SVM-DT-ELM collaboratively determines the identification results, which can effectively utilize the characteristics of each classifier and improve the classification accuracy. The effectiveness and universality of the proposed method are verified by Augsburg database and database for emotion analysis using physiological (DEAP) signals. The experimental results uniformly indicated that the proposed method combining fused nonlinear features and team-collaboration identification strategy presents better performance than the existing methods.

This article proposes a rapider OpenPose model (ROpenPose) to solve the posture detection problem of astronauts in a space capsule in a weightless environment. The ROpenPose model has three innovations as follows: 1) It uses MobileNets instead of VGG-19 to achieve lighter calculations while ensuring the accuracy of model recognition. 2) Three small convolution kernels replace the large convolution kernel of the original OpenPose, which significantly reduces the computational complexity of the model. 3) Through the parameter sharing of a convolution process, the original two-branch structure is changed to a single-branch structure, which obviously improves the calculation speed of the model. A residual network is proposed to suppress the hidden danger of gradient disappearance. The deployment of ROpenPose greatly improves astronauts' detection efficiency while ensuring their high detection performance, and thereby realizing the real-time monitoring of their operation attitude. Experimental results show that ROpenPose runs at speed higher than and detection performance comparable to a number of the existing state-of-the-art models.

Tactual length perception is not only affected by the distance between two physical stimulation points, but also related with the patterns of the tactile stimuli. The vibration parameters which we study include spatial distribution (regular or irregular), vibratory rhythm (simultaneously or sequentially), number of stimulators, burst duration and vibration intensity. This paper mainly presents three experiments aimed to study the perceived length evoked by vibrotactile stimuli. In the pair-wise comparison experiments, standard (fixed) and comparison lines were respectively presented on the back of subjects. The standard line is the linearly arranged tactors with constant vibration parameters, while the vibration parameters of the comparison ones are variable. The results showed that the vibration parameters contributed differently to the length estimation performance. Vibration intensity, spatial distribution, rhythm have little effect on tactile length perception. While the number of stimulators and vibratory duration both having significant effect on the length estimation performance. There is a tendency to overestimate the vibration lines which contains more tactors or has longer stimulus duration.

The COVID-19 pandemic has highlighted the importance of social distancing to prevent the spread of infectious diseases. However, enforcing social distancing in public spaces with traditional methods, such as hiring workers or using robots to remind people, can be unwelcome. In this paper, we propose a new technique to help maintain social distancing using augmented reality. We hypothesize that visualizing respiratory droplets with an augmented-reality interface can help individuals avoid getting too close to others. To test this hypothesis, we developed an augmented reality prototype that combines real-time head pose tracking with particle systems to align special effects to the subjects in the real world. We conducted user studies to evaluate the effectiveness of our method in daily communication scenes. Experimental results validated the efficiency of the proposed method and showed the potential of augmented reality for future epidemiological protection.