Search results

Limited by the types of sensors, the state information available for musculoskeletal robots with highly redundant, nonlinear muscles is often incomplete, which makes the control face a bottleneck problem. The aim of this paper is to design a method to improve the motion performance of musculoskeletal robots in partially observable scenarios, and to leverage the ontology knowledge to enhance the algorithm’s adaptability to musculoskeletal robots that have undergone changes.

A memory and attention-based reinforcement learning method is proposed for musculoskeletal robots with prior knowledge of muscle synergies. First, to deal with partially observed states available to musculoskeletal robots, a memory and attention-based network architecture is proposed for inferring more sufficient and intrinsic states. Second, inspired by muscle synergy hypothesis in neuroscience, prior knowledge of a musculoskeletal robot’s muscle synergies is embedded in network structure and reward shaping.

Based on systematic validation, it is found that the proposed method demonstrates superiority over the traditional twin delayed deep deterministic policy gradients (TD3) algorithm. A musculoskeletal robot with highly redundant, nonlinear muscles is adopted to implement goal-directed tasks. In the case of 21-dimensional states, the learning efficiency and accuracy are significantly improved compared with the traditional TD3 algorithm; in the case of 13-dimensional states without velocities and information from the end effector, the traditional TD3 is unable to complete the reaching tasks, while the proposed method breaks through this bottleneck problem.

In this paper, a novel memory and attention-based reinforcement learning method with prior knowledge of muscle synergies is proposed for musculoskeletal robots to deal with partially observable scenarios. Compared with the existing methods, the proposed method effectively improves the performance. Furthermore, this paper promotes the fusion of neuroscience and robotics.

The purpose of this paper is to focus on an engineering application of the vertebrate musculoskeletal system. The musculoskeletal system has unique mechanisms such as bi‐articular muscle, antagonistic muscle pairs and muscle‐tendon elasticity. The “artificial musculoskeletal system” is achieved through the use of the pneumatic artificial muscles. The study provides a novel method to describe the force property of the articulated mechanism driven by muscle actuator and a transmission.

A musculoskeletal system consists of multiple bodies connected together with rotational joints and driven by mono‐ and bi‐articular actuators. The paper analyzes properties of the musculoskeletal system with statically calculated omni‐directional output forces. A set of experiments has been performed to demonstrate the physical ability of the musculoskeletal robot.

A method to design a musculoskeletal system is proposed based on an analysis of the profile of convex polygon of maximum output forces. The result shows that the well‐designed musculoskeletal system enables the legged robot to jump 0.6 m high and land softly from 1.0 m drop off.

The paper provides a design principle for a musculoskeletal robot. The musculoskeletal system is the bio‐inspired mechanism for all multi‐degrees‐of‐freedom articulated devices, and has the advantages of optimized actuator configuration and force control.

Human-like musculoskeletal robots can fulfill flexible movement and manipulation with the help of multi joints and actuators. However, in general, sophisticated structures, accurate sensors and well-designed control are all necessary for a musculoskeletal robot to achieve high-precision movement. How to realize the reliable and accurate movement of the robot under the condition of limited sensing and control accuracy is still a bottleneck problem. This paper aims to improve the movement performance of musculoskeletal system by bio-inspired method.

Inspired by two kinds of natural constraints, the convergent force field found in neuroscience and attractive region in the environment found in information science, the authors proposed a structure transforming optimization algorithm for constructing constraint force field in musculoskeletal robots. Due to the characteristics of rigid-flexible coupling and variable structures, a constraint force field can be constructed in the task space of the musculoskeletal robot by optimizing the arrangement of muscles.

With the help of the constraint force field, the robot can complete precise and robust movement with constant control signals, which brings in the possibility to reduce the requirement of sensing feedback during the motion control of the robot. Experiments are conducted on a musculoskeletal model to evaluate the performance of the proposed method in movement accuracy, noise robustness and structure sensitivity.

A novel concept, constraint force field, is proposed to realize high-precision movements of musculoskeletal robots. It provides a new theoretical basis for improving the performance of robotic manipulation such as assembly and grasping under the condition that the accuracy of control and sensory are limited.

Jumping robots with coordinated multiple legs have been a hot research subject during the past years because of their excellent abilities in fast moving and obstacle-climbing. However, dynamics of jumping process of these coordinated legged robots are complex because of collisions between coordinated legs and the ground. This paper aims to analyze features of jumping process and to present the kinematic and dynamic models of a novel sole-type quadruped jumping robot with variable coordinated joints.

A complete jumping period of is divided into several subphases according to contact status of different coordinated legs to the ground. Continuous dynamics and discrete dynamics are established in different subphases. Simulations are performed in MATLAB software and ADAMS environment.

Comparison between two-set simulated results acquired from ADAMS and MATLAB demonstrates the validity of kinematic and dynamic equations.

The established dynamics establish the foundation of further research in motion planning and controller design of coordinated multiple legs.

This study aims to improve the muscle model control performance of a tendon-driven musculoskeletal system (TDMS) to overcome disadvantages such as multisegmentation and strong coupling. An adaptive network controller (ANC) with a disturbance observer is established to reduce the modeling error of the musculoskeletal model and improve its antidisturbance ability.

In contrast to other control technologies adopted for musculoskeletal humanoids, which use geometric relationships and antagonist inhibition control, this study develops a method comprising of three parts. (1) First, a simplified musculoskeletal model is constructed based on the Taylor expansion, mean value theorem and Lagrange–d’Alembert principle to complete the decoupling of the muscle model. (2) Next, for this simplified musculoskeletal model, an adaptive neuromuscular controller is designed to acquire the muscle-activation signal and realize stable tracking of the endpoint of the muscle-driven robot relative to the desired trajectory in the TDMS. For the ANC, an adaptive neural network controller with a disturbance observer is used to approximate dynamical uncertainties. (3) Using the Lyapunov method, uniform boundedness of the signals in the closed-loop system is proved. In addition, a tracking experiment is performed to validate the effectiveness of the adaptive neuromuscular controller.

The experimental results reveal that compared with other control technologies, the proposed design techniques can effectively improve control accuracy. Moreover, the proposed controller does not require extensive considerations of the geometric and antagonistic inhibition relationships, and it demonstrates anti-interference ability.

Musculoskeletal robots with humanoid structures have attracted considerable attention from numerous researchers owing to their potential to avoid danger for humans and the environment. The controller based on bio-muscle models has shown great performance in coordinating the redundant internal forces of TDMS. Therefore, adaptive controllers with disturbance observers are designed to improve the immunity of the system and thus directly regulate the internal forces between the bio-muscle models.

This study aims to propose a novel lightweight tendon-driven musculoskeletal arm (LTDM-arm) robot with a flexible series–parallel mixed skeletal joint structure and modularized artificial muscle system (MAMS). The proposed LTDM-arm exhibits human-like flexibility, safety and operational accuracy. In addition, to improve the safety and stability of the LTDM-arm, a control method is proposed to solve local artificial muscle overload accidents.

The proposed LTDM-arm comprises seven degrees of freedom skeletons, 15 MAMSs and various sensor systems (joint sensing, muscle tension sensing, visual sensing, etc.). It retains the morphology of a human skeleton (humerus, ulna and radius) and a simplified muscle configuration. This study proposes an input saturation control with full-state constraints to reduce local artificial muscle overload accidents caused by redundant muscle tension calculations.

3D circular trajectory experiments were conducted to verify the stability of the control method and the flexibility of the LTDM-arm. The results showed that the average error of the muscle length was approximately 0.35 mm (0.38%), which indicates that the proposed control scheme can make the output follow the target trajectory while ensuring constraint satisfaction.

The human arm is capable of performing compliant operations rapidly, flexibly and robustly in unstructured environments. Existing musculoskeletal arm robots lack simulations of the full morphology of the human arm and are insufficient in dexterity. However, the flexibility and safety features of the proposed LTDM-arm were consistent with that of the human arm. Therefore, this study offers a new approach for investigating the advantages of the musculoskeletal system and the concepts of muscle control.

This study aims to conduct text mining and semantic network analysis of muscle-supportive and posture-corrective wearable robots for the elderly to understand key terms related to the topic and to identify considerations for developing these types of clothing.

The authors searched and identified the key terms wearable robot, muscle-supportive, posture correction and elderly using the text-mining software Textom to extract terms as well as the network analysis software UCINET 6 to process and visualize the relationships among the terms. The authors compared and analyzed the term frequency (TF), the TF-inverse document frequency and the degree centrality of the terms, and the authors visualized and summarized the terms using NetDraw.

The key terms and their relationships in 3–4 groups were identified: wearable robot, muscle-supportive, posture correction and elderly. The authors identified the aspects of designing muscle-supportive and posture-corrective wearable robots for the elderly.

This study contributes to the field of muscle-supportive clothing and wearable robotics by deriving insights into what people are discussing and interested in, and by offering recommendations when developing these types of clothing for the elderly.

Recognizing construction workers' activities is critical for on-site performance and safety management. Thus, this study presents the potential of automatically recognizing construction workers' actions from activations of the erector spinae muscles.

A lab study was conducted wherein the participants (n = 10) performed rebar task, which involved placing and tying subtasks, with and without a wearable robot (exoskeleton). Trunk muscle activations for both conditions were trained with nine well-established supervised machine learning algorithms. Hold-out validation was carried out, and the performance of the models was evaluated using accuracy, precision, recall and F1 score.

Results indicate that classification models performed well for both experimental conditions with support vector machine, achieving the highest accuracy of 83.8% for the “exoskeleton” condition and 74.1% for the “without exoskeleton” condition.

The study paves the way for the development of smart wearable robotic technology which can augment itself based on the tasks performed by the construction workers.

This study contributes to the research on construction workers' action recognition using trunk muscle activity. Most of the human actions are largely performed with hands, and the advancements in ergonomic research have provided evidence for relationship between trunk muscles and the movements of hands. This relationship has not been explored for action recognition of construction workers, which is a gap in literature that this study attempts to address.

Snake-inspired robots are of great significance in many fields because of their great adaptability to the environment. This paper aims to systematically illustrate the research progress of snake-inspired robots according to their application environments. It classifies snake-inspired robots according to the numbers of degrees of freedom in each joint and briefly describes the modeling and control of snake-inspired robots. Finally, the application fields and future development trends of snake-inspired robots are analyzed and discussed.

This paper summarizes the research progress of snake-inspired robots and clarifies the requirements of snake-inspired robots for self-adaptive environments and multi-functional tasks. By equipping various sensors and tool modules, snake-inspired robots are developed from fixed-point operation in a single environment to autonomous operation in an amphibious environment. Finally, it is pointed out that snake-inspired robots will be developed in terms of rigid and flexible deformable structure, long endurance and multi-function and intelligent autonomous control.

Inspired by the modular and reconfigurable concepts of biological snakes, snake-inspired robots are well adapted to unknown and changing environments. Therefore, snake-inspired robots will be widely used in industrial, military, medical, post-disaster search and rescue applications. Snake-inspired robots have become a hot research topic in the field of bionic robots.

This paper summarizes the research status of snake-inspired robots, which facilitates the reader to be a comprehensive and systematic understanding of the research progress of snake-inspired robots. This helps the reader to gain inspiration from biological perspectives.

The purpose of this review paper is to address the substantial challenges of the outdated exoskeletons used for rehabilitation and further study the current advancements in this field. The shortcomings and technological developments in sensing the input signals to enable the desired motions, actuation, control and training methods are explained for further improvements in exoskeleton research.

Search platforms such as Web of Science, IEEE, Scopus and PubMed were used to collect the literature. The total number of recent articles referred to in this review paper with relevant keywords is filtered to 143.

Exoskeletons are getting smarter often with the integration of various modern tools to enhance the effectiveness of rehabilitation. The recent applications of bio signal sensing for rehabilitation to perform user-desired actions promote the development of independent exoskeleton systems. The modern concepts of artificial intelligence and machine learning enable the implementation of brain–computer interfacing (BCI) and hybrid BCIs in exoskeletons. Likewise, novel actuation techniques are necessary to overcome the significant challenges seen in conventional exoskeletons, such as the high-power requirements, poor back drivability, bulkiness and low energy efficiency. Implementation of suitable controller algorithms facilitates the instantaneous correction of actuation signals for all joints to obtain the desired motion. Furthermore, applying the traditional rehabilitation training methods is monotonous and exhausting for the user and the trainer. The incorporation of games, virtual reality (VR) and augmented reality (AR) technologies in exoskeletons has made rehabilitation training far more effective in recent times. The combination of electroencephalogram and electromyography-based hybrid BCI is desirable for signal sensing and controlling the exoskeletons based on user intentions. The challenges faced with actuation can be resolved by developing advanced power sources with minimal size and weight, easy portability, lower cost and good energy storage capacity. Implementation of novel smart materials enables a colossal scope for actuation in future exoskeleton developments. Improved versions of sliding mode control reported in the literature are suitable for robust control of nonlinear exoskeleton models. Optimizing the controller parameters with the help of evolutionary algorithms is also an effective method for exoskeleton control. The experiments using VR/AR and games for rehabilitation training yielded promising results as the performance of patients improved substantially.

Robotic exoskeleton-based rehabilitation will help to reduce the fatigue of physiotherapists. Repeated and intention-based exercise will improve the recovery of the affected part at a faster pace. Improved rehabilitation training methods like VR/AR-based technologies help in motivating the subject.

The paper describes the recent methods for signal sensing, actuation, control and rehabilitation training approaches used in developing exoskeletons. All these areas are key elements in an exoskeleton where the review papers are published very limitedly. Therefore, this paper will stand as a guide for the researchers working in this domain.