Review of machine learning methods in soft robotics

Unsupervised Sim-to-Real Adaptation of Soft Robot Proprioception Using a Dual Cross-Modal Autoencoder

Data-driven calibration methods have shown promising results for accurate proprioception in soft robotics. This process can be greatly benefited by adopting numerical simulation for computational efficiency. However, the gap between the simulated and real domains limits the accurate, generalized application of the approach. Herein, we propose an unsupervised domain adaptation framework as a data-efficient, generalized alignment of these heterogeneous sensor domains. A dual cross-modal autoencoder was designed to match the sensor domains at a feature level without any extensive labeling process, facilitating the computationally efficient transferability to various tasks. Moreover, our framework integrates domain adaptation with anomaly detection, which endows robots with the capability for external collision detection. As a proof-of-concept, the methodology was adopted for the famous soft robot design, a multigait soft robot, and two fundamental perception tasks for autonomous robot operation, involving high-fidelity shape estimation and collision detection. The resulting perception demonstrates the digital-twinned calibration process in both the simulated and real domains. The proposed design outperforms the existing prevalent benchmarks for both perception tasks. This unsupervised framework envisions a new approach to imparting embodied intelligence to soft robotic systems via blending simulation.

Soft Materials and Devices Enabling Sensorimotor Functions in Soft Robots

Sensorimotor functions, the seamless integration of sensing, decision-making, and actuation, are fundamental for robots to interact with their environments. Inspired by biological systems, the incorporation of soft materials and devices into robotics holds significant promise for enhancing these functions. However, current robotics systems often lack the autonomy and intelligence observed in nature due to limited sensorimotor integration, particularly in flexible sensing and actuation. As the field progresses toward soft, flexible, and stretchable materials, developing such materials and devices becomes increasingly critical for advanced robotics. Despite rapid advancements individually in soft materials and flexible devices, their combined applications to enable sensorimotor capabilities in robots are emerging. This review addresses this emerging field by providing a comprehensive overview of soft materials and devices that enable sensorimotor functions in robots. We delve into the latest development in soft sensing technologies, actuation mechanism, structural designs, and fabrication techniques. Additionally, we explore strategies for sensorimotor control, the integration of artificial intelligence (AI), and practical application across various domains such as healthcare, augmented and virtual reality, and exploration. By drawing parallels with biological systems, this review aims to guide future research and development in soft robots, ultimately enhancing the autonomy and adaptability of robots in unstructured environments.

Research Progress in Electroactive Polymers for Soft Robotics and Artificial Muscle Applications

Soft robots, constructed from deformable materials, offer significant advantages over rigid robots by mimicking biological tissues and providing enhanced adaptability, safety, and functionality across various applications. Central to these robots are electroactive polymer (EAP) actuators, which allow large deformations in response to external stimuli. This review examines various EAP actuators, including dielectric elastomers, liquid crystal elastomers (LCEs), and ionic polymers, focusing on their potential as artificial muscles. EAPs, particularly ionic and electronic varieties, are noted for their high actuation strain, flexibility, lightweight nature, and energy efficiency, making them ideal for applications in mechatronics, robotics, and biomedical engineering. This review also highlights piezoelectric polymers like polyvinylidene fluoride (PVDF), known for their flexibility, biocompatibility, and ease of fabrication, contributing to tactile and pressure sensing in robotic systems. Additionally, conducting polymers, with their fast actuation speeds and high strain capabilities, are explored, alongside magnetic polymer composites (MPCs) with applications in biomedicine and electronics. The integration of machine learning (ML) and the Internet of Things (IoT) is transforming soft robotics, enhancing actuation, control, and design. Finally, the paper discusses future directions in soft robotics, focusing on self-healing composites, bio-inspired designs, sustainability, and the continued integration of IoT and ML for intelligent, adaptive, and responsive robotic systems.

Optimized inverse kinematics modeling and joint angle prediction for six-degree-of-freedom anthropomorphic robots with Explainable AI

Inverse kinematics, crucial in robotics, involves computing joint configurations to achieve specific end-effector positions and orientations. This task is particularly complex for six-degree-of-freedom (six-DoF) anthropomorphic robots due to complicated mathematical equations, nonlinear behaviours, multiple valid solutions, physical constraints, non-generalizability and computational demands. The primary contribution of this work is to address the complex inverse kinematics problem for six-DoF anthropomorphic robots through the systematic exploration of AI models. This study involves rigorous evaluation and Bayesian optimization for hyperparameter tuning to identify the optimal regressor, balancing both accuracy and computational efficiency. Utilizing five-fold cross-validation on a publicly available dataset, the selected model demonstrates exceptional performance in predicting six joint angles for end effector configuration, yielding an average mean square error of 1.934 × 10-3 to 3.522 × 10-3. Its computational efficiency, with a prediction time of approximately 1.25 ms per sample, makes it a practical choice. Additionally, the study employs Explainable AI, using SHAP (SHapley Additive exPlanations) analysis to gain an understanding of feature importance. This analysis not only enhances model interpretability but also reaffirms the efficacy in this challenging multi-input multi-output predictive task. This research advances state-of-the-art models and neural networks by prioritizing computational efficiency alongside accuracy-a critical yet often overlooked factor. Pioneering a significant advancement in anthropomorphic robot kinematics, it balances accuracy and efficiency, offering practical robotic automation solutions.

Cloud and IoT based smart agent-driven simulation of human gait for detecting muscles disorder

Motion disorders affect a significant portion of the global population. While some symptoms can be managed with medications, these treatments often impact all muscles uniformly, not just the affected ones, leading to potential side effects including involuntary movements, confusion, and decreased short-term memory. Currently, there is no dedicated application for differentiating healthy muscles from abnormal ones. Existing analysis applications, designed for other purposes, often lack essential software engineering features such as a user-friendly interface, infrastructure independence, usability and learning ability, cloud computing capabilities, and AI-based assistance. This research proposes a computer-based methodology to analyze human motion and differentiate between healthy and unhealthy muscles. First, an IoT-based approach is proposed to digitize human motion using smartphones instead of hardly accessible wearable sensors and markers. The motion data is then simulated to analyze the neuromusculoskeletal system. An agent-driven modeling method ensures the naturalness, accuracy, and interpretability of the simulation, incorporating neuromuscular details such as Henneman's size principle, action potentials, motor units, and biomechanical principles. The results are then provided to medical and clinical experts to aid in differentiating between healthy and unhealthy muscles and for further investigation. Additionally, a deep learning-based ensemble framework is proposed to assist in the analysis of the simulation results, offering both accuracy and interpretability. A user-friendly graphical interface enhances the application's usability. Being fully cloud-based, the application is infrastructure-independent and can be accessed on smartphones, PCs, and other devices without installation. This strategy not only addresses the current challenges in treating motion disorders but also paves the way for other clinical simulations by considering both scientific and computational requirements.

Versatile graceful degradation framework for bio-inspired proprioception with redundant soft sensors

Reliable proprioception and feedback from soft sensors are crucial for enabling soft robots to function intelligently in real-world environments. Nevertheless, soft sensors are fragile and are susceptible to various damage sources in such environments. Some researchers have utilized redundant configuration, where healthy sensors compensate instantaneously for lost ones to maintain proprioception accuracy. However, achieving consistently reliable proprioception under diverse sensor degradation remains a challenge. This paper proposes a novel framework for graceful degradation in redundant soft sensor systems, incorporating a stochastic Long Short-Term Memory (LSTM) and a Time-Delay Feedforward Neural Network (TDFNN). The LSTM estimates readings from healthy sensors to compare them with actual data. Then, statistically abnormal readings are zeroed out. The TDFNN receives the processed sensor readings to perform proprioception. Simulation experiments with a musculoskeletal leg that contains 40 nonlinear soft sensors demonstrate the effectiveness of the proposed framework. Results show that the knee angle proprioception accuracy is retained across four distinct degradation scenarios. Notably, the mean proprioception error increases by less than 1.91°(1.36%) when 30 % of the sensors are degraded. These results suggest that the proposed framework enhances the reliability of soft sensor proprioception, thereby improving the robustness of soft robots in real-world applications.

Upgrading and extending the life cycle of soft robots with in situ free-form liquid three-dimensional printing

Soft robotics hardware, with numerous applications ranging from health care to exploration of unstructured environments, suffers from limited life cycles, which lead to waste generation and poor sustainability. Soft robots combine soft or hybrid components via complex assembly and disassembly workflows, which complicate the repair of broken components, hinder upgradability, and ultimately reduce their life spans. In this work, an advanced extrusion-based additive manufacturing process, in situ free-form liquid three-dimensional printing (iFL3DP), was developed to facilitate functional upgrades and repairs in soft robots. A yield-stress hydrogel-a type of material that can maintain its shape until sufficient stress is applied-was first printed directly onto the robot surface, serving as a support for printing new components. This technique enabled the fabrication of advanced components with seamless integration onto already assembled robots. These components could combine multiple materials with intricate geometries, including overhangs and high-aspect ratio shapes, that are considerably challenging to manufacture and integrate via traditional methods such as casting. This approach was successfully applied to upgrade an existing soft robot by adding three advanced functionalities: whisker-like sensors for tactile feedback, a grasping mechanism, and a multifunctional passive whisker array. This study showcases the easy repairability of features, new and old, substantially extending the robot's life span. This workflow has potential to enhance the sustainable development of soft robots.

Learning-Based Control for Soft Robot-Environment Interaction with Force/Position Tracking Capability

Soft robotics promises to achieve safe and efficient interactions with the environment by exploiting its inherent compliance and designing control strategies. However, effective control for the soft robot-environment interaction has been a challenging task. The challenges arise from the nonlinearity and complexity of soft robot dynamics, especially in situations where the environment is unknown and uncertainties exist, making it difficult to establish analytical models. In this study, we propose a learning-based optimal control approach as an attempt to address these challenges, which is an optimized combination of a feedforward controller based on probabilistic model predictive control and a feedback controller based on nonparametric learning methods. The approach is purely data-driven, without prior knowledge of soft robot dynamics and environment structures, and can be easily updated online to adapt to unknown environments. A theoretical analysis of the approach is provided to ensure its stability and convergence. The proposed approach enabled a soft robotic manipulator to track target positions and forces when interacting with a manikin in different cases. Moreover, comparisons with other data-driven control methods show a better performance of our approach. Overall, this work provides a viable learning-based control approach for soft robot-environment interactions with force/position tracking capability.

Explainable machine learning model of disorganisation in swarms of drones

The main challenges when managing a fleet of unmanned aerial vehicles are to ensure the relative stability of its formation and to minimise disorganisation, specifically when undergoing an intrusion. When planning the mission it is beneficial for the operator to set the parameters of the formation to balance the needs of the mission with the disorganisation that an intruder may cause. The model developed in this research predicts the anticipated disturbance as a function of the parameters of the formation. The effectiveness of six machine learning methods are compared with a previously established baseline, using data obtained from simulations. CatBoost (categorical boosting) delivered the best results, with an R 2 (coefficient of determination) value of 83.3%, representing an improvement of 80% over the baseline. The SHAP (Shapley Additive Explanations) method was used to extend the model beyond predictability for particular combinations of values of parameters, towards generalised recommendations for the operator of the formation.

Advancing materials science through next-generation machine learning

For over a decade, machine learning (ML) models have been making strides in computer vision and natural language processing (NLP), demonstrating high proficiency in specialized tasks. The emergence of large-scale language and generative image models, such as ChatGPT and Stable Diffusion, has significantly broadened the accessibility and application scope of these technologies. Traditional predictive models are typically constrained to mapping input data to numerical values or predefined categories, limiting their usefulness beyond their designated tasks. In contrast, contemporary models employ representation learning and generative modeling, enabling them to extract and encode key insights from a wide variety of data sources and decode them to create novel responses for desired goals. They can interpret queries phrased in natural language to deduce the intended output. In parallel, the application of ML techniques in materials science has advanced considerably, particularly in areas like inverse design, material prediction, and atomic modeling. Despite these advancements, the current models are overly specialized, hindering their potential to supplant established industrial processes. Materials science, therefore, necessitates the creation of a comprehensive, versatile model capable of interpreting human-readable inputs, intuiting a wide range of possible search directions, and delivering precise solutions. To realize such a model, the field must adopt cutting-edge representation, generative, and foundation model techniques tailored to materials science. A pivotal component in this endeavor is the establishment of an extensive, centralized dataset encompassing a broad spectrum of research topics. This dataset could be assembled by crowdsourcing global research contributions and developing models to extract data from existing literature and represent them in a homogenous format. A massive dataset can be used to train a central model that learns the underlying physics of the target areas, which can then be connected to a variety of specialized downstream tasks. Ultimately, the envisioned model would empower users to intuitively pose queries for a wide array of desired outcomes. It would facilitate the search for existing data that closely matches the sought-after solutions and leverage its understanding of physics and material-behavior relationships to innovate new solutions when pre-existing ones fall short.

Integrated Actuation and Sensing: Toward Intelligent Soft Robots

Soft robotics has received substantial attention due to its remarkable deformability, making it well-suited for a wide range of applications in complex environments, such as medicine, rescue operations, and exploration. Within this domain, the interaction of actuation and sensing is of utmost importance for controlling the movements and functions of soft robots. Nonetheless, current research predominantly focuses on isolated actuation and sensing capabilities, often neglecting the critical integration of these 2 domains to achieve intelligent functionality. In this review, we present a comprehensive survey of fundamental actuation strategies and multimodal actuation while also delving into advancements in proprioceptive and haptic sensing and their fusion. We emphasize the importance of integrating actuation and sensing in soft robotics, presenting 3 integration methodologies, namely, sensor surface integration, sensor internal integration, and closed-loop system integration based on sensor feedback. Furthermore, we highlight the challenges in the field and suggest compelling directions for future research. Through this comprehensive synthesis, we aim to stimulate further curiosity among researchers and contribute to the development of genuinely intelligent soft robots.

Towards high performance and durable soft tactile actuators

Soft actuators are gaining significant attention due to their ability to provide realistic tactile sensations in various applications. However, their soft nature makes them vulnerable to damage from external factors, limiting actuation stability and device lifespan. The susceptibility to damage becomes higher with these actuators often in direct contact with their surroundings to generate tactile feedback. Upon onset of damage, the stability or repeatability of the device will be undermined. Eventually, when complete failure occurs, these actuators are disposed of, accumulating waste and driving the consumption of natural resources. This emphasizes the need to enhance the durability of soft tactile actuators for continued operation. This review presents the principles of tactile feedback of actuators, followed by a discussion of the mechanisms, advancements, and challenges faced by soft tactile actuators to realize high actuation performance, categorized by their driving stimuli. Diverse approaches to achieve durability are evaluated, including self-healing, damage resistance, self-cleaning, and temperature stability for soft actuators. In these sections, current challenges and potential material designs are identified, paving the way for developing durable soft tactile actuators.

Modeling of a light-fueled self-paddling boat with a liquid crystal elastomer-based motor

Active materials possess unique properties of being able to respond autonomously to external stimuli, yet realizing and regulating the motion behavior of active machines remains a major challenge. Conventional control approaches, including sensor control and external device control, are both complex and difficult to implement. In contrast, active materials-based self-oscillators offer distinct properties such as periodic motion and ease of regulation. Inspired by paddle boats, we have proposed a conceptual light-fueled self-paddling boat with a photothermally responsive liquid crystal elastomer (LCE)-based motor that operates under steady illumination and incorporates an LCE fiber. Based on the well-established dynamic LCE model and rotation dynamics, the dynamic equations for governing the self-paddling of the LCE-steered boat are derived, and the driving torque of the LCE-based motor and the paddling velocity of the LCE-steered boat are formulated successively. The numerical results show that two motion modes of the boat under steady illumination: the static mode and the self-paddling mode. The self-paddling regime arises from the competition between the light-fueled driving torque and the frictional torque. Moreover, the critical conditions required to trigger the self-paddling are quantitatively examined as well as the significant system parameters affecting the driving torque, angular velocity, and paddling velocity. The proposed conceptual light-fueled self-paddling LCE-steered boat exhibits benefits including customizable size and being untethered and ambient powered, which provides valuable insights into the design and application of micromachines, soft robotics, energy harvesters, and beyond.

Adaptive Online Learning and Robust 3-D Shape Servoing of Continuum and Soft Robots in Unstructured Environments

In this article, we present a novel and generic data-driven method to servo-control the 3-D shape of continuum and soft robots based on proprioceptive sensing feedback. Developments of 3-D shape perception and control technologies are crucial for continuum and soft robots to perform tasks autonomously in surgical interventions. However, owing to the nonlinear properties of continuum robots, one main difficulty lies in the modeling of them, especially for soft robots with variable stiffness. To address this problem, we propose a versatile learning-based adaptive shape controller by leveraging proprioception of 3-D configuration from fiber Bragg grating (FBG) sensors, which can online estimate the unknown model of continuum robot against unexpected disturbances and exhibit an adaptive behavior to the unmodeled system without priori data exploration. Based on a new composite adaptation algorithm, the asymptotic convergences of the closed-loop system with learning parameters have been proven by Lyapunov theory. To validate the proposed method, we present a comprehensive experimental study using two continuum and soft robots both integrated with multicore FBGs, including a robotic-assisted colonoscope and multisection extensible soft manipulators. The results demonstrate the feasibility, adaptability, and superiority of our controller in various unstructured environments, as well as phantom experiments.

Chorda Dorsalis System as a Paragon for Soft Medical Robots to Design Echocardiography Probes with a New SOM-Based Steering Control

Continuum robots play the role of end effectors in various surgical robots and endoscopic devices. While soft continuum robots (SCRs) have proven advantages such as safety and compliance, more research and development are required to enhance their capability for specific medical scenarios. This research aims at designing a soft robot, considering the concepts of geometric and kinematic similarities. The chosen application is a semi-invasive medical application known as transesophageal echocardiography (TEE). The feasibility of fabrication of a soft endoscopic device derived from the Chorda dorsalis paragon was shown empirically by producing a three-segment pneumatic SCR. The main novelties include bioinspired design, modeling, and a navigation control strategy presented as a novel algorithm to maintain a kinematic similarity between the soft robot and the rigid counterpart. The kinematic model was derived based on the method of transformation matrices, and an algorithm based on a self-organizing map (SOM) network was developed and applied to realize kinematic similarity. The simulation results indicate that the control method forces the soft robot tip to follow the path of the rigid probe within the prescribed distance error (5 mm). The solution provides a soft robot that can surrogate and succeed the traditional rigid counterpart owing to size, workspace, and kinematics.

Machine-learned wearable sensors for real-time hand-motion recognition: toward practical applications

Soft electromechanical sensors have led to a new paradigm of electronic devices for novel motion-based wearable applications in our daily lives. However, the vast amount of random and unidentified signals generated by complex body motions has hindered the precise recognition and practical application of this technology. Recent advancements in artificial-intelligence technology have enabled significant strides in extracting features from massive and intricate data sets, thereby presenting a breakthrough in utilizing wearable sensors for practical applications. Beyond traditional machine-learning techniques for classifying simple gestures, advanced machine-learning algorithms have been developed to handle more complex and nuanced motion-based tasks with restricted training data sets. Machine-learning techniques have improved the ability to perceive, and thus machine-learned wearable soft sensors have enabled accurate and rapid human-gesture recognition, providing real-time feedback to users. This forms a crucial component of future wearable electronics, contributing to a robust human-machine interface. In this review, we provide a comprehensive summary covering materials, structures and machine-learning algorithms for hand-gesture recognition and possible practical applications through machine-learned wearable electromechanical sensors.

Latent Representation-Based Learning Controller for Pneumatic and Hydraulic Dual Actuation of Pressure-Driven Soft Actuators

The pneumatic and hydraulic dual actuation of pressure-driven soft actuators (PSAs) is promising because of their potential to develop novel practical soft robots and expand the range of soft robot applications. However, the physical characteristics of air and water are largely different, which makes it challenging to quickly adapt to a selected actuation method and achieve method-independent accurate control performance. Herein, we propose a novel LAtent Representation-based Feedforward Neural Network (LAR-FNN) for dual actuation. The LAR-FNN consists of an autoencoder (AE) and a feedforward neural network (FNN). The AE generates a latent representation of a PSA from a 30-s stairstep response. Subsequently, the FNN provides an individual inverse model of the target PSA and calculates feedforward control input by using the latent representation. The experimental results with PSAs demonstrate that the LAR-FNN can meet the requirements of dual actuation control (i.e., accurate control performance regardless of the actuation method with a short adaptation time) with a single neural network. The results suggest that a LAR-FNN can contribute to soft dual-actuation robot development and the field of soft robotics.

Designing Metasurfaces for Efficient Solar Energy Conversion

Metasurfaces have recently emerged as a promising technological platform, offering unprecedented control over light by structuring materials at the nanoscale using two-dimensional arrays of subwavelength nanoresonators. These metasurfaces possess exceptional optical properties, enabling a wide variety of applications in imaging, sensing, telecommunication, and energy-related fields. One significant advantage of metasurfaces lies in their ability to manipulate the optical spectrum by precisely engineering the geometry and material composition of the nanoresonators' array. Consequently, they hold tremendous potential for efficient solar light harvesting and conversion. In this Review, we delve into the current state-of-the-art in solar energy conversion devices based on metasurfaces. First, we provide an overview of the fundamental processes involved in solar energy conversion, alongside an introduction to the primary classes of metasurfaces, namely, plasmonic and dielectric metasurfaces. Subsequently, we explore the numerical tools used that guide the design of metasurfaces, focusing particularly on inverse design methods that facilitate an optimized optical response. To showcase the practical applications of metasurfaces, we present selected examples across various domains such as photovoltaics, photoelectrochemistry, photocatalysis, solar-thermal and photothermal routes, and radiative cooling. These examples highlight the ways in which metasurfaces can be leveraged to harness solar energy effectively. By tailoring the optical properties of metasurfaces, significant advancements can be expected in solar energy harvesting technologies, offering new practical solutions to support an emerging sustainable society.

A Deep Learning Framework for Soft Robots with Synthetic Data

Data-driven methods with deep neural networks demonstrate promising results for accurate modeling in soft robots. However, deep neural network models rely on voluminous data in discovering the complex and nonlinear representations inherent in soft robots. Consequently, while it is not always possible, a substantial amount of effort is required for data acquisition, labeling, and annotation. This article introduces a data-driven learning framework based on synthetic data to circumvent the exhaustive data collection process. More specifically, we propose a novel time series generative adversarial network with a self-attention mechanism, Transformer TimeGAN (TTGAN) to precisely learn the complex dynamics of a soft robot. On top of that, the TTGAN is incorporated with a conditioning network that enables it to produce synthetic data for specific soft robot behaviors. The proposed framework is verified on a widely used pneumatic-based soft gripper as an exemplary experimental setup. Experimental results demonstrate that the TTGAN generates synthetic time series data with realistic soft robot dynamics. Critically, a combination of the synthetic and only partially available original data produces a data-driven model with estimation accuracy comparable to models obtained from using complete original data.

A review on model-based and model-free approaches to control soft actuators and their potentials in colonoscopy

Colorectal cancer (CRC) is the third most common cancer worldwide and responsible for approximately 1 million deaths annually. Early screening is essential to increase the chances of survival, and it can also reduce the cost of treatments for healthcare centres. Colonoscopy is the gold standard for CRC screening and treatment, but it has several drawbacks, including difficulty in manoeuvring the device, patient discomfort, and high cost. Soft endorobots, small and compliant devices thatcan reduce the force exerted on the colonic wall, offer a potential solution to these issues. However, controlling these soft robots is challenging due to their deformable materials and the limitations of mathematical models. In this Review, we discuss model-free and model-based approaches for controlling soft robots that can potentially be applied to endorobots for colonoscopy. We highlight the importance of selecting appropriate control methods based on various parameters, such as sensor and actuator solutions. This review aims to contribute to the development of smart control strategies for soft endorobots that can enhance the effectiveness and safety of robotics in colonoscopy. These strategies can be defined based on the available information about the robot and surrounding environment, control demands, mechanical design impact and characterization data based on calibration.

Soft DAgger: Sample-Efficient Imitation Learning for Control of Soft Robots

This paper presents Soft DAgger, an efficient imitation learning-based approach for training control solutions for soft robots. To demonstrate the effectiveness of the proposed algorithm, we implement it on a two-module soft robotic arm involved in the task of writing letters in 3D space. Soft DAgger uses a dynamic behavioral map of the soft robot, which maps the robot's task space to its actuation space. The map acts as a teacher and is responsible for predicting the optimal actions for the soft robot based on its previous state action history, expert demonstrations, and current position. This algorithm achieves generalization ability without depending on costly exploration techniques or reinforcement learning-based synthetic agents. We propose two variants of the control algorithm and demonstrate that good generalization capabilities and improved task reproducibility can be achieved, along with a consistent decrease in the optimization time and samples. Overall, Soft DAgger provides a practical control solution to perform complex tasks in fewer samples with soft robots. To the best of our knowledge, our study is an initial exploration of imitation learning with online optimization for soft robot control.

Reduced order modeling and model order reduction for continuum manipulators: an overview

Soft robot's natural dynamics calls for the development of tailored modeling techniques for control. However, the high-dimensional configuration space of the geometrically exact modeling approaches for soft robots, i.e., Cosserat rod and Finite Element Methods (FEM), has been identified as a key obstacle in controller design. To address this challenge, Reduced Order Modeling (ROM), i.e., the approximation of the full-order models, and Model Order Reduction (MOR), i.e., reducing the state space dimension of a high fidelity FEM-based model, are enjoying extensive research. Although both techniques serve a similar purpose and their terms have been used interchangeably in the literature, they are different in their assumptions and implementation. This review paper provides the first in-depth survey of ROM and MOR techniques in the continuum and soft robotics landscape to aid Soft Robotics researchers in selecting computationally efficient models for their specific tasks.

The neuromuscular control for lower limb exoskeleton- a 50-year perspective

Historically, impaired lower limb function has resulted in heavy health burden and large economic loss in society. Although experts from various fields have put large amounts of effort into overcoming this challenge, there is still not a single standard treatment that can completely restore the lost limb function. During the past half century, with the advancing understanding of human biomechanics and engineering technologies, exoskeletons have achieved certain degrees of success in assisting and rehabilitating patients with loss of limb function, and therefore has been spotlighted in both the medical and engineering fields. In this article, we review the development milestones of lower limb exoskeletons as well as the neuromuscular interactions between the device and wearer throughout the past 50 years. Fifty years ago, the lower-limb exoskeletons just started to be devised. We review several prototypes and present their designs in terms of structure, sensor and control systems. Subsequently, we introduce the development milestones of modern lower limb exoskeletons and discuss the pros and cons of these differentiated devices. In addition, we summarize current important neuromuscular control systems and sensors; and discuss current evidence demonstrating how the exoskeletons may affect neuromuscular control of wearers. In conclusion, based on our review, we point out the possible future direction of combining multiple current technologies to build lower limb exoskeletons that can serve multiple aims.

Implementation of a Hybrid Intelligence System Enabling the Effectiveness Assessment of Interaction Channels Use in HMI

The article presents a novel idea of Interaction Quality Sensor (IQS), introduced in the complete solution of Hybrid INTelligence (HINT) architecture for intelligent control systems. The proposed system is designed to use and prioritize multiple information channels (speech, images, videos) in order to optimize the information flow efficiency of interaction in HMI systems. The proposed architecture is implemented and validated in a real-world application of training unskilled workers-new employees (with lower competencies and/or a language barrier). With the help of the HINT system, the man-machine communication information channels are deliberately chosen based on IQS readouts to enable an untrained, inexperienced, foreign employee candidate to become a good worker, while not requiring the presence of either an interpreter or an expert during training. The proposed implementation is in line with the labor market trend, which displays significant fluctuations. The HINT system is designed to activate human resources and support organizations/enterprises in the effective assimilation of employees to the tasks performed on the production assembly line. The market need of solving this noticeable problem was caused by a large migration of employees within (and between) enterprises. The research results presented in the work show significant benefits of the methods used, while supporting multilingualism and optimizing the preselection of information channels.

Soft-Sensor System for Grasp Type Recognition in Underactuated Hand Prostheses

This paper presents the development of an intelligent soft-sensor system to add haptic perception to the underactuated hand prosthesis PrHand. Two sensors based on optical fiber were constructed, one for finger joint angles and the other for fingertips' contact force. Three sensor fabrications were tested for the angle sensor by axially rotating the sensors in four positions. The configuration with the most similar response in the four rotations was chosen. The chosen sensors presented a polynomial response with R2 higher than 92%. The tactile force sensors tracked the force made over the objects. Almost all sensors presented a polynomial response with R2 higher than 94%. The system monitored the prosthesis activity by recognizing grasp types. Six machine learning algorithms were tested: linear regression, k-nearest neighbor, support vector machine, decision tree, k-means clustering, and hierarchical clustering. To validate the algorithms, a k-fold test was used with a k = 10, and the accuracy result for k-nearest neighbor was 98.5%, while that for decision tree was 93.3%, enabling the classification of the eight grip types.

Use of Finite Elements in the Training of a Neural Network for the Modeling of a Soft Robot

Soft bioinspired manipulators have a theoretically infinite number of degrees of freedom, providing considerable advantages. However, their control is very complex, making it challenging to model the elastic elements that define their structure. Finite elements (FEA) can provide a model with sufficient accuracy but are inadequate for real-time use. In this context, Machine Learning (ML) is postulated as an option, both for robot modeling and for its control, but it requires a very high number of experiments to train the model. A linked combination of both options (FEA and ML) can be an approach to the solution. This work presents the implementation of a real robot made up of three flexible modules and actuated with SMA (shape memory alloy) springs, the development of its model through finite elements, its use to adjust a neural network, and the results obtained.

An Interdisciplinary Tutorial: A Self-Healing Soft Finger with Embedded Sensor

In the field of soft robotics, knowledge of material science is becoming more and more important. However, many researchers have a background in only one of both domains. To aid the understanding of the other domain, this tutorial describes the complete process from polymer synthesis over fabrication to testing of a soft finger. Enough background is provided during the tutorial such that researchers from both fields can understand and sharpen their knowledge. Self-healing polymers are used in this tutorial, showing that these polymers that were once a specialty, have become accessible for broader use. The use of self-healing polymers allows soft robots to recover from fatal damage, as shown in this tutorial, which increases their lifespan significantly.

Recent Developments of Actuation Mechanisms for Continuum Robots: A Review

Traditional rigid robots face significant challenges in congested and tight environments, including bulky size, maneuverability, and safety limitations. Thus, soft continuum robots, inspired by the incredible capabilities of biological appendages such as octopus arms, starfish, and worms, have shown promising performance in complex environments due to their compliance, adaptability, and safety. Different actuation techniques are implemented in soft continuum robots to achieve a smoothly bending backbone, including cable-driven actuators, pneumatic actuators, and hydraulic actuation systems. However, designing and developing efficient actuation mechanisms, motion planning approaches, and control algorithms are challenging due to the high degree of redundancy and non-linearity of soft continuum robots. This article profoundly reviews the merits and drawbacks of soft robots' actuation systems concerning their applications to provide the readers with a brief review reference to explore the recent development of soft robots' actuation mechanisms technology. Moreover, the authors have surveyed the recent review studies in controller design of continuum robots as a guidance for future applications.

SoftSAR: The New Softer Side of Socially Assistive Robots-Soft Robotics with Social Human-Robot Interaction Skills

When we think of "soft" in terms of socially assistive robots (SARs), it is mainly in reference to the soft outer shells of these robots, ranging from robotic teddy bears to furry robot pets. However, soft robotics is a promising field that has not yet been leveraged by SAR design. Soft robotics is the incorporation of smart materials to achieve biomimetic motions, active deformations, and responsive sensing. By utilizing these distinctive characteristics, a new type of SAR can be developed that has the potential to be safer to interact with, more flexible, and uniquely uses novel interaction modes (colors/shapes) to engage in a heighted human-robot interaction. In this perspective article, we coin this new collaborative research area as SoftSAR. We provide extensive discussions on just how soft robotics can be utilized to positively impact SARs, from their actuation mechanisms to the sensory designs, and how valuable they will be in informing future SAR design and applications. With extensive discussions on the fundamental mechanisms of soft robotic technologies, we outline a number of key SAR research areas that can benefit from using unique soft robotic mechanisms, which will result in the creation of the new field of SoftSAR.

Specified-duration shapers for suppressing residual vibrations

Input-shaping control has received considerable research attention for suppressing residual vibrations. Although numerous studies have been conducted on designing input shapers with arbitrary robustness to modeling errors, no studies have focused on the design of input shapers with arbitrarily specified shaping times. In this study, a specified-duration (SD) shaper, which is an input shaper with an arbitrarily specified shaping time, and a systematic method to design an SD shaper using impulse vectors are proposed. As the specified shaping time increases, the SD shaper increases the number of impulses one by one according to the number of added derivative constraints, thereby improving robustness to modeling errors. The performance of the SD shaper was evaluated for a second-order system through computer simulations. The simulation results revealed that the SD shaper suppresses residual vibrations of the vibratory system at the specified shaping time. The validity of the SD shaper was experimentally verified using a horizontal beam vibration apparatus. The results of this study provide insight into the development of vibration suppression strategies with input shaping control.

Aeroelastics-aware compensation system for soft aerial vehicle stabilization

This paper describes a compensation system for soft aerial vehicle stabilization. Balancing the arms is one of the main challenges of soft UAVs since the propeller is freely tilting together with the flexible arm. In comparison with previous designs, in which the autopilot was adjusted to deal with these imbalances with no extra actuation, this work introduces a soft tendon-actuated system to achieve in-flight stabilization in an energy-efficient way. The controller is specifically designed for disturbance rejection of aeroelastic perturbations using the Ziegler-Nichols method, depending on the flight mode and material properties. This aerodynamics-aware compensation system allows to further bridge the gap between soft and aerial robotics, leading to an increase in the flexibility of the UAV, and the ability to deal with changes in material properties, increasing the useful life of the drone. In energetic terms, the novel system is 15–30% more efficient, and is the basis for future applications such as object grasping.

4D Printing: A Cutting-edge Platform for Biomedical Applications

Nature's materials have evolved over time to be able to respond to environmental stimuli by generating complex structures that can change their functions in response to distance, time, and direction of stimuli. A number of technical efforts are currently being made to improve printing resolution, shape fidelity, and printing speed to mimic the structural design of natural materials with three-dimensional (3D) printing. Unfortunately, this technology is limited by the fact that printed objects are static and cannot be reshaped dynamically in response to stimuli. In recent years, several smart materials have been developed that can undergo dynamic morphing in response to a stimulus, thus resolving this issue. Four-dimensional (4D) printing refers to a manufacturing process involving additive manufacturing, smart materials, and specific geometries. It has become an essential technology for biomedical engineering and has the potential to create a wide range of useful biomedical products. This paper will discuss the concept of 4D bioprinting and the recent developments in smart matrials, which can be actuated by different stimuli and be exploited to develop biomimetic materials and structures, with significant implications for pharmaceutics and biomedical research, as well as prospects for the future.

Prediction of arabica coffee production using artificial neural network and multiple linear regression techniques

Crop yield and its prediction are crucial in agricultural production planning. This study investigates and predicts arabica coffee yield in order to match the market demand, using artificial neural networks (ANN) and multiple linear regression (MLR). Data of six variables, including areas, productivity zones, rainfalls, relative humidity, and minimum and maximum temperature, were collected for the recent 180 months between 2004 and 2018. The predicted yield of the cherry coffee crop continuously increases each year. From the dataset, it was found that the prediction accuracy of the R2 and RMSE from ANN was 0.9524 and 0.0784 tons, respectively. The ANN model showed potential in determining the cherry coffee yields.

Review of Artificial Intelligence and Machine Learning Technologies: Classification, Restrictions, Opportunities and Challenges

Artificial intelligence (AI) is an evolving set of technologies used for solving a wide range of applied issues. The core of AI is machine learning (ML)—a complex of algorithms and methods that address the problems of classification, clustering, and forecasting. The practical application of AI&ML holds promising prospects. Therefore, the researches in this area are intensive. However, the industrial applications of AI and its more intensive use in society are not widespread at the present time. The challenges of widespread AI applications need to be considered from both the AI (internal problems) and the societal (external problems) perspective. This consideration will identify the priority steps for more intensive practical application of AI technologies, their introduction, and involvement in industry and society. The article presents the identification and discussion of the challenges of the employment of AI technologies in the economy and society of resource-based countries. The systematization of AI&ML technologies is implemented based on publications in these areas. This systematization allows for the specification of the organizational, personnel, social and technological limitations. This paper outlines the directions of studies in AI and ML, which will allow us to overcome some of the limitations and achieve expansion of the scope of AI&ML applications.

Bipedal Walking of Underwater Soft Robot Based on Data-Driven Model Inspired by Octopus

The soft organisms in nature have always been a source of inspiration for the design of soft arms and this paper draws inspiration from the octopus's tentacle, aiming at a soft robot for moving flexibly in three-dimensional space. In the paper, combined with the characteristics of an octopus's tentacle, a cable-driven soft arm is designed and fabricated, which can motion flexibly in three-dimensional space. Based on the TensorFlow framework, a data-driven model is established, and the data-driven model is trained using deep reinforcement learning strategy to realize posture control of a single soft arm. Finally, two trained soft arms are assembled into an octopus-inspired biped walking robot, which can go forward and turn around. Experimental analysis shows that the robot can achieve an average speed of 7.78 cm/s, and the maximum instantaneous speed can reach 12.8 cm/s.

A Survey of Human Gait-Based Artificial Intelligence Applications

We performed an electronic database search of published works from 2012 to mid-2021 that focus on human gait studies and apply machine learning techniques. We identified six key applications of machine learning using gait data: 1) Gait analysis where analyzing techniques and certain biomechanical analysis factors are improved by utilizing artificial intelligence algorithms, 2) Health and Wellness, with applications in gait monitoring for abnormal gait detection, recognition of human activities, fall detection and sports performance, 3) Human Pose Tracking using one-person or multi-person tracking and localization systems such as OpenPose, Simultaneous Localization and Mapping (SLAM), etc., 4) Gait-based biometrics with applications in person identification, authentication, and re-identification as well as gender and age recognition 5) “Smart gait” applications ranging from smart socks, shoes, and other wearables to smart homes and smart retail stores that incorporate continuous monitoring and control systems and 6) Animation that reconstructs human motion utilizing gait data, simulation and machine learning techniques. Our goal is to provide a single broad-based survey of the applications of machine learning technology in gait analysis and identify future areas of potential study and growth. We discuss the machine learning techniques that have been used with a focus on the tasks they perform, the problems they attempt to solve, and the trade-offs they navigate.

Underwater Soft Robotics: A Review of Bioinspiration in Design, Actuation, Modeling, and Control

Nature and biological creatures are some of the main sources of inspiration for humans. Engineers have aspired to emulate these natural systems. As rigid systems become increasingly limited in their capabilities to perform complex tasks and adapt to their environment like living creatures, the need for soft systems has become more prominent due to the similar complex, compliant, and flexible characteristics they share with intelligent natural systems. This review provides an overview of the recent developments in the soft robotics field, with a focus on the underwater application frontier.

Pneumatic Soft Actuators With Kirigami Skins

Soft pneumatic actuators have become indispensable for many robotic applications due to their reliability, safety, and design flexibility. However, the currently available actuator designs can be challenging to fabricate, requiring labor-intensive and time-consuming processes like reinforcing fiber wrapping and elastomer curing. To address this issue, we propose to use simple-to-fabricate kirigami skins-plastic sleeves with carefully arranged slit cuts-to construct pneumatic actuators with pre-programmable motion capabilities. Such kirigami skin, wrapped outside a cylindrical balloon, can transform the volumetric expansion from pneumatic pressure into anisotropic stretching and shearing, creating a combination of axial extension and twisting in the actuator. Moreover, the kirigami skin exhibits out-of-plane buckling near the slit cut, which enables high stretchability. To capture such complex deformations, we formulate and experimentally validates a new kinematics model to uncover the linkage between the kirigami cutting pattern design and the actuator's motion characteristics. This model uses a virtual fold and rigid-facet assumption to simplify the motion analysis without sacrificing accuracy. Moreover, we tested the pressure-stroke performance and elastoplastic behaviors of the kirigami-skinned actuator to establish an operation protocol for repeatable performance. Analytical and experimental parametric analysis shows that one can effectively pre-program the actuator's motion performance, with considerable freedom, simply by adjusting the angle and length of the slit cuts. The results of this study can establish the design and analysis framework for a new family of kirigami-skinned pneumatic actuators for many robotic applications.

4D printing in biomedical applications: emerging trends and technologies

Nature's material systems during evolution have developed the ability to respond and adapt to environmental stimuli through the generation of complex structures capable of varying their functions across direction, distances and time. 3D printing technologies can recapitulate structural motifs present in natural materials, and efforts are currently being made on the technological side to improve printing resolution, shape fidelity, and printing speed. However, an intrinsic limitation of this technology is that printed objects are static and thus inadequate to dynamically reshape when subjected to external stimuli. In recent years, this issue has been addressed with the design and precise deployment of smart materials that can undergo a programmed morphing in response to a stimulus. The term 4D printing was coined to indicate the combined use of additive manufacturing, smart materials, and careful design of appropriate geometries. In this review, we report the recent progress in the design and development of smart materials that are actuated by different stimuli and their exploitation within additive manufacturing to produce biomimetic structures with important repercussions in different but interrelated biomedical areas.