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.

Can physical and cognitive training based on episodic memory be combined in a new protocol for daily training?

BACKGROUND: Cognitive training (CT) is defined as guided practice on a set of standard tasks designed to stimulate particular cognitive functions. Recent studies have shown that physical exercise is beneficial for cognitive activity in older adults and patients with degenerative diseases. AIMS: The main objective of the present study is to create a new cognitive tool able to provide training for cognitive functions that take advantage of the physical activity involved in the execution of the task. A study concerning the application of a new CT tool for episodic memory is presented and divided in two parts. The first one aims at developing a new sensorized device, called SmartTapestry, for physical and cognitive training. The second part aims at understanding its technical viability and level of sensitivity in stimulating the same cognitive domain covered by the standardized tests, despite the introduction of the physical activity variable. METHODS: The SmartTapestry device was tested with a total of 53 subjects, 29 healthy subjects and 24 subjects suffering from mild cognitive impairment. RESULTS AND DISCUSSIONS: The results show a good correlation between the two approaches (p < 0.005), suggesting that SmartTapestry can stimulate the same cognitive functions of traditional cognitive tasks, with the addition of physical exercise. CONCLUSIONS: The results of this study may be useful in designing ecological and combined cognitive-physical tools, which can be used daily at home, reducing the presence of clinical staff, to train at the same time the brain and the body so as to improve the cognitive treatments efficacy.

Early Intervention to Improve Sucking in Preterm Newborns: A Systematic Review of Quantitative Studies.

BACKGROUND: Premature birth is associated with feeding difficulties due to inadequate coordination of sucking, swallowing, and breathing. Nonnutritive sucking (NNS) and oral stimulation interventions may be effective for oral feeding promotion, but the mechanisms of the intervention effects need further clarifications. PURPOSE: We reviewed preterm infant intervention studies with quantitative outcomes of sucking performance to summarize the evidence of the effect of interventions on specific components of sucking. METHODS: PubMed, CINAHL, MEDLINE, EMBASE, and PSYCOLIST databases were searched for English language publications through August 2017. Studies were selected if they involved preterm infants, tested experimental interventions to improve sucking or oral feeding skills, and included outcome as an objective measure of sucking performance. Specific Medical Subject Headings (MeSH) terms were utilized. RESULTS: Nineteen studies were included in this review: 15 randomized, 1 quasi-randomized, and 3 crossover randomized controlled trials. Intervention types were grouped into 6 categories (i) NNS, (ii) NNS with auditory reinforcement, (iii) sensorimotor stimulation, (iv) oral support, (v) combined training, and (vi) nutritive sucking. Efficiency parameters were positively influenced by most types of interventions, though appear to be less affected by trainings based on NNS alone. IMPLICATIONS FOR PRACTICE: These findings may be useful in the clinical care of infants requiring support to achieve efficient sucking skills through NNS and oral stimulation interventions. IMPLICATIONS FOR RESEARCH: Further studies including quantitative measures of sucking performance outcome measures are needed in order to best understand the needs and provide more tailored interventions to preterm infants.

A new system for quantitative evaluation of infant gaze capabilities in a wide visual field.

BACKGROUND: The visual assessment of infants poses specific challenges: many techniques that are used on adults are based on the patient's response, and are not suitable for infants. Significant advances in the eye-tracking have made this assessment of infant visual capabilities easier, however, eye-tracking still requires the subject's collaboration, in most cases and thus limiting the application in infant research. Moreover, there is a lack of transferability to clinical practice, and thus it emerges the need for a new tool to measure the paradigms and explore the most common visual competences in a wide visual field. This work presents the design, development and preliminary testing of a new system for measuring infant's gaze in the wide visual field called CareToy C: CareToy for Clinics. METHODS: The system is based on a commercial eye tracker (SmartEye) with six cameras running at 60 Hz, suitable for measuring an infant's gaze. In order to stimulate the infant visually and audibly, a mechanical structure has been designed to support five speakers and five screens at a specific distance (60 cm) and angle: one in the centre, two on the right-hand side and two on the left (at 30° and 60° respectively). Different tasks have been designed in order to evaluate the system capability to assess the infant's gaze movements during different conditions (such as gap, overlap or audio-visual paradigms). Nine healthy infants aged 4-10 months were assessed as they performed the visual tasks at random. RESULTS: We developed a system able to measure infant's gaze in a wide visual field covering a total visual range of ±60° from the centre with an intermediate evaluation at ±30°. Moreover, the same system, thanks to different integrated software, was able to provide different visual paradigms (as gap, overlap and audio-visual) assessing and comparing different visual and multisensory sub-competencies. The proposed system endowed the integration of a commercial eye-tracker into a purposive setup in a smart and innovative way. CONCLUSIONS: The proposed system is suitable for measuring and evaluating infant's gaze capabilities in a wide visual field, in order to provide quantitative data that can enrich the clinical assessment.

An active one-lobe pulmonary simulator with compliance control for medical training in neonatal mechanical ventilation.

Mechanical ventilation is a current support therapy for newborns affected by respiratory diseases. However, several side effects have been observed after treatment, making it mandatory for physicians to determine more suitable approaches. High fidelity simulation is an efficient educational technique that recreates clinical experience. The aim of the present study is the design of an innovative and versatile neonatal respiratory simulator which could be useful in training courses for physicians and nurses as for mechanical ventilation. A single chamber prototype, reproducing a pulmonary lobe both in size and function, was designed and assembled. Volume and pressure within the chamber can be tuned by the operator through the device control system, in order to simulate both spontaneous and assisted breathing. An innovative software-based simulator for training neonatologists and nurses within the continuing medical education program on respiratory disease management was validated. Following the clinical needs, three friendly graphic user interfaces were implemented for simulating three different clinical scenarios (spontaneous breathing, controlled breathing and triggered/assisted ventilation modalities) thus providing physicians with an active experience. The proposed pulmonary simulator has the potential to be included in the range of computer-driven technologies used in medical training, adding novel functions and improving simulation results.

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.

Soft Robotics: A Route to Equality, Diversity, and Inclusivity in Robotics.

Robotics is entering our daily lives. The discipline is increasingly crucial in fields such as agriculture, medicine, and rescue operations, impacting our food, health, and planet. At the same time, it is becoming evident that robotic research must embrace and reflect the diversity of human society to address these broad challenges effectively. In recent years, gender inclusivity has received increasing attention, but it still remains a distant goal. In addition, awareness is rising around other dimensions of diversity, including nationality, religion, and politics. Unfortunately, despite the efforts, empirical evidence shows that the field has still a long way to go before achieving a sufficient level of equality, diversity, and inclusion across these spectra. This study focuses on the soft robotics community-a growing and relatively recent subfield-and it outlines the present state of equality and diversity panorama in this discipline. The article argues that its high interdisciplinary and accessibility make it a particularly welcoming branch of robotics. We discuss the elements that make this subdiscipline an example for the broader robotic field. At the same time, we recognize that the field should still improve in several ways and become more inclusive and diverse. We propose concrete actions that we believe will contribute to achieving this goal, and provide metrics to monitor its evolution.

Octopus-inspired sensorized soft arm for environmental interaction.

Octopuses can whip their soft arms with a characteristic "bend propagation" motion to capture prey with sensitive suckers. This relatively simple strategy provides models for robotic grasping, controllable with a small number of inputs, and a highly deformable arm with sensing capabilities. Here, we implemented an electronics-integrated soft octopus arm (E-SOAM) capable of reaching, sensing, grasping, and interacting in a large domain. On the basis of the biological bend propagation of octopuses, E-SOAM uses a bending-elongation propagation model to move, reach, and grasp in a simple but efficient way. E-SOAM's distal part plays the role of a gripper and can process bending, suction, and temperature sensory information under highly deformed working states by integrating a stretchable, liquid-metal-based electronic circuit that can withstand uniaxial stretching of 710% and biaxial stretching of 270% to autonomously perform tasks in a confined environment. By combining this sensorized distal part with a soft arm, the E-SOAM can perform a reaching-grasping-withdrawing motion across a range up to 1.5 times its original arm length, similar to the biological counterpart. Through a wearable finger glove that produces suction sensations, a human can use just one finger to remotely and interactively control the robot's in-plane and out-of-plane reaching and grasping both in air and underwater. E-SOAM's results not only contribute to our understanding of the function of the motion of an octopus arm but also provide design insights into creating stretchable electronics-integrated bioinspired autonomous systems that can interact with humans and their environments.

Bioinspired soft robots for deep-sea exploration.

The deep ocean, Earth's untouched expanse, presents immense challenges for exploration due to its extreme pressure, temperature, and darkness. Unlike traditional marine robots that require specialized metallic vessels for protection, deep-sea species thrive without such cumbersome pressure-resistant designs. Their pressure-adaptive forms, unique propulsion methods, and advanced senses have inspired innovation in designing lightweight, compact soft machines. This perspective addresses challenges, recent strides, and design strategies for bioinspired deep-sea soft robots. Drawing from abyssal life, it explores the actuation, sensing, power, and pressure resilience of multifunctional deep-sea soft robots, offering game-changing solutions for profound exploration and operation in harsh conditions.

What we look for at Science Robotics.

Science Robotics welcomes papers demonstrating technical and scientific advances, with potential for influence beyond robotics.

Evaluation of the Electroglottographic Signal Variability in Organic and Functional Dysphonia.

OBJECTIVES: To confirm the data reported in our previous studies on the analysis of the variability of the electroglottographic signal in the pathological voice; to evaluate possible differences in variability between organic and functional pathologies; to identify any distinctive/typical EGG patterns for these pathologies. METHODS: One hundred twenty-five subjects were enrolled (36 euphonic and 89 pathological: 24 functional dysphonia, 21 bilateral vocal nodules, 23 unilateral polyps and 21 unilateral cysts). All subjects were studied with videolaryngostroboscopy, spectrographic analysis of voice and electroglottography (EGG). The EGG signal variability was then investigated using amplitude-speed combined analysis, by means of a proprietary software algorithm. Amplitude and Speed variation were expressed as a new parameter, the Variability Index (VI), calculated both for the whole EGG signal recorded (VI-tot) and in each phase of the glottic cycle (VI-Q, absolute value; VI-Q%, percentage value). RESULTS: In the comparison of VI values between pathological and normal groups, VI-tot and VI-Q2% (which corresponds to the final phase of vocal fold contact) were significantly greater in pathological subjects (P= 0.002). The comparison of VI values among subgroups of the various pathologies showed a difference for VI-tot (P< 0.0001) and VI-Q2% (P= 0.001); this difference was more marked in the cysts than in the functional dysphonia. The cut-off values of VI-tot and VI-Q2% were 0.191 and 18.17%, respectively (sensitivity and specificity 65.2% and 66.7% for VI-tot and 84.3% and 77.8% for VI-Q2%). CONCLUSIONS: The variability of the EGG signal investigated through the combined analysis of the amplitude and the speed of vibration using a proprietary algorithm software has proved useful not only to distinguish the normal voice from the pathological voice, but also to characterize which phases are more altered in the various voice pathologies studied, both functional and organic. Furthermore, the analysis of the VI parameter allowed to propose cut-off values characterized by a good sensitivity and specificity to discriminate dysphonia from the euphonic voice. Larger groups of patients will be needed to confirm these results.

Distinct neural systems involved in agency and animacy detection.

We designed an fMRI experiment comparing perception of human faces and robotic faces producing emotional expressions. The purpose of our experiment was to investigate engagement of different parts of the social brain by viewing these animate and inanimate agents. Both human and robotic face expressions evoked activity in face-responsive regions in the fusiform gyrus and STS and in the putative human mirror neuron system. These results suggest that these areas mediate perception of agency, independently of whether the agents are living or not. By contrast, the human faces evoked stronger activity than did robotic faces in the medial pFC and the anterior temporal cortex--areas associated with the representation of others' mental states (theory of mind), whereas robotic faces evoked stronger activity in areas associated with perception of objects and mechanical movements. Our data demonstrate that the representation of the distinction between animate and inanimate agents involves areas that participate in attribution of mental stance.

Non-invasive study of Octopus vulgaris arm morphology using ultrasound.

Octopus arms are extremely dexterous structures. The special arrangements of the muscle fibers and nerve cord allow a rich variety of complex and fine movements under neural control. Historically, the arm structure has been investigated using traditional comparative morphological ex vivo analysis. Here, we employed ultrasound imaging, for the first time, to explore in vivo the arms of the cephalopod mollusc Octopus vulgaris. Sonographic examination (linear transducer, 18 MHz) was carried out in anesthetized animals along the three anatomical planes: transverse, sagittal and horizontal. Images of the arm were comparable to the corresponding histological sections. We were able, in a non-invasive way, to measure the dimensions of the arm and its internal structures such as muscle bundles and neural components. In addition, we evaluated echo intensity signals as an expression of the difference in the muscular organization of the tissues examined (i.e. transverse versus longitudinal muscles), finding different reflectivity based on different arrangements of fibers and their intimate relationship with other tissues. In contrast to classical preparative procedures, ultrasound imaging can provide rapid, destruction-free access to morphological data from numerous specimens, thus extending the range of techniques available for comparative studies of invertebrate morphology.

Smarter materials for smarter robots.

Advances in materials science will blur the boundaries between robots and the materials from which they are composed.

Sharpness recognition based on synergy between bio-inspired nociceptors and tactile mechanoreceptors.

Touch and pain sensations are complementary aspects of daily life that convey crucial information about the environment while also providing protection to our body. Technological advancements in prosthesis design and control mechanisms assist amputees to regain lost function but often they have no meaningful tactile feedback or perception. In the present study, we propose a bio-inspired tactile system with a population of 23 digital afferents: 12 RA-I, 6 SA-I, and 5 nociceptors. Indeed, the functional concept of the nociceptor is implemented on the FPGA for the first time. One of the main features of biological tactile afferents is that their distal axon branches in the skin, creating complex receptive fields. Given these physiological observations, the bio-inspired afferents are randomly connected to the several neighboring mechanoreceptors with different weights to form their own receptive field. To test the performance of the proposed neuromorphic chip in sharpness detection, a robotic system with three-degree of freedom equipped with the tactile sensor indents the 3D-printed objects. Spike responses of the biomimetic afferents are then collected for analysis by rate and temporal coding algorithms. In this way, the impact of the innervation mechanism and collaboration of afferents and nociceptors on sharpness recognition are investigated. Our findings suggest that the synergy between sensory afferents and nociceptors conveys more information about tactile stimuli which in turn leads to the robustness of the proposed neuromorphic system against damage to the taxels or afferents. Moreover, it is illustrated that spiking activity of the biomimetic nociceptors is amplified as the sharpness increases which can be considered as a feedback mechanism for prosthesis protection. This neuromorphic approach advances the development of prosthesis to include the sensory feedback and to distinguish innocuous (non-painful) and noxious (painful) stimuli.

Five years of Science Robotics.

Looking back at the last 5 years of Science Robotics and looking forward to the next 5.

A vision for future bioinspired and biohybrid robots.

Bioinspired and biohybrid robots can help respond to diverse, sustainable application needs.

Cerebellar adaptive mechanisms explain the optimal control of saccadic eye movements.

Cerebellar synaptic plasticity is vital for adaptability and fine tuning of goal-directed movements. The perceived sensory errors between desired and actual movement outcomes are commonly considered to induce plasticity in the cerebellar synapses, with an objective to improve desirability of the executed movements. In rapid goal-directed eye movements called saccades, the only available sensory feedback is the direction of reaching error information received only at end of the movement. Moreover, this sensory error dependent plasticity can only improve the accuracy of the movements, while ignoring other essential characteristics such as reaching in minimum-time. In this work we propose a rate based, cerebellum inspired adaptive filter model to address refinement of both accuracy and movement-time of saccades. We use optimal control approach in conjunction with information constraints posed by the cerebellum to derive bio-plausible supervised plasticity rules. We implement and validate this bio-inspired scheme on a humanoid robot. We found out that, separate plasticity mechanisms in the model cerebellum separately control accuracy and movement-time. These plasticity mechanisms ensure that optimal saccades are produced by just receiving the direction of end reaching error as an evaluative signal. Furthermore, the model emulates encoding in the cerebellum of movement kinematics as observed in biological experiments.