Keywords: Human-machine interfaces, Machine learning and reinforcement learning, Neuro robotics
Abstract: Brain Machine Interfaces (BMI) are used to establish a communication pathway between the human brain and machines. Using BMI, signals from the brain are transmitted to an external processing unit where they are decoded into meaningful actions (e.g., browsing the internet using a PC or grasping an object with a prosthetic hand). BMIs are used to increase intuitiveness of the control of technical devices that can help individuals with motor or sensory impairments to regain their lost dexterity or able-bodied people to augment their capabilities. In this work, we present an Electromyography (EMG) based method for decoding object motion in dexterous, in-hand manipulation tasks. To do that, we use EMG signals derived from specific muscles of the human hand and forearm, and an optical motion capture system that records the object motion. The decoding is formulated as a regression problem using the Random Forests methodology that is based on a combination of decision trees. The model was trained using time-domain features, namely: root mean square, waveform length and zero crossings. A 5-fold cross validation procedure is used for model assessment purposes. This preliminary study achieves significantly high estimation accuracies, proving that object motion can be directly decoded from myoelectric activations of the muscles of the human hand and forearm. This work can support the formulation of EMG based telemanipulation schemes for advanced robotic and prosthetic hands.

Keywords: Human-machine interfaces, Wearable devices, Prostheses
Abstract: Hand gesture recognition based on myoelectric (EMG) signals is an innovative approach for the development of intuitive interaction devices, ranging from poliarticulated prosthetic hands to intuitive robot and mobile interfaces. Their study and development in controlled environments provides promising results, but effective real-world adoption is still limited due to reliability problems. such as motion artifacts and arm posture, temporal variability and issues caused by the re-positioning of sensors at each use. In this work, we present an EMG dataset collected with the aim to explore postural and temporal variability in the recognition of arm gestures. Its collection of gestures executed in 4 arm postures over 8 days allows to evaluate the impact of such variability on classification performance. We implemented and tested state-of-the-art recognition approaches analyzing the impact of different training strategies. Moreover, we compared the computational and memory requirements of the considered algorithms, providing an additional evaluation criteria useful for real-time implementation. Results show a decrease in the recognition of inter-posture and inter-day gestures up to 20%. The provided dataset will allow further exploration of such effects and the development of effective training and recognition strategies.

Keywords: Design and control, Neuro robotics, Biomechanics
Abstract: This study presents a method of providing deep muscle stimulation with an interferential current via surface electrodes, and analyzes the response of the skeletal muscles and the pain sensation experienced by subjects. Three healthy adult males were subjected to isometric contraction to generate the interferential current in the soleus muscle group, and the degree of muscle contraction by the plantarflexion torque and the pain sensation of the subjects were evaluated. Two square waves with different pulse widths were generated at the same time, and acted as the stimulation current with a relatively small amplitude of +-16 mA. The experiments confirmed that a larger output torque could be obtained by generating interferential waves, compared to the condition where two identical waves (non-interferential current) were used. In addition, it was confirmed that the output torque and the pain sensation of the subjects decreased with the increase in the average frequency of the interferential current. When an interferential current was generated under the condition of interferential frequency f_Low = 30 Hz and average frequency f_High = 50 - 100 Hz with the proposed electrode arrangement, subjects produced the largest output without pain. The results can contribute to assisting ankle pushoff movement in walking during the rehabilitation of patients with central motor paralysis.

Keywords: Human-machine interfaces, Design and control
Abstract: Estimation of a model that relates the surface electromyography (sEMG) to the joint angle is of major importance to design human robot interfaces, to control biomimetic mechanisms and to model muscular systems. However, it is challenging to obtain information from electromyographic signals due to nonlinearities, unmodeled muscle dynamics, noise and interference. In this paper, we applied system identification methods to estimate the parameters of a Hammerstein Wiener model that relates sEMG from three muscles of the arm (biceps brachii, triceps brachii and brachioradialis) to the elbow joint angle. In order to determine an estimation model and a calibration procedure, a set of experiments were carried out with two subjects. The experiment consisted of a series of elbow flexion and extension movements with different weights (0kg, 1.5kg and 3kg) and two different types of movement: continuous and intermittent, with controlled velocity. The sEMG and angular data were recorded during the experiment. The chosen model for system identification was the Hammerstein-Wiener with Wavelet Network as input and output nonlinearities. A second experiment was conducted in a different day in order to validate the estimated models. The results show a procedure to estimate a model for the EMG-to-Joint Angle, showing high correlations and coefficient of determination as well as low Root-Mean-Square Errors (RMSE) with respect to the measured data: correlations of 94.90 pm 3.92%, coefficients of determination of 0.8814 pm 0.0978 and RMSE values of 10.82 pm 3.73^circ. These models are interesting candidates to help in the control of actuated prosthetic and orthotic devices.

Keywords: Human-machine interfaces, Design and control, Locomotion and manipulation in robots and biological systems
Abstract: Robotic algorithms that augment movement errors have been proposed as promising training strategies to enhance motor training and neurorehabilitation. However, research effort has mainly focused on rehabilitation of upper limbs. In this study, we investigated the effect of training with novel error modulating strategies on learning an asymmetric gait pattern. Thirty healthy young participants walked in the robotic exoskeleton Lokomat, while learning a foot target-tracking task which required an increased hip and knee flexion in the dominant leg. Learning with three different strategies was evaluated: (i) No guidance: no disturbance/guidance was applied, (ii) Haptic error amplification: dangerous and discouraging large errors were limited with haptic guidance, while awareness of task relevant errors was enhanced with error amplification, and (iii) Visual error amplification: visually perceived errors were amplified in a virtual reality environment. We also evaluated whether increasing the movement variability during training by adding randomly-varying haptic disturbances on top of the other training strategies further enhanced learning. We found that training with the novel haptic error amplification strategy limited large errors during training, did not hamper learning and enhanced transfer of the learned asymmetric gait pattern. Training with visual error amplification, on the other hand, increased errors during training and hampered motor learning. Adding haptic disturbances did not have a significant effect on learning. The novel haptic error modulating controller that amplifies small task-relevant errors while limiting large errors provided the best framework to enhance motor learning.

Keywords: Human-machine interfaces, Dynamics and control
Abstract: In this work we present a V-REP simulator for the da Vinci Research Kit (dVRK). The simulator contains a full robot kinematic model and integrated sensors. A robot operating system (ROS) interface has been created for easy use and development of common software components. Moreover, several scenes have been implemented to illustrate the performance and potentiality of the developed simulator. Both the simulator and the example scenes are available to the community as an open source software.

Keywords: Prostheses, Design and control, Human-centered design
Abstract: The various drawbacks of complex myoelectric prosthetic hands have led to low adoption rates by upper limb amputees, with upwards of 35% pediatric device rejection and 23% adult device rejection rates. This paper describes the design of a novel 50th percentile female sized single-actuator anthropomorphic myoelectric hand that was created to address issues commonly associated with myoelectric prosthesis. The hand uses underactuated coupling mechanisms to enable three distinct passively-adaptive grasp types, from a single DC motor and single input control, selected through a simple manual reconfiguration of the digits. The hand was evaluated in a preliminary assessment that included benchtop evaluation and a five subject able-bodied combined Box and Blocks and SHAP test using a bypass socket to evaluate effectiveness on activities of daily living. The able-bodied subject testing results are presented and compared to published SHAP results from commercial powered hook, single-actuator anthropomorphic and multi-actuator anthropomorphic devices.

Keywords: Flexible instruments, Design
Abstract: Single port and natural orifice surgeries gain popularity compared to multi-port ones. The main reason is that single port surgeries are less invasive and thus less harmful to the patient than multiport surgeries. However, one of the drawbacks of devices for single port surgery is their limited access to regions proximate to the point of entrance. Moreover, they are strongly limited in choosing (controlling) the end-effector’s orientation. Thus, inspecting a point of interest inside the endoscope’s workspace from different angles is challenging. To overcome the existing limitations, we propose an EtherCAT-based master-slave robotic system. This master-slave system consists of an articulated endoscope with 2 work channels that can be manipulated by a force-sensing joystick. The articulated endoscope is capable of realizing a large range of orientations (max ±220°) throughout the workspace, can turn back, and look upon itself within a radius of 40 mm. The suggested robotic system will be useful in single port interventions such as inspections and treatments of the cardiac sphincter, the fundus in the stomach, or the neck of the bladder. In this paper, we present the design of the endoscope, analyze its kinematics, and present the workspace including the feasible orientations. Finally, the functionality of the master-slave system EndoCAT is demonstrated in 2 experiments on a mockup stomach.

Keywords: Flexible instruments, Biologically-inspired systems, Surgical navigation
Abstract: The improved natural hemodynamics offered by mitral valve (MV) repair strategies aims to prevent heart failure and to minimize the use of long-term anticoagulant. This combined with reduced patient trauma offered by minimally invasive surgical (MIS) interventions, requires an increase in capabilities of MIS MV repair. The use of robotic catheters have been described in MIS applications such as navigational tasks, ablation and MV repair. The majority of the robotic catheters are evaluated in testbeds capable of partially mimicking the cardiac environment, while validation in a clinical scenario is associated with significant preparation time and limited availability. Therefore, catheter development could be aided by an accessible and available testbed capable of reproducing beating heart motions, circulation and the relevant anatomy in MIS cardiovascular interventions. In this study, we contribute a beating heart testbed for the evaluation of robotic catheters in MIS cardiovascular interventions. Our work describes a heart model with relevant interior structures and an integrated realistic MV model, which is attached to a Stewart platform to reproduce beating heart motions based on pre-operative patient data. The heart model is extended with an aortic valve, a systemic arterial model, a venous reservoir and a pulsatile pump to mimic the systemic circulation. Experimental evaluation showed systemic circulation and beating heart motion reproduction for 70 BPM with a mean absolute distance error of 1.26 mm, while a robotic catheter in the heart model is observed by ultrasound imaging and electromagnetic position tracking. Therefore, the presented testbed is capable of evaluating MIS robotic cardiovascular interventions such as MV repair, navigation tasks and ablation.

Keywords: Exoskeletons, Human-machine interaction, Wearable and augmenting devices
Abstract: Exoskeletons have shown their ability to assist locomotion and augment human performances. However, the benefits of wearing these devices depend on how effectively power can be transmitted from the device to the user's biological structures. Recent studies have shown evidence of inefficient power transmission, with losses of up to 50%. The problem of power transmission can be mitigated by designing interfaces that increase contact stiffness and reduce relative motion between the limb and the robot. In this contribution, the design and development of physical interfaces for rigid lower limb exoskeletons is presented. The relative motion between the human and the developed interface is evaluated using a motion capture system and compared to the performances of a commercially available interface. Results indicate a clear reduction in relative motion between the user and the exoskeleton when the customized interface is worn.

Keywords: Wearable and augmenting devices, Exoskeletons, Novel actuators
Abstract: Abstract—Hand exoskeletons could potentially improve hand use after stroke but are typically obtrusive and non-intuitive. Here, we provide a rationale for a control strategy suitable for a minimalistic hand exoskeleton. We also report on pilot testing of the strategy, and on a soft actuator design for implementing the strategy. The strategy is based on four experimental observations from studies conducted in our laboratory with unimpaired individuals and stroke survivors. First, using only a pinch grip, unimpaired people can achieve a substantial level of hand function when measured with clinical assessments. Second, the level of achieved function corresponds well with what is necessary to drive daily hand use, as measured by a novel wearable sensor with stroke survivors. Third, even people with severe hand impairment after stroke have a well-preserved ability to control isometric finger flexion force, even though they cannot use their hand to manipulate objects. Fourth, such individuals also exhibit highly correlated forces between fingers. From these observations we propose a control strategy that measures residual finger flexion of digits 3-5 (middle-pinky fingers) to control the force of an exoskeleton assisting in pinch grip. We implemented this “residual force control” strategy using the FINGER exoskeleton and found that unimpaired subjects could intuitively use this strategy to pick up an object and learn to amplify their grip force (Repeated Measures ANOVA, p < .004). We have also begun developing a soft exoskeleton to implement the residual force control strategy, and we report on the actuator design here. The results of preliminary testing of the actuator show that the actuator could produce a sufficient amount of force (13.06 N ± .33 SD) to assist the hand.

Keywords: Mobility, Technology assessment, Neurological disease
Abstract: Service robots have the potential to support the personal mobility of elderly population. Monitoring and measuring grasping force in older adults is an important issue both from robotic and clinical perspectives. From robot point of view, new adaptive control strategies can be implemented based on the users’ force; clinicians can monitor the changes in the grasp strength over time to evaluate abnormal conditions, which can be associated with geriatric syndromes. In this context, this work focused on the design, development and testing of a sensorized smart handle able to enhance the robotic mobility support service provided by the robot, called ASTRO. The primary goal of this paper is to design the sensorized handle according to clinical and technical specifications in terms of working range, sensitivity and clinical requirements. Then, the smart handle was tested with 19 subjects to investigate whether the system is able to detect forces correlated to the ones measured with a traditional tool. Additionally, further analysis were conducted to analyse how the forces were distributed to refine and optimize the design. The study shows meaningful results as the grasp forces measured with the smart handle and the traditional tool were significantly correlated

Keywords: Human-machine interaction, Human-machine interfaces, Exoskeletons
Abstract: The design of comfortable and effective physical human robot interaction (pHRI) interfaces for force transfer is a prominent challenge for coupled human-robot systems. Forces applied by the robot at the fingers create reaction forces on the dorsal surface of the hand, often leading to high pressure concentrations which can cause pain and discomfort. In this paper, the interaction between the pHRI interface and the dorsal surface of the hand is systematically characterized, and a new method for the design of comfortable interfaces is presented. The variability of the stiffness of the hand dorsum is quantified experimentally, and this data is used to minimize the peak pressure exerted on the hand dorsum, by varying the stiffness profile of the pHRI interface. This optimized design is demonstrated to improve the pressure distribution over the hand dorsum where the robot is attached to the hand. Additionally, to enable informed design choices, the effects of varying the stiffness of the pHRI interface on relative displacement between the robot and the hand dorsum are also characterized. This optimization approach to designing pHRI interface can be extended to different limbs as well, especially when there is a transfer of high moment loads to the human body, provided the appropriate stiffness data is available.

Keywords: Exoskeletons, Design and control, Human-machine interaction
Abstract: Exoskeletal systems become more important for the rehabilitation of people with lower limb paralysis. Today these systems are mainly used in clinical settings, but also devices to support activities of daily living are becoming reality. The uncontrolled environment of home use leads to new design challenges and loading situations that are not yet well known. In this paper a novel orthotic system that supports people with lower limb paralysis in their everyday life is introduced. To learn more about the mechanical stress on the device, the knee joint unit was equipped with various sensors to measure motion, forces and power in real-world situations. The influence of the user’s residual function on these parameters was investigated. First results from a clinical trial with 3 paralyzed patients show that knee power is in the same range as for healthy humans, with peak values up to 5.9 W/kg. Peak torque on the knee joint can be as high as 1.8 Nm/kg. There are large differences between patients depending on their diagnosis. All systems were exposed to impacts with accelerations higher than 85 m/s² at exceptional events. These results show that patient pathology significantly influences system loads, and that mechanical robustness is important for the design of supportive systems.

Keywords: Neuroengineering, Neuro robotics, Human-machine interfaces
Abstract: In this paper we compare three approaches to solve the hand-eye and robot-world calibration problem, for their application to a Transcranial Magnetic Stimulation (TMS) system. The selected approaches are: i) non-orthogonal approach (QR24); ii) stochastic global optimization (SGO); iii) quaternion-based (QUAT) method. Performance were evaluated in term of translation and rotation errors, and computational time. The experimental setup is composed of a 7 dof Panda robot (by Franka Emika GmbH) and a Polaris Vicra camera (by Northern Digital Inc) combined with the SofTaxic Optic software (by E.M.S. srl). The SGO method resulted to have the best performance, since it provides lowest errors and high stability over different datasets and number of calibration points. The only drawback is its computational time, which is higher than the other two, but this parameter is not relevant for TMS application. Over the different dataset used in our tests, the small workspace (sphere with radius of 0.05m) and a number of calibration points around 150 allow to achieve the best performance with the SGO method, with an average error of 0.83±0.35mm for position and 0.22±0.12deg for orientation.

Keywords: Dynamics and control, Biologically-inspired systems, Locomotion and manipulation in robots and biological systems
Abstract: This paper proposes a nonlinear control of aerial attitude regulation for hopping robots that have a freely-rotating body with a pair of actuated arms that are connected to each other. These kinds of robots are so-called underactuated robots. The proposed controller is based on an output zeroing control that stabilizes angular momentum and the body pitch angle to be zero. Compared with attitude stabilizing methods using axisymmetric reaction wheels driven by a linear control law, utilizing arms for attitude compensation, which is not axisymmetric, will achieve more natural motions like animals. As the first stage of the development, a controller has been designed for an underactuated inverted pendulum model, where the robot body is supported a freely-rotating joint and the arms is actuated by a motor. The control law is simplified based on simulation analysis so as to be implemented to low-cost microcomputers. To show the validity of the proposed method, an inverted-pendulum-type robot with a pair of arms driven by an inexpensive hobby-use motor has been prototyped, the attitude of which is measured by using a 9-axis motion sensor, and the control law has been implemented to a microcomputer, Renesas RX63N (GR-SAKURA II). Experiments to control the pitch angle of the robot has been conducted, and the experimental results showed that the proposed method well works to stabilize the amplitude of the pitch angle to be less than about ±13 degrees.

Keywords: Locomotion and manipulation in robots and biological systems, Biologically-inspired systems
Abstract: In this paper an insect-inspired body size learning algorithm is adopted in a humanoid robot and a control system, mainly developed with spiking neurons, is proposed. It implements an evaluation of distances by using the typical parallax method performed by different insect species, such as Drosophila melanogaster. A Darwin-OP robot was used as test-bed to demonstrate the potential application of the learning method on a humanoid structure. The robot, equipped with a hand extension, was free to move in an environment to discover objects. As consequence, it was able to learn, using an operant conditioning, which objects can be reached, via the estimation of their distance on varying the length of the equipped tool. The learning scheme was tested both in a dynamical simulation environment and with the Darwin-OP robot.

Keywords: Locomotion and manipulation in robots and biological systems, Dynamics and control, Surgical navigation
Abstract: Intracochlear electrode array insertion is a crucial process for cochlear implant surgery. However, the behavior of the intracochlear electrode array during the insertion remains unclear to surgeons. In order to minimize or eliminate the trauma induced by electrode array insertion, we propose an electrode capacitive sensing method to sense the behaviors of the electrode array during the robotic insertion process. To this end, we take a single capacitance measurement between electrode pair 1 and 2 during the robotic insertion and show experimentally that capacitance signal curves are systematically affected by intracochlear forces between the scala tympani wall and the contact electrode. Therefore, electrode capacitance measurements help track the motion between the electrode array and the cochlear lateral wall during surgeries.

Keywords: Biologically-inspired systems, Manipulation
Abstract: Rodents like rats and mice use their mystacial vibrissae for tactile perception. Several information of objects varying in (geometrical) size can be detected. For instance, the animals are able to recognize the shape of an object as well as to determine very fine surface textures by contacting the object with their vibrissae. The vibrissal kinematics differ in these tasks, the vibrissae can be brushed/swept against an object (protraction and retraction) or they can be dabbed against it. Here, a vibrissa inspired sensor (cylindrical spring steel wire) is swept along a rectangular object. Examining the influence of the displacement of the sensor support on the measured signals, translational and rotatory scanning scenarios are analyzed in simulations and experiments. In a first step, a frictionless contact between sensor and object is assumed. Then, friction is taken into account. In dependence on the scanning scenario the measured signal is amplified. Furthermore, the orientation of the sensor within one sweep is important in case of the frictional contact. The results imply that the scanning scenario and sensor orientation can be used to tune the measured signals.

Keywords: Wearable devices, Wearable and augmenting devices
Abstract: To precisely track human motion, today's state-of-the-art employs either well-calibrated sensors tightly strapped to the body or high-speed cameras confined to a finite capture volume. These restrictions make such systems less mobile. In this paper, we aim to break this usability barrier around motion-capture technology through a wearable system that has sensors integrated directly into garments. We develop a pose-estimation approach based on classic kinematics and show that it is insufficient to analyze motion in such a system, leading to mean Euler angle errors of up to +-60 degrees and standard deviations of 120degrees. Thus, we motivate the need for data-driven algorithms in this domain. Through a quantitative study, we attribute motion-estimation errors to the high-degree of sensor displacement (up to 118 degrees standard deviation from the nominal value) with respect to the body segments that are present when human poses change. Based on controlled experiments, we develop a new dataset for such systems comprising over 3 hours of biomechanical motion recordings from 215 trials on 12 test subjects.

Keywords: Human-machine interaction, Human-machine interfaces, Neural networks
Abstract: This paper introduces a method which uses feedforward neural networks (FNNs) for estimating gait cycle progress using data recorded from inertial and muscle activity sensors attached to one side of the lower body. Three-axis inertial measurement unit (IMU) readings from accelerometers and gyroscopes located above the outer ankle and knee were fused with mechanomyogram (MMG) sensor readings from across major muscle groups on the left leg. Validation was against ground truth gathered concurrently with VICON motion capture. The performance was characterised by rms error (Erms) and max error (Emax), averaged across four cross-validated trials, and enhanced by adjusting number of sliding window frames and hidden layer neurons. The final configuration estimated gait cycle progress with Erms of 1.6% and Emax of 6.8%. This demonstrates promise for such a method to be used for control of unilateral robotic prostheses and exoskeletons, providing state estimation of gait progress from low power sensors limited to one side of the lower body.

Keywords: Wearable and augmenting devices, Human-machine interaction, Human-machine interfaces
Abstract: Contextualization: Identification of gait events is a key step towards enhancing control of robotic devices for rehabilitation and/or assistance of lower limb function. Gap: Several approaches and techniques that can properly identify gait events and phases have been developed and successfully tested in healthy subjects, but current algorithms had limited success for impaired gait. Purpose: Here we studied the feasibility of real-time identification of gait phases for impaired subjects using a single inertial sensor on the paretic foot. Methodology: We carried out a pilot experiment evaluating seven algorithms proposed in the literature for detection of two main events (heel-strike [HS] and toe-off [TO]) for normal and hemiparetic gait. Results: We obtained a high performance for healthy gait indicating their suitability for real-time implementation for that population. However, we obtained a lower accuracy for all the algorithms during hemiparetic-like gait. Conclusions: We performed a comprehensive review of the literature and evaluated the current algorithms in gait segmentation using a single inertial sensor on foot or shank for both healthy and paretic gait. Existing algorithms worked as expected in healthy but not paretic gait.

Keywords: Neuro robotics, Human-machine interaction, Neurological disease
Abstract: Rehabilitation robotics and neuromuscular stimulation have become widespread technologies for rehabilitation training of stroke and spinal cord injured patients. In this context, real-time tracking of the performed motion facilitates real-time control of the motion support and biofeedback about undesired compensatory motions. We consider a cable-driven robotic system for upper limb rehabilitation and extend it by two wearable inertial sensors. By sensor fusion of the robotic and inertial measurements, we obtain accurate estimates of the forearm and upper arm orientation and position, which cannot be obtained by either of both measurement systems alone. A real-time biofeedback is introduced to prevent undesired compensatory motions of the trunk and shoulder. The proposed methods are evaluated with respect to an optical reference system in a series of experimental trials with and without compensatory motions. Using only the robotic sensors yields average measurement errors of up to 24 cm for the shoulder position and 19° for the elbow angle. In contrast, the proposed hybrid sensor fusion achieves accuracies better than 6 cm and 4°, respectively.

Keywords: Wearable devices, Machine learning and reinforcement learning, Technology assessment
Abstract: Quantitative gait assessment typically involves optical motion capture systems and force plates, which result in high operating costs. Footwear-based motion tracking systems can provide a portable and affordable solution for real-time gait analysis in unconstrained environments. However, the relatively low accuracy of these systems still represents a barrier to their widespread use. In this paper, we show that linear and learning-based regression models can substantially improve the raw estimates of a set of kinematic gait parameters obtained with instrumented insoles (SportSole) from a group of N=9 healthy subjects who walked at different speeds. Least Absolute Shrinkage and Selection Operator (LASSO) and Support Vector Regression (SVR) models are compared in terms of accuracy, precision, and robustness to change in gait speed, using gold-standard equipment to generate reference data. Results indicate that SVR is superior to LASSO. Indeed, the mean absolute errors (MAE) in stride length, velocity and foot-ground clearance were 1.28+/-0.19%, 1.62+/-0.42% and 3.72+/-0.87% for LASSO, 1.06+/-0.08%, 1.13+/-0.08% and 3.00+/-0.87% for SVR, respectively. These findings provide further evidence that footwear-based systems may represent valid alternatives to laboratory equipment for assessing a basic set of gait parameters in unconstrained environments.

Keywords: Technology assessment, Biomechanics, Wearable devices
Abstract: An increasing diversity of available motion capture technologies allows for measurement of human kinematics in various environments. However, little is known about the differences in quality of measured kinematics by such technologies. Therefore, this work presents a comparison between three motion capture approaches, based on inertial-magnetic measurement units (processed with Xsens MVN Analyze) and optical markers (processed using Plug-In Gait and OpenSim Gait2392). It was chosen to evaluate the different motion capture approaches in running, as such kinematics are preferably measured in the natural running environment and involve challenging dynamics. An evaluation was done using data of 8 subjects running on a treadmill at three different speeds, namely 10, 12 and 14 km/h. The sagittal plane results show excellent correlation (ρ > 0.96) and RMSDs are smaller than 5 degrees for 6 out of the 8 subjects. However, results in the frontal and transversal planes were less correlated between the different motion capture approaches. This shows that sagittal kinematics can be measured consistently using any of the three analyzed motion capture approaches, but ambiguities exist in the analysis of frontal and transversal planes.

Keywords: Machine learning and reinforcement learning, Human-machine interfaces, Locomotion and manipulation in robots and biological systems
Abstract: Human-robot interaction (HRI) for gait rehabilitation would benefit from data-driven gait models that account for gait phases and gait dynamics. Here we address the current limitation in gait models driven by kinematic data, which do not model interlimb gait dynamics and have not been shown to precisely identify gait events. We used Switched Linear Dynamical Systems (SLDS) to model joint angle kinematic data from healthy individuals walking on a treadmill with normal gaits and with gaits perturbed by electrical stimulation. We compared the model-inferred gait phases to gait phases measured externally via a force plate. We found that SLDS models accounted for over 88% of the variation in each joint angle and labeled the joint kinematics with the correct gait phase with 84% precision on average. The transitions between hidden states matched measured gait events, with a median absolute difference of 25ms. To our knowledge, this is the first time that SLDS inferred gait phases have been validated by an external measure of gait, instead of against predefined gait phase durations. SLDS provide individual-specific representations of gait that incorporate both gait phases and gait dynamics. SLDS may be useful for developing control policies for HRI aimed at improving gait by allowing for changes in control to be precisely timed to different gait phases.

Keywords: Biologically-inspired systems, Haptics, Soft robotics
Abstract: Rats rely heavily on tactile information from their whiskers to acquire information about their surroundings. A whisker has no sensors along its length. Instead, mechanical deformation of the whisker is sensed via receptors at its base. The present study introduces a micro-sensor developed specifically to imitate the sensing of biological rat whiskers. The sensor responds to bending moments resulting from touch and/or airflow in two axes. The sensor was designed based on analytical models from cantilever beam theory, and the models were validated with finite-element analysis. Sensors were then fabricated using micro-milled molds and integrated into an Arduino-based circuit for simple signal acquisition. The present work begins to develop the technology to allow investigation of important engineering aspects of the rat vibrissal system at 1x scale. In addition to its potential use in novel engineering applications, the sensor could aid neuroscientists in their understanding of the rat vibrissal-trigeminal pathway.

Keywords: Biomechanics, Wearable devices, Manipulation
Abstract: Force Sensing Resistors (FSRs) are obtained by randomly dispersing conductive nanoparticles in an insulating polymer matrix. The low–profile and low–weight characteristics of FSRs have been exploited in the manufacturing of pressure–mapping insoles. However, FSRs exhibit low repeatability which limits the extensive usage of such devices in certain applications that require high levels of accuracy and repeatability. In order to enhance the performance of FSRs, a custom–designed insole has been manufactured from commercial FSRs. The purpose of this study is to assess the influence of the sourcing voltage (Vs) and the driving circuit on the repeatability of pressure measurements when collecting gait data. It was found that setting a low Vs improves the repeatability of the system, whereas setting a large sourcing voltage causes a degradation of sensors’ sensitivity. Some possible reasons for sensitivity degradation are discussed on the basis of modelling sensor dynamics with equations for quantum tunneling conduction and constriction resistance. Enhancing the repeatability of FSRs opens the way for a broader usage of such devices in applications demanding accurate pressure measurements, such as: rehabilitation and object manipulation in robotics.

Keywords: Prostheses, Wearable devices, Haptics
Abstract: Using a hand prosthesis means grasping without tactile information. Although supplementary sensory feedback has been investigated extensively, few study results could translate into clinical applications. Unreliable and imprecise feedforward control of current hand prostheses hinders the investigation of supplementary sensory feedback, so an ideal feedforward tool should be used. Thus, we aimed to create a device that would allow to use the sensory deprived human hand as an ideal tool without the need for local anesthesia. For this, we fashioned silicone digit extensions with integrated force sensors and tested the performance of 12 volunteers in grasping with these extensions. Two tests were performed: a simple pick and lift test to compare performance to anesthetized digits, and a virtual egg test to assess grasping efficiency. We found that the extensions significantly alter grasping. In future studies, these extensions will help us investigate how to artificially restore the information necessary for successful and efficient grasping with an ideal feedforward tool.

Keywords: Wearable devices, Locomotion and manipulation in robots and biological systems, Mobility
Abstract: Gait analysis has been considered in various scenarios to provide information about the ambulatory physical activity. In this regard, studying gait phases can provide valuable information about the quality of gait. Force myography (FMG) techniques have been successfully employed to detect gait events using pattern recognition methods. This paper explores how the accuracy of detecting gait phases is correlated with the parameters of gait and FMG signal. To this end, FMG data were collected from 11 volunteers walking on a treadmill with a custom-designed FMG ankle band. The collected FMG data were classified into four gait phases using Linear Discriminant Analysis (LDA) algorithm. The correlation between the error in classification and the parameters of gait and FMG signal was then investigated. The results show that in comparison with other studied parameters, variations in stride length have the most impact on the accuracy of gait phase classification with a coefficient of determination (R2) of 0.80. Such an effect is more pronounced when signal power-related features, such as root mean square (RMS), are used in the classification algorithm. This study provides insight into the factors affecting the accuracy of FMG-based techniques for gait analysis and is a preliminary step towards developing high performance FMG-based wearable ambulatory activity monitoring systems.

Keywords: Haptics, Dynamics and control
Abstract: Humans have multiple ways to adapt their arm dynamics to the task they have to perform. One way of doing this is through co-contraction of antagonist muscles. In telemanipulation this ability is easily lost due to time delays, quantization effects, bandwidth or hardware limitations. In this work a new concept for telemanipulation is presented. The end-point stiffness of a (simulated) telerobot is controlled via a variable impedance controller. The end effector stiffness scales with an estimate of the co-contraction around the elbow of the teleoperator. The telemanipulation concept was evaluated with ten subjects that performed two telemanipulation tasks in six different conditions. Three impedance levels: low, high, and variable, and two delay settings. The first task was on positioning accuracy, the second task on impact minimization. We have shown that low and variable impedance performed significantly better on the force task than high impedance for both delay conditions. We have also shown that high and variable impedance performed better on the position task. In the position task only one delay setting yielded significant results. This shows that the human ability to control arm stiffness can effectively be transferred to a telemanipulated robot.

Keywords: Design and control, Force control, Human-machine interaction
Abstract: In this paper, we deal with the human-robot interaction control problem. Levels of actuation of the user are considered in the human-robot interaction model from a stochastic point of view. It is given in terms of a Markovian approach. Electromyographic signals are used to compute jump parameters between different levels of interaction. In this way, human neuromuscular system defines the behavior of the Markov chain. A unified approach composed by robust Kalman filter and robust regulator for discrete-time Markovian jump linear systems is proposed. Also, a serious game is used to generate visual feedback and promote the active participation of the user. Experimental results show high accuracy in the Markovian compliance control for a robotic platform applied in ankle rehabilitation.

Keywords: Prostheses, Prostheses control, Design and control
Abstract: In this paper, the goal is to develop a high level controller for active prosthetic feet that can continuously estimate the ankle motion based on the shank motion. The proposed controller does not require speed determination, gait percent identification, input data manipulation, look-up tables or switching rules. To do this, the Gaussian process (GP) regression is used. The performance of the controller has been tested for walking speed of 0.6, 0.9, 1.2, 1.4 and 1.6 m/s. The results showed that the controller had lower estimation quality when input was only shank angular velocity or shank angle. However, the aggregated angular velocity and angle input resulted in high output estimation quality. Furthermore, for each speed, the estimation quality was more acceptable when the controller was trained for it. Accordingly, when the high level controller was tested without previous training, the estimation quality was less acceptable

Keywords: Dynamics and control, Design and control
Abstract: During human gait, the mechanical impedances of lower limb joints are modulated to achieve better stabilization and energy efficiency. Therefore, modulating impedance is a promising method to assist the patients with impaired gait. In this paper, we have attempted to implement the impedance modulation of ankle joint by using an impedance control scheme based on functional electrical stimulation. The control scheme, position-based impedance control, contains desired impedance model and a robust inner loop position controller using time delay estimation. The feasibility of the proposed impedance modulation using impedance control was shown through an experiment.

Keywords: Locomotion and manipulation in robots and biological systems, Haptics
Abstract: Due to the increase in number of patients with significant gait deficits, the need for sophisticated tools to assist the patient to perform different kinds of locomotion training exercises is highly relevant. Ground walking platforms (GWP) are some of the new robotic gait rehabilitation systems that aim to simulate different ground trajectories for a patient (e.g. plane ground, hill ...etc.) with different haptic materials (e.g. water, sand, ice,...etc.). The system targeted in this study aims for providing the user with simulated ground reaction forces based on the users movements in a virtual reality environment by deploying two 6 Degree of Freedoms robotic footplates. This paper presents a part of such control study that aims to simulate the behavior and the interaction of the user, the GWP, and the virtual reality implemented in Unity3D. Using real data, results show good detection of interaction between foot and different medium, while the simulation of the robot gives actual results concerning the properties of simulated medium

Keywords: Design, Wearable devices, Prostheses
Abstract: Urinary incontinence (UI) is a dysfunction related to an involuntary urine leakage, mainly caused by a deficient action of the urethral sphincter muscles. When UI is particularly severe, it must be managed via invasive surgical procedures. The most popular artificial urinary sphincters (AUS) are surgically placed around the urethra (i.e. extra-urethral AUS) and they squeeze it whenever necessary to restore continence. Current solutions do not present a unisex design and are constituted of several components, which must be installed in a rather invasive way and which imply a non-comfortable device management and control. The aim of this paper is to present the design of novel AUS prototypes, all of them magnetically actuated and controlled, by focusing on a unisex design suitable for both female and male anatomies, and by reducing device dimensions and to minimize the implantation invasiveness. In particular, magnetically-activated AUS based on an endo-urethral and an extra-urethral approaches are proposed. Preliminary results demonstrated that magnetically-controlled AUS could be appropriate solutions for different typologies of patients and requirements, depending on endo- or extra-urethral design approach.

Keywords: Surgical navigation, Design, Computer vision
Abstract: This study aims to propose an innovative force and motion control system for use in the virtual surgical training simulators with force display. In surgical operations using a bone chisel, impact forces are applied to the bone by pounding the chisel with a mallet. To virtually represent this situation in a surgical training simulator, the force display device must exhibit high stiffness and be able to react instantaneously to the large impact force and quick acceleration. For this reason, the force display device with high stiffness and force and motion control system that react instantaneously to an impact force are developed for a surgical training simulator involving a bone chisel in this study. In this approach, this force display device is constructed with a ball-screw mechanism for translational motion and a biaxial rotation mechanism for rotational motion of the chisel while allowing high stiffness. However, the force display device's response can be reduced by increasing the mass of moving parts in the device, due to its high stiffness. Therefore, the 2 degree-of-freedom admittance control is proposed for reacting instantaneously to the impact force. The efficiency of the proposed device and control system is verified by creating a virtual experience to contact and split hard tissue using a chisel.

Keywords: Micro- and nano-robotics, Design and control, Locomotion and manipulation in robots and biological systems
Abstract: Recently, a wireless capsule endoscope with active locomotion has become an effective endoscopic method for diagnosis and treatment of diseases of gastrointestinal (GI) tract. Various modules such as biopsy and drug delivery were developed for the wireless capsule endoscope (WCE) to extend its application. In this paper, we present a marking module so-called tattooing module for WCE to localize the lesions and tumors in digestive organs before the laparoscopic surgery. The WCE with tattooing module is manipulated by an Electromagnetic Actuation (EMA) system, where a moderate magnetic field intensity is generated to drive the WCE reaching to a target of the digestive organs. The tattooing module is capable of stowing the needle inside the WCE’s body to avoid pathway organs damage during locomotion and extruding to puncture the target for tattooing. The magnetic field is controlled to activate the micro-reed switch and triggers a chemical reaction that generates gas pressure. The produced gas increases the pressure in the propellant room and pushes the piston to eject the ink into the target. The prototype of the tattooing capsule endoscope is fabricated with dimension of 13 mm in diameter and 33 mm in length. The working principle and the mechanism of the tattooing module are suggested and the feasibility test with the prototype is demonstrated through in-vitro experiments.

Keywords: Flexible instruments, Design
Abstract: Surgical robotics have helped surgeons for more than two decades using sophisticated and operation-based devices. Different kinds of surgical robots have been developed for specific purposes. In recent times, stiffness changing robots are in the spotlight due to their necessity. The interaction force on neighboring tissues during navigation to surgical target can be reduced owing to the low stiffness and the stiffness can be increased to provide high payloads as it reaches the surgical site. In this work, a 2-DOF soft robot with stiffness changing capability is presented for tumor removal in neurosurgery under MRI-guidance. A floating fixed-point approach is used that changes the physical length of the manipulator to decrease deflection and increase stiffness. Experimental results confirmed that the stiffness could be varied more than three and a half times than the initial value. The robot can bend 37.46° in right and left and 38.56° in up and down direction. The current version of the robot is joystick-operated and can be controlled manually. Finally, the manipulator is composed of MR-compatible materials allowing it to be used in MR-guided interventions.

Keywords: Design and control
Abstract: Ultrasound scanning provides a noninvasive solution for additional screening in breast cancer detection and could potentially be improved through utilization of a human-robot collaborative system. Alternative ultrasound modalities, such as elastography, offer promising improvements over current sonography cancer detection rates but require stability and knowledge of applied force. A human-robot scanning system could leverage the sonographer's capabilities to select transducer placement while the robot maintains stability during scanning. This paper presents a novel hybrid force velocity controller for ultrasound scanning which utilizes elastography to provide compliance feedback and improve controller performance. We explore the sensitivity of the elastography algorithm to initial elasticity assumptions and analyze the performance gains of compliance feedback. The results of our study show that our proposed controller provides a performance improvement even when poor initial tissue compliance estimates are used.

Keywords: Technology assessment, Soft robotics, Design and control
Abstract: The increasing use of soft materials in robotics applications requires the development of mathematical models to describe their viscoelastic and nonlinear properties. The traditional linear viscoelastic models are unable to describe nonlinear strain-dependent behaviors. This limitation has been addressed by implementing a piecewise linearization (PL) in the simplest viscoelastic model, the Standard Linear Solid (SLS). In this work, we aim to implement the PL in a more complex model, the Wiechert model and compare the stress response of both linearized models. Therefore, the experimental data from the stress relaxation and tensile strength tests of six rubber-based materials is used to approximate the spring and dashpot constants of the SLS and the Wiechert model. Prior to implement the PL into the stress-strain curve of each material, the stress response from the Maxwell branches must be subtracted from this curve. By using the parameters obtained from fitting the Wiechert model into the stress relaxation curve, the response of both linearized models was improved. Due to the selection of constitutive equations evaluated, the linearized SLS model described the stress-strain curve more accurately. Finally, this work describes in details every step of the fitting process and highlights the benefits of using linearization methods to improve known models as an alternative of using highly complex models to describe the mechanical properties of soft materials.

Keywords: Haptics, Neural networks, Human-machine interfaces
Abstract: Children with physical impairments face great challenges to play because of their limitations, for example, in reaching and grasping objects. Studies have shown that children with physical impairments can improve their independence, cognitive, and social skills by playing using robots. Haptic robots can provide the sense of touch, and implement haptic guidance to help users reach objects. In this study, we developed a telerobotic haptic system with two haptic robots. The goal of this study was to do preliminary tests of the haptic guidance method and the prediction of targets. Another goal was to analyze the visual attention of the participants during the activity when eye-hand discoordination was induced. Five adults without disabilities played a whack-a-mole game using the robotic system, to assure that the robot works adequately before children with disabilities use it. The robots were programmed to induce eye-hand discoordination, so that haptic guidance would be required. A multi-layer perceptron neural network was implemented to predict the target moles that the participants had to reach, which in future versions, will allow generation of forbidden region virtual fixtures (FRVF) to guide the user towards the target moles. Analysis of participant's eye gaze led to the hypothesis that the less control a person has over the teleoperation system, the less they will look at the target. On average, the accuracy of the target prediction by the neural network was 70.7%. The predicting of targets will allow the robot to assist children during movement of the robot towards the target toy, without needing the children to explicitly point out with their gaze which toy they want to reach. This will potentially lead to a more intuitive and faster human-robot interaction.

Keywords: Micro- and nano-robotics, Biomechanics, Manipulation
Abstract: In this paper, actively controllable stem cell spheroid-based microrobot is proposed to improve targeting the stem cells to specific desired sites using external electromagnetic actuation (EMA) system. Mouse mesenchymal stem cell (MSCs) were labeled with 0.2 mg/ml magnetic nanoparticles (MNPs) and no cytotoxicity was observed up to MNPs concentration of 0.2 mg/ml on 2D and 3D culture. The stem cell spheroids were fabricated with diameters of 353.62 μm ± 24.61 (MNP unlabeled) and 336.84 μm ± 28.2 (MNP labeled) by ultra-low attachment plate culture. In addition, mouse MSCs spheroids were actively controlled by the EMA system. Therefore, these results demonstrated that stem cell spheroid-based microrobot can be potential therapeutic agent for tissue regeneration or repairing, such as, articular cartilage regeneration.

Keywords: Locomotion and manipulation in robots and biological systems, Dynamics and control
Abstract: In recent years, because the development of space technology has been increasing for the purpose of improving the social infrastructure, the expansion of space transportation systems based on low-cost and high-frequency rockets is important. Solid propellants used in solid-fuel rockets have properties of the compactness, inexpensiveness, and easy-handling. However, solid propellants are highly viscous slurries and highly explosive. As there is no device capable of continuously and safely transporting solid propellant, the process of manufacturing solid propellant is a batch process. We focused on the movement of human intestines that knead and transport with a small force as part of the development process. In this report, we design a mechanism for the mixing process by using a peristaltic mixing transporting device for efficiency and by automating the equipment. Specifically, we conduct two experiments with samples using an adjusted fluid ratio: a comparison experiment on fluidity and a pressure response experiment. We investigate the possibility of automatic control and efficiency by using the factors of pressure and flow rate

Keywords: Biologically-inspired systems, Locomotion and manipulation in robots and biological systems, Neural networks
Abstract: To date, controlled flight of very small, insect-sized (∼100 mg) Micro Aerial Vehicles (MAVs) has required off-board sensors and computation. Achieving autonomy in more general environments that do not have such resources available will require integrating these components. In this work we present advances toward this goal by demonstrating a new, custom-built, low-weight (26 mg) camera mounted on a 74 mg flapping-wing robot. We implemented a convolution neural network (CNN) to classify images. Using this system, we demonstrated how the camera-equipped robot could repeatably move toward flower images and away from predator images. An estimate of the computational requirements of our network indicates that it could be performed using low-weight microcontrollers that are compatible with the payload and power constraints of insect-scale MAVs. Many desired capabilities for aerial vehicles, such as landing site selection and obstacle detection and avoidance, are ill-defined because the boundary between positive and negative classes are unclear. This work shows that CNNs operating on input from vision, which have previously been deployed only on larger robots, can be used at insect-scale for such tasks.

Keywords: Micro- and nano-robotics, Locomotion and manipulation in robots and biological systems, Biologically-inspired systems
Abstract: The mechanism of swimming at very low Reynolds number conditions is a topic of interest to biologists and engineering community. We develop a novel kinematic model of a slender flexible swimmer which locomotes in a low Reynolds number regime. In contrast to existing techniques that model such systems as a connected set of straight, rigid links, the novelty of our technique stems from the fact that we model the swimmer with two components - one is a straight, rigid body (the head) and the other is a flexible member (the tail). Using Cox theory we model the gradient of the forces as a function of the instantaneous shape of the swimmer and its velocity. By virtue of the low inertia conditions, an expression for the translational and rotational velocity of the head is obtained for the planar motion in the form of a Lie algebra of the Special Euclidean group. We explain the principal fiber bundle structure of the configuration space of the swimmer and use that to show a weak controllability result for a type of slender flexible swimmer where the shape space is the space of all continuous curves of a given length. A set of simulation results is presented showing the variation of the swimmer head velocity for a bump function moving along the swimmer length.

Keywords: Biologically-inspired systems, Dynamics and control, Locomotion and manipulation in robots and biological systems
Abstract: This paper presents the optimisation of explosive jumping motions on a 3-DoF leg prototype. The leg is based on the recently introduced asymmetric compliant actuator scheme, in which a series-elastic main drive is augmented with a parallel adjustable compliant branch with significantly different stiffness and energy storage capacity properties. The leg prototype implements two such actuation configurations, one of which includes a biarticulated branch, and they are compared to conventional series-elastic based actuation. An optimisation problem is formulated to optimise the joint trajectories and elastic element pretension to maximise jumping height. A simulation study demonstrates that the biarticulated configuration yields maximum jumping height, and that it achieves the highest peak joint power. Compared to series-elastic based actuation, the augmented leg jumps 4% higher with a monoarticulated parallel compliance configuration while using less energy, and over 10% higher in biarticulated configuration.

Keywords: Biologically-inspired systems, Locomotion and manipulation in robots and biological systems
Abstract: Infrastructure pipes require inspection to prevent accidents. However, it is difficult to inspect a 1-in-diameter gas and water pipe because it is long, narrow, and complicated. To inspect 1-in pipe, we have developed pipe inspection robots based on the movement of earthworms or inchworms. However, in either of these robots, one is not able to output high propulsive force, and one is not able to output high traction force. To run and inspect a robot in a complex long-distance pipe, the robot that has both propulsive force and traction force is necessary. In this paper, we propose an active traction type　flexible inchworm robot (PI-RO I). The proposed robot has a mechanism that outputs the propulsive force and the traction force actively as a method to make both forces high. To achieve this mechanism, a grip unit for holding the robot inside the pipe and an extension unit for moving the robot were developed, and basic characteristics experiments on these units were conducted. In addition, the propulsive force and the traction force of the robot were measured by running experiments, and it was confirmed that both can output actively.

Keywords: Biologically-inspired systems, Locomotion and manipulation in robots and biological systems
Abstract: Some aquatic insects can rapidly dash over the water surface by secreting chemical material that lowers the surface tension behind. This locomotion is commonly known as Marangoni propulsion, and we built a non-tethered miniature robot inspired by their mobility. The robot had six circular footpads with equilateral triangular cross section, and weighed 14.8 gram including on-board electronics, a battery, and a servo motor. Although the robot successfully skimmed over the water surface by dripping alcohol (e.g., 3-Methyl-1-butanol), the robot could not maintain a linear motion by itself. Therefore, we designed and attached flexural joints at the hind legs of the robot to compensate its linear motion; the asymmetric force applied to the hind legs subsequently induced another counter moment due to the bending of flexural joints. During the experiments,these joints were effective at reducing undesired lateral deviation more than 3-fold compared to one without flexural joints. Also, the characteristics of the robot’s locomotion was similar with the locomotion of aquatic arthropods according to the dimensionless number analysis.

Keywords: Soft robotics, Biologically-inspired systems, Design and control
Abstract: Due to safety concerns in human-robot interaction, researchers are moving from rigid-component robotics to soft robotics. This paper presents the design of a novel linear pneumatic robot structure based on the eversion principle. Experiments are carried out to investigate the forces that this robot can exert.

Keywords: Novel actuators, Wearable and augmenting devices, Exoskeletons
Abstract: Mechatronic rehabilitative devices have been proven to provide cost effective solutions to long term physical therapy for patients with musculoskeletal disorders. However, current actuator technologies limit the minimization of the overall size and weight of these devices preventing innovation into unobtrusive wearable form factors that are also effective and comfortable. This study is focused on a recently discovered smart actuator made from flexible nylon thread, which has exhibited a great potential for use in wearable mechatronic devices. This is known as the twisted coiled actuator (TCA) due to the hyper twisting and induced coiling involved in its fabrication process. One of the limiting factors of the TCA, is the thermal activation mechanism, which results in a slow cooling phase and a low working bandwidth. This paper is focused on optimizing an active cooling design using numerical analysis. To do this, a simple pipe geometry was designed and tested using fluid dynamics software. Three off-the-shelf fluidic pumps were simulated using varying tube diameters to find a sufficient cooling rate, a minimum fluid volume, and to select a proper pump for future testing. The results indicate that a global maximum cooling rate exists for each specific pump at a unique tube diameter. Additionally, the speed of cooling was under 500 ms concluding that the pumps tested can sufficiently provide the cooling rates required to assist motion in wearable devices. Furthermore, the process developed here provides quantitative support for the optimal selection of initial design parameters and can be translated to designs using different form factors and fluid properties.