Marcia K. O’Malley

BACKGROUND: Damage to the corticospinal tract after stroke and spinal cord injury (SCI) often results in persistent upper extremity (UE) impairment. Transcutaneous spinal stimulation (TSS) and robotic technologies have been explored as approaches to facilitate motor training; however, their combined effects on UE sensorimotor recovery remain poorly understood. The purpose of this study was to examine the effects of TSS combined with UE robotic training in individuals with chronic stroke or SCI. METHODS: Five participants with stroke and six with SCI completed a 14-week, sham controlled, single blind crossover study consisting of four total weeks of assessments (one week each pre and post for both training phases), four weeks of UE training with sham TSS, a two-week washout period, and four weeks of UE training with active TSS. Each one-hour session (three days/week) included robotic exoskeleton-assisted UE movements and hand grip training, performed concurrently with sham or active TSS. Assessments included electrophysiological measurements and standardized rehabilitation outcomes. RESULTS: Descriptive analysis revealed meaningful individual improvements masked by group-level heterogeneity. In the stroke group, three participants showed grip strength improvement (assessed without stimulation) after the active phase (+ 9.4 Newtons [N] to + 23.9 N), with two-to-four-fold increases in forearm muscle activation. Mean Fugl-Meyer overall UE scores improved from 89 to 94.2. In the SCI group, two participants showed grip strength gains. One participant exhibited a six-fold immediate force increase (1.0 N to 6.2 N) during stimulation. Another participant achieved improved grip strength without stimulation (23.9 N to 36.8 N) and a three-fold increase in electromyography (EMG) activity from the flexor carpi radialis and first dorsal interosseous muscles, alongside partial pin-prick sensory recovery and self-reported restoration of previously affected perspiration during the active TSS phase. CONCLUSIONS: Varied outcomes in participants confirm that therapeutic effects of combined TSS and robotic UE training are highly individualized. Three critical elements must be blended for the best outcomes of this combinatorial approach: residual UE function, a curated stimulation paradigm, and tailored UE training that provides appropriate challenge, intensity, and salience. The results suggest TSS with UE robotic training hold key potential when considered in the context of the physiological and functional profile of each participant.

Vision and sound often dominate human-robot interaction systems, but touch can convey nuanced, real-time task and environment information. Vibrotactile feedback is increasingly common in commercial devices, and the wrist offers a promising location for wearable haptics: it is unobtrusive and sensitive to tactile input. While prior work has studied vibrotactile feedback in navigation and action confirmation tasks, quantifying information transfer for wrist-worn vibrotactile cues remains underexplored. This study estimates information transfer on the wrist (bits per stimulus) using a confusion-matrix-based metric alongside accuracy and pleasantness ratings. We conducted three linked studies with the same participants: identifying single-tactor cues spanning frequency, amplitude, and modulation; identifying the same cues under sequential vibrotactile masking and during a concurrent typing task; and identifying multi-tactor spatiotemporal patterns and rating pleasantness. Participants reliably discriminated complex signals, with amplitude and feature interactions playing key roles. Information transfer ranged from 0.75 bits/stimulus for multi-tactor patterns to 2.28 bits/stimulus for single-tactor testing across all participants (2.56 for experienced participants); masking and typing yielded 1.64 and 1.85 bits/stimulus, respectively. Performance was not solely driven by amplitude-normalized intensity.

Since the 1980s, the power law has been the dominant view of the trajectory of skill acquisition. More recent research has challenged this "law," suggesting other models may better capture individual-level data. Furthermore, the motor learning and recovery literature suggests dynamical models might better capture nonmonotonic behavior and the effect of feedback. This study compares the fits of six models on data from two mirror-tracing experiments with different feedback metrics delivered through haptics. This includes two power models, the exponential model, a hybrid power and exponential model, and two more recent dynamical models that allow for nonmonotonic learning curves and can incorporate the role of feedback. Like others before, these results show that the "power law" model is not necessarily the best way to describe individual learning. However, none of the models examined showed a clear advantage in fitting individual-level data, and we discuss multiple reasons why this might be the case. (PsycInfo Database Record (c) 2026 APA, all rights reserved).

Training for complex motor tasks, such as those encountered in minimally invasive surgery, benefits from effective performance feedback mechanisms to accelerate skill acquisition and ensure retention. Prior work has demonstrated that haptic feedback based on movement smoothness quantified by the metric spectral arc length (SPARC), when provided in real-time as trainees perform complex motor tasks, can cause beneficial changes in task completion strategies resulting in faster completion times without loss of accuracy. The concept of movement smoothness is abstract, however, and more intuitive measures of movement smoothness like idle time and average velocity can be good alternatives to SPARC. Here, we demonstrate the effect of real-time objective performance feedback of movement smoothness, conveyed through a vibrotactile cue encoding alternative measures of movement smoothness, compared to feedback based on SPARC. Subjects receiving smoothness-based feedback based on average velocity performed the task fastest, but their accuracy was lower than the other two groups. We evaluated the effect of removing feedback for additional trials, and showed that performance improvements ceased. After training, the three groups were indistinguishable from each other.

Background Stroke is a leading cause of long-term adult disability, with approximately 80% of survivors experiencing upper extremity (UE) motor impairments. Conventional tools like the Fugl-Meyer Assessment (FMA) are widely used but limited by ordinal scales and subjective visual observation. While wearable robotics offer high-resolution data, their clinical translation is hindered by a lack of standardized protocols and limited interpretability for clinical decision-making. Objective This study aimed to develop an objective, standardized, and clinically interpretable method to quantify UE motor qualities by integrating wearable robotic technology with traditional clinical assessment tasks. Methods Ten healthy individuals and ten stroke survivors performed seven standardized tasks (six from the FMA-UE and one additional elbow task) while wearing the HARMONY exoskeleton. We developed a “trajectory pattern similarity score” based on the root mean square error between individual joint trajectories and normative averages. Additionally, kinematic synergy analysis was performed using non-negative matrix factorization to evaluate alterations in multi-joint coordination. Results The trajectory pattern similarity score showed a strong negative correlation with clinical FMA-UE scores (r = −0.93, p < 0.01) and demonstrated excellent test-retest reliability (ICC = 0.98). The number of identified kinematic synergies decreased significantly as motor impairment severity increased (r = 0.79, p < 0.01). Furthermore, kinematic synergy analysis provided a mechanistic explanation for reduced individual joint control. Post-stroke synergies could be explained through the merging (linear combinations of healthy kinematic patterns), preservation, or loss of healthy kinematic synergies, reflected as pathological joint coupling and loss of specific individual joint control. Conclusions This study presents a novel, standardized assessment framework that integrates wearable robotic technology with conventional clinical tasks. By bridging the gap between objective robotic data and clinical interpretability, this approach would enable robust motor impairment assessment and intuitive phenotyping of motor characteristics to guide personalized rehabilitation strategies.

Hybrid FES-exoskeleton systems can drive users through trajectories that mimic functional upper limb movements with higher accuracy and reduced motor torque compared to either system alone, situating them well as assistive devices. Limitations in FES-driven movement accuracy, exoskeleton size, and electrical energy requirements prevent the widespread adoption of hybrid systems as assistive devices in real-world settings. Additionally, these systems are limited in their adoption due to difficulties in efficiently coordinating the two subsystems, and given the heterogeneity of motor capabilities across neurologically-impaired individuals. We present a methodology for nonlinear trajectory optimization to create feasible, personalized trajectories between desired set-points that preemptively considers each user's dynamic constraints. We evaluate personalized and minimum jerk trajectories experimentally with eight neurologically intact participants and in a case study with one individual with a cervical spinal cord injury, extending our prior simulation-based validation. The personalized trajectories maintain significant exoskeleton torque reduction from the exoskeleton-alone system to the hybrid FES-exoskeleton system without sacrificing functional tracking accuracy.

Individuals who lack tactile and/or proprioceptive sensations in their lower limbs commonly report postural and locomotive imbalance. To mitigate imbalance, haptic feedback has been implemented using devices that employ external sensors and waist-worn actuators for sensory augmentation. Analyses of the effectiveness of these devices have primarily focused on balance outcomes and generally disregard the potential effect of the user's varying activities or anthropometric data on their perception of the haptic feedback. Since motor activity can influence haptic perception, we investigate vibrotactile cue localizability in a waist-worn haptic device for two conditions, standing and walking on a treadmill at a self-selected speed. In this preliminary study, ten participants without sensorimotor deficits wore a waistband equipped with seven vibrotactile actuators. Vibrotactile cues were played at randomized locations during standing and walking, and participants reported the perceived stimulation location. In addition, we recorded relevant anthropometric data for each participant. In both standing and walking conditions, participants correctly localized 51% of the vibrotactile cues on average. When considering zonal accuracy, or localization within one vibrotactor position, participants, on average, achieved an accuracy of 91 % in standing and 90% in walking.

Wearable haptic devices can provide haptic feedback in both active and passive touch interactions. In virtual or extended reality environments, wearable devices also enable referred haptic feedback, where touch sensation expected in one place on the body is directed to another. How people perceive the distinction between these feedback forms has been relatively under-explored, especially considering the additional role that vision plays in our multisensory understanding of the world. To explore how active and passive touch affect the perception of referred haptic feedback, we conducted an experiment in VR where participants chose the stiffer of two springs in a 2-interval, 2-alternative forced-choice design. We find that participants can be categorized into two groups based on the different strategies they employ for making the decision - those with a haptic or a visual prior, similar to related work. We also considered both active and passive feedback conditions. Notably, people are more accurate in judging haptic stiffness during the active case. Our results have implications for designers of virtual systems and simulations where users receive various sensory inputs, both via active and passive interactions, with potential mismatches due to latency, bandwidth, or design issues.

Movement smoothness is a commonly employed metric of motor impairment in stroke patients due to a strong correlation with Fugl-Meyer Assessment score, and because it can be assessed with functional-oriented tasks like wrist pointing. Robotic assessments of wrist-pointing tasks typically use targets located at fixed points in the workspace, which can exceed the range of motion (ROM) of stroke patients who exhibit severe motor impairment. We hypothesize that movement smoothness assessed based on movements that do not reach visual targets may conflate assessment of movement smoothness with ROM. We propose that scaling the locations of wrist pointing targets for wrist flexion-extension and radial-ulnar movements from a ROM assessment will better reflect movement characteristics. In this paper, we analyze wrist-pointing movements for fixed and ROM-scaled target locations. We evaluate wrist-pointing performance in both conditions for 8 neurologically-intact participants whose wrist ROM was constrained with a wrist brace simulating ROM deficits present after stroke. Results show that failing to reach targets during wrist pointing has a significant negative impact on movement smoothness. Additionally, scaling target placement to ROM increases the proportion of targets reached while producing practically equivalent movement smoothness to successfully reached fixed target locations. These findings support incorporation of ROM-scaled target placement into movement smoothness assessment.

Automated tracking of surgical tool keypoints in robotic surgery videos is an essential task for various downstream use cases such as skill assessment, expertise assessment, and the delineation of safety zones. In recent years, the explosion of deep learning for vision applications has led to many works in surgical instrument segmentation, while lesser focus has been on tracking specific tool keypoints, such as tool tips. In this work, we propose a novel, multi-frame context-driven deep learning framework to localize and track tool keypoints in surgical videos. We train and test our models on the annotated frames from the 2015 EndoVis Challenge dataset, resulting in state-of-the-art performance. By leveraging sophisticated deep learning models and multi-frame context, we achieve 90% keypoint detection accuracy and a localization RMS error of 5.27 pixels. Results on a self-annotated JIGSAWS dataset with more challenging scenarios also show that the proposed multi-frame models can accurately track tool-tip and tool-base keypoints, with <4.2-pixel RMS error overall. Such a framework paves the way for accurately tracking surgical instrument keypoints, enabling further downstream use cases. Project and dataset webpage: https://tinyurl.com/mfc-tracker

Haptic devices enable communication via touch, augmenting visual and auditory displays, or by offering alternative channels of communication when vision and hearing are unavailable. Because of the different types of haptic stimuli that are perceivable by users — vibration, skin stretch, pressure and temperature, among others — devices can be designed to communicate complex information by delivering multiple types of haptic stimuli simultaneously. These multi-sensory haptic devices are often designed to be wearable and have been developed for use in a wide variety of applications, including communication, entertainment and rehabilitation. Multi-sensory haptic devices present unique challenges to designers because human perceptual acuity can vary widely depending on the wearable location on the body and/or the heterogeneity in human perceptual performance, particularly when multiple cues are presented simultaneously. Additionally, packaging haptic systems in a wearable form factor presents its own engineering challenges such as cue masking, device mounting and actuator capabilities, among others. Thus, in this Review, we discuss the state-of-the-art and specific obstacles present in the field to produce multi-sensory devices that enhance the human capacity for haptic interaction and information transmission. Haptic devices enable communication via touch, augmenting visual and auditory displays, or by offering alternative channels of communication when vision and hearing are unavailable. This Review discusses multi-sensory wearable haptics, focusing on body-worn devices that convey multiple types of cutaneous haptic feedback. The translation of wearable multi-sensory haptic devices relies on a robust understanding of human haptic perception so that feedback modalities can be combined to optimally enhance user performance in a given application domain. The contact mechanics at the haptic interface between device and skin varies between users with an unknown effect on haptic perception. In addition to traditional electromechanical actuation, new methods, such as polymeric, fluidic and thermal actuation, now exist. When designing wearable haptic devices, body location, the device interface to the body, user comfort and preserving the integrity of the haptic feedback must be considered. The translation of wearable multi-sensory haptic devices relies on a robust understanding of human haptic perception so that feedback modalities can be combined to optimally enhance user performance in a given application domain. The contact mechanics at the haptic interface between device and skin varies between users with an unknown effect on haptic perception. In addition to traditional electromechanical actuation, new methods, such as polymeric, fluidic and thermal actuation, now exist. When designing wearable haptic devices, body location, the device interface to the body, user comfort and preserving the integrity of the haptic feedback must be considered.

Automated tracking of surgical tool keypoints in robotic surgery videos is an essential task for various downstream use cases such as skill assessment, expertise assessment, and the delineation of safety zones. In recent years, the explosion of deep learning for vision applications has led to many works in surgical instrument segmentation, while lesser focus has been on tracking specific tool keypoints, such as tool tips. In this work, we propose a novel, multi-frame context-driven deep learning framework to localize and track tool keypoints in surgical videos. We train and test our models on the annotated frames from the 2015 EndoVis Challenge dataset, resulting in state-of-the-art performance. By leveraging sophisticated deep learning models and multi-frame context, we achieve 90\% keypoint detection accuracy and a localization RMS error of 5.27 pixels. Results on a self-annotated JIGSAWS dataset with more challenging scenarios also show that the proposed multi-frame models can accurately track tool-tip and tool-base keypoints, with ${&lt;}4.2$-pixel RMS error overall. Such a framework paves the way for accurately tracking surgical instrument keypoints, enabling further downstream use cases. Project and dataset webpage: https://tinyurl.com/mfc-tracker

Robot-assisted minimally invasive surgery (RAMIS) offers numerous benefits over traditional open surgery, resulting in greater prevalence of use and range of approved procedures. The proliferation of RAMIS has highlighted a need for effective, robust, and objective methods for assessing robotic surgical skills. Traditionally, assessment has relied on expert observation using structured grading rubrics. Although validated and widely used, this method is also resource intensive and subject to reviewer bias. In response, recent work has explored the potential for more robust assessment methods, including the development of skill-based metrics, crowd-sourced assessment techniques, and automated evaluation systems. This review summarizes recent developments in robotic surgical technical skill assessment, focusing on studies using the da Vinci platform. Assessment methods are grouped into four categories: structured rubrics, skill-based metrics, crowd-sourcing techniques, and automated assessment models. Trends of note include adaptation of established rubrics for specific areas of specialty, the implementation of deep learning models for automated assessment, and a move to integrate crowd-sourcing platforms for efficient and inexpensive evaluation. While traditional grading rubric structures remain the standard, multilevel assessment strategies and objective feedback systems are gaining traction. Future work should seek to integrate task- and movement-based assessment into procedure-level evaluations to create more robust and generalizable models for assessment. These advances show a shift towards data-driven and objective assessment methods, which could improve surgical training and patient outcomes.

Holistic consideration of the human and the robot is necessary to overcome hurdles in human-robot interaction.

Wearable haptic devices provide touch feedback to users for applications including virtual reality, prosthetics, and navigation. When these devices are designed for experimental validation in research settings, they are often highly specialized and customized to the specific application being studied. As such, it can be difficult to replicate device hardware due to the associated high costs of customized components and the complexity of their design and construction. In this work, we present Snaptics, a simple and modular platform designed for rapid prototyping of fully wearable multi-sensory haptic devices using 3D-printed modules and inexpensive off-the-shelf components accessible to the average hobbyist. We demonstrate the versatility of the modular system and the salience of haptic cues produced by wearables constructed with Snaptics modules in two human subject experiments. First, we report on the identification accuracy of multi-sensory haptic cues delivered by a Snaptics device. Second, we compare the effectiveness of the Snaptics Vibrotactile Bracelet to the Syntacts Bracelet, a high-fidelity wearable vibration feedback bracelet, in assisting participants with a virtual reality sorting task. Results indicate that participant performance was comparable in perceiving cue sets and in completing tasks when interacting with low-cost Snaptics devices as compared to a similar research-grade haptic wearables.

Multiscale Haptic Interactions Most wearable haptic devices (top left) deliver haptic notifications to the body; perception of haptics is depicted as brain activity (pink region). In the multiscale paradigm (bottom right), a user actively interacts with the wearable device to perceive a greater amount of information, both via passive receipt of haptic notifications through the band (pink region) and via active exploration of the textile haptic band with fingertips (yellow region). Background letters and numbers represent information transmission to the user via the device. For more information on multiscale haptics, see article number 2300897 by Marcia K. O’Malley and co-workers.

Wearable haptic devices transmit information via touch receptors in the skin, yet devices located on parts of the body with high densities of receptors, such as fingertips and hands, impede interactions. Other locations that are well‐suited for wearables, such as the wrists and arms, suffer from lower perceptual sensitivity. The emergence of textile‐based wearable devices introduces new techniques of fabrication that can be leveraged to address these constraints and enable new modes of haptic interactions. This article formalizes the concept of “multiscale” interaction, an untapped paradigm for haptic wearables, enabling enhanced delivery of information via textile‐based haptic modules. In this approach, users choose the depth and detail of their haptic experiences by varying their interaction mode. Flexible prototyping methods enable multiscale haptic bands that provide both body‐scale interactions (on the forearm) and hand‐scale interactions (on the fingers and palm). A series of experiments assess participants’ ability to identify pressure states and spatial locations delivered by these bands across these interaction scales. A final experiment demonstrates the encoding of three‐bit information into prototypical multiscale interactions, showcasing the paradigm's efficacy. This research lays the groundwork for versatile haptic communication and wearable design, offering users the ability to select interaction modes for receiving information.

Wearable haptic devices on the forearm can relay information from virtual agents, robots, and other humans while leaving the hands free. We introduce and test a new wearable haptic device that uses soft actuators to provide normal and shear force to the skin of the forearm. A rigid housing and gear motor are used to control the direction of the shear force. A 6-axis force/torque sensor, distance sensor, and pressure sensors are integrated to quantify how the soft tactor interacts with the skin. When worn by participants, the device delivered consistent shear forces of up to 0.64 N and normal forces of up to 0.56 N over distances as large as 14.3 mm. To understand cue saliency, we conducted a user study asking participants to identify linear shear directional cues in a 4-direction task and an 8-direction task with different cue speeds, travel distances, and contact patterns. Participants identified cues with longer travel distances best, with an 85.1% accuracy in the 4-direction task, and a 43.5% accuracy in the 8-direction task. Participants had a directional bias, with a preferential response in the axis towards and away from the wrist bone.

Individuals who have suffered a spinal cord injury often require assistance to complete daily activities, and for individuals with tetraplegia, recovery of upper-limb function is among their top priorities. Hybrid functional electrical stimulation (FES) and exoskeleton systems have emerged as a potential solution to provide upper limb movement assistance. These systems leverage the user's own muscles via FES and provide additional movement support via an assistive exoskeleton. To date, these systems have focused on single joint movements, limiting their utility for the complex movements necessary for independence. In this paper, we extend our prior work on model predictive control (MPC) of hybrid FES-exo systems and present a multi degree of freedom (DOF) hybrid controller that uses the controller's cost function to achieve desired behavior. In studies with neurologically intact individuals, the hybrid controller is compared to an exoskeleton acting alone for movement assistance scenarios incorporating multiple degrees-of-freedom of the limb to explore the potential for exoskeleton power consumption reduction and impacts on tracking accuracy. Additionally, each scenario is explored in simulation using the models required to generate the MPC formulation. The two DOF hybrid controller implementation saw reductions in power consumption and satisfactory trajectory tracking in both the physical and simulated systems. In the four DOF implementation, the experimental results showed minor improvements for some joints of the upper limb. In simulation, we observed comparable performance as in the two DOF implementation.

Transcutaneous spinal stimulation (TSS) is a promising rehabilitative intervention to restore motor function and coordination for individuals with spinal cord injury (SCI). The effects of TSS are most commonly assessed by evaluating muscle response to stimulation using surface electromyography (sEMG). Given the increasing use of robotic devices to deliver therapy and the emerging potential of hybrid rehabilitation interventions that combine neuromodulation with robotic devices, there is an opportunity to leverage the on-board sensors of the robots to measure kinematic and torque changes of joints in the presence of stimulation. This paper explores the potential for robotic assessment of the effects of TSS delivered to the cervical spinal cord. We used a four degree-of-freedom exoskeleton to measure the torque response of upper limb (UL) joints during stimulation, while simultaneously recording sEMG. We analyzed joint torque and electromyography data generated during TSS delivered over individual sites of the cervical spinal cord in neurologically intact participants. We show that site-specific effects of TSS are manifested not only by modulation of the amplitude of spinally evoked motor potentials in UL muscles, but also by changes in torque generated by individual UL joints. We observed preferential resultant action of proximal muscles and joints with stimulation at the rostral site, and of proximal joints with rostral-lateral stimulation. Robotic assessment can be used to measure the effects of TSS, and could be integrated into complex control algorithms that govern the behavior of hybrid neuromodulation-robotic systems.

Wearable devices increasingly incorporate vibrotactile feedback notifications to users, which are limited by the frequency-dependent response characteristics of the low-cost actuators that they employ. To increase the range and type of information that can be conveyed to users via vibration feedback, it is crucial to understand user perception of vibration cue intensity across the narrow range of frequencies that these actuators operate. In this paper, we quantify user perception of vibration cues conveyed via a linear resonant actuator embedded in a bracelet interface using two psychophysical experiments. We also experimentally determine the frequency response characteristics of the wearable device. We then compare user perceived intensity of vibration cues delivered by the bracelet when the cues undergo frequency-specific amplitude modulation based on user perception compared to modulation based on the experimental or manufacturer-reported characterization of the actuator dynamic response. For applications in which designers rely on user perception of cue amplitudes across frequencies to be equivalent, it is recommended that a perceptual calibration experiment be conducted to determine appropriate modulation factors. For applications in which only relative perceived amplitudes are important, basing amplitude modulation factors on manufacturer data or experimentally determined dynamic response characteristics of the wearable device should be sufficient.

Realistically improving upper-limb prostheses is only possible if we listen to users' actual technological needs.

Haptic feedback offers a useful mode of communication in visually or auditorily noisy environments. The adoption of haptic devices in our everyday lives, however, remains limited, motivating research on haptic wearables constructed from materials that enable comfortable and lightweight form factors. Textiles, a material class fitting these needs and already ubiquitous in clothing, have begun to be used in haptics, but reliance on arrays of electromechanical controllers detracts from the benefits that textiles offer. Here, we mitigate the requirement for bulky hardware by developing a class of wearable haptic textiles capable of delivering high-resolution information on the basis of embedded fluidic programming. The designs of these haptic textiles enable tailorable amplitudinal, spatial, and temporal control. Combining these capabilities, we demonstrate wearables that deliver spatiotemporal cues in four directions with an average user accuracy of 87%. Subsequent demonstrations of washability, repairability, and utility for navigational tasks exemplify the capabilities of our approach.

A goal of wearable haptic devices has been to enable haptic communication, where individuals learn to map information typically processed visually or aurally to haptic cues via a process of cross-modal associative learning. Neural correlates have been used to evaluate haptic perception and may provide a more objective approach to assess association performance than more commonly used behavioral measures of performance. In this article, we examine Representational Similarity Analysis (RSA) of electroencephalography (EEG) as a framework to evaluate how the neural representation of multifeatured haptic cues changes with association training. We focus on the first phase of cross-modal associative learning, perception of multimodal cues. A participant learned to map phonemes to multimodal haptic cues, and EEG data were acquired before and after training to create neural representational spaces that were compared to theoretical models. Our perceptual model showed better correlations to the neural representational space before training, while the feature-based model showed better correlations with the post-training data. These results suggest that training may lead to a sharpening of the sensory response to haptic cues. Our results show promise that an EEG-RSA approach can capture a shift in the representational space of cues, as a means to track haptic learning.

Haptic cues delivered via wearable devices have great potential to enhance a user's experience by transmitting task information and touch sensations in domains such as virtual reality, teleoperation, and prosthetics. Much is still unknown on how haptic perception, and consequently optimal haptic cue design, varies between individuals. In this work we present three contributions. First, we propose a new metric, the Allowable Stimulus Range (ASR), as a way to capture subject-specific magnitudes for a given cue, using the method of adjustments and the staircase method. Second, we present a modular, grounded, 2-DOF, haptic testbed designed to conduct psychophysical experiments in multiple control schemes and with rapidly-interchangeable haptic interfaces. Third, we demonstrate an application of the testbed and our ASR metric, together with just noticeable differences (JND) measurements, to compare perception of haptic cues delivered via position or force control schemes. Our findings show that users demonstrate higher perceptual resolution in the position-control case, though survey results suggest that force-controlled haptic cues are more comfortable. The results of this work outline a framework to define perceptible and comfortable cue magnitudes for an individual, providing the groundwork to understand haptic variability, and compare the effectiveness of different types of haptic cues.

Vibration is a widely used mode of haptic communication, as vibrotactile cues provide salient haptic notifications to users and are easily integrated into wearable or handheld devices. Fluidic textile-based devices offer an appealing platform for the incorporation of vibrotactile haptic feedback, as they can be integrated into clothing and other conforming and compliant wearables. Fluidically driven vibrotactile feedback has primarily relied on valves to regulate actuating frequencies in wearable devices. The mechanical bandwidth of such valves limits the range of frequencies that can be achieved, particularly in attempting to reach the higher frequencies realized with electromechanical vibration actuators ( 100 Hz). In this paper, we introduce a soft vibrotactile wearable device constructed entirely of textiles and capable of rendering vibration frequencies between 183 and 233 Hz with amplitudes ranging from 23 to 114 g. We describe our methods of design and fabrication and the mechanism of vibration, which is realized by controlling inlet pressure and harnessing a mechanofluidic instability. Our design allows for controllable vibrotactile feedback that is comparable in frequency and greater in amplitude relative to state-of-the-art electromechanical actuators while offering the compliance and conformity of fully soft wearable devices.

Introduction: Individuals who have suffered a cervical spinal cord injury prioritize the recovery of upper limb function for completing activities of daily living. Hybrid FES-exoskeleton systems have the potential to assist this population by providing a portable, powered, and wearable device; however, realization of this combination of technologies has been challenging. In particular, it has been difficult to show generalizability across motions, and to define optimal distribution of actuation, given the complex nature of the combined dynamic system. Methods: In this paper, we present a hybrid controller using a model predictive control (MPC) formulation that combines the actuation of both an exoskeleton and an FES system. The MPC cost function is designed to distribute actuation on a single degree of freedom to favor FES control effort, reducing exoskeleton power consumption, while ensuring smooth movements along different trajectories. Our controller was tested with nine able-bodied participants using FES surface stimulation paired with an upper limb powered exoskeleton. The hybrid controller was compared to an exoskeleton alone controller, and we measured trajectory error and torque while moving the participant through two elbow flexion/extension trajectories, and separately through two wrist flexion/extension trajectories. Results: The MPC-based hybrid controller showed a reduction in sum of squared torques by an average of 48.7 and 57.9% on the elbow flexion/extension and wrist flexion/extension joints respectively, with only small differences in tracking accuracy compared to the exoskeleton alone. Discussion: To realize practical implementation of hybrid FES-exoskeleton systems, the control strategy requires translation to multi-DOF movements, achieving more consistent improvement across participants, and balancing control to more fully leverage the muscles' capabilities.

Vibration is ubiquitous as a mode of haptic communication, and is used widely in handheld devices to convey events and notifications. The miniaturization of electromechanical actuators that are used to generate these vibrations has enabled designers to embed such actuators in wearable devices, conveying vibration at the wrist and other locations on the body. However, the rigid housings of these actuators mean that such wearables cannot be fully soft and compliant at the interface with the user. Fluidic textile-based wearables offer an alternative mechanism for haptic feedback in a fabric-like form factor. To our knowledge, fluidically driven vibrotactile feedback has not been demonstrated in a wearable device without the use of valves, which can only enable low-frequency vibration cues and detract from wearability due to their rigid structure. We introduce a soft vibrotactile wearable, made of textile and elastomer, capable of rendering high-frequency vibration. We describe our design and fabrication methods and the mechanism of vibration, which is realized by controlling inlet pressure and harnessing a mechanical hysteresis. We demonstrate that the frequency and amplitude of vibration produced by our device can be varied based on changes in the input pressure, with 0.3 to 1.4 bar producing vibrations that range between 160 and 260 Hz at 13 to 38 g, the acceleration due to gravity. Our design allows for controllable vibrotactile feedback that is comparable in frequency and outperforms in amplitude relative to electromechanical actuators, yet has the compliance and conformity of fully soft wearable devices.

Recent advances in extended reality (XR) technologies make seeing and hearing virtual objects commonplace, yet strategies for synthesizing haptic interactions with virtual objects continue to be limited. Two design principles govern the rendering of believable and intuitive haptic feedback: movement through open space must feel “free” while contact with virtual objects must feel stiff. Herein, a novel multisensory approach that conveys proprioception and effort through illusory visual feedback and refers to the wrist, via a bracelet interface, discrete and continuous interaction forces that would otherwise occur at the hands and fingertips, is presented. Results demonstrate that users reliably discriminate the stiffness of virtual buttons when provided with multisensory pseudohaptic feedback, comprising tactile pseudohaptic feedback (discrete vibrotactile feedback and continuous squeeze cues in a bracelet interface) and visual pseudohaptic illusions of touch interactions. Compared to the use of tactile or visual pseudohaptic feedback alone, multisensory pseudohaptic feedback expands the range of physical stiffnesses that are intuitively associated with the rendered virtual interactions and reduces individual differences in physical‐to‐virtual stiffness mappings. This multisensory approach, which leaves users' hands unencumbered, provides a flexible framework for synthesizing a wide array of touch‐enabled interactions in XR, with great potential for enhancing user experiences.

Background: Despite the positive results in upper limb (UL) motor recovery after using electrical neuromodulation in individuals after cervical spinal cord injury (SCI) or stroke, there has been limited exploration of potential benefits of combining task-specific hand grip training with transcutaneous electrical spinal stimulation (TSS) for individuals with UL paralysis. Objectives: This study investigates the combinatorial effects of task-specific hand grip training and noninvasive TSS to enhance hand motor output after paralysis. Methods: Four participants with cervical SCI classified as AIS A and B and two participants with cerebral stroke were recruited in this study. The effects of cervical TSS without grip training and during training with sham stimulation were contrasted with hand grip training with TSS. TSS was applied at midline over cervical spinal cord. During hand grip training, 5 to 10 seconds of voluntary contraction were repeated at a submaximum strength for approximately 10 minutes, three days per week for 4 weeks. Signals from hand grip dynamometer along with the electromyography (EMG) activity from UL muscles were recorded and displayed as visual feedback. Results: Our case study series demonstrated that combined task-specific hand grip training and cervical TSS targeting the motor pools of distal muscles in the UL resulted in significant improvements in maximum hand grip strength. However, TSS alone or hand grip training alone showed limited effectiveness in improving grip strength. Conclusion: Task-specific hand grip training combined with TSS can result in restoration of hand motor function in paralyzed upper limbs in individuals with cervical SCI and stroke.