Aaron J. Young

Individuals post-stroke with hemiparetic gait often experience impaired mobility, which can be restored using powered exoskeletons. Specifically, hip exoskeletons can deliver positive work to the leg to help initiate leg swing; however, it is unclear which limb assistance strategy (e.g., unilateral or bilateral assistance) best improves walking performance due to the asymmetric characteristics of hemiparetic gait. Therefore, the purpose of this study was to investigate how different hip exoskeleton limb assistance strategies affect lower-limb joint biomechanics during walking for individuals post-stroke. We hypothesized that bilateral assistance would best improve late stance paretic leg orientation and swing initiation, unilateral paretic limb assistance would improve interlimb asymmetries and unilateral nonparetic limb assistance would indirectly improve late stance kinematics. We found bilateral assistance significantly increased paretic hip flexion angle, providing an improvement in leg swing initiation, and improved ankle joint work symmetry, suggesting an improvement in propulsion mechanics. Unilateral paretic limb assistance significantly increased paretic knee flexion during swing, which is a common rehabilitative target to improve stiff-knee gait for individuals post-stroke, and significantly improved joint work symmetry between limbs, which is beneficial for individuals who have increased reliance on the nonparetic limb. Unilateral nonparetic assistance did not improve late stance kinematics. These results provide insight into joint-level responses of individuals post-stroke and can help tailor hip exoskeleton assistance for individuals post-stroke based on rehabilitation targets.

OBJECTIVE: The study seeks to determine whether a powered, cable-driven exosuit has the potential to lower the lumbar muscle activity and overall metabolic expenditure of symmetric and asymmetric lifting tasks. METHODS: A lightweight, cable-driven back exosuit, using a three-state impedance controller, was developed to provide variable assistance based on user posture. Experimental electromyography (EMG), metabolic cost, and user preference data were recorded for ten participants evaluated wearing the powered back exosuit versus the backX, a commercially available passive back support exoskeleton, and a no exo baseline. RESULTS: Both exoskeletons significantly reduced (p$< $0.05) muscle activation of certain lumbar flexor and extensor muscles when compared to a no exo condition across all conditions tested, though neither significantly reduced the metabolic cost associated with lifting. Users tended to prefer lifting with the powered device as opposed to the passive or no exo condition. CONCLUSION: Despite the increased mass of powered back support exoskeletons, these devices can reduce lumbar muscle activity to a similar degree as passive exoskeletons, and are favored by users over their passive counterparts. SIGNIFICANCE: While current powered back support devices tend to incur the cost of being heavy, rigid, and inconvenient for certain lifting postures, these results show that cable-driven powered devices may minimize these factors to the point that they are favored over the currently popular passive devices on the market.

Data-driven methods have transformed our ability to assess and respond to human movement with wearable robots, promising real-world rehabilitation and augmentation benefits. However, the proliferation of data-driven methods, with the associated demand for increased personalization and performance, requires vast quantities of high-quality, device-specific data. Procuring these data is often intractable because of resource and personnel costs. We propose a framework that overcomes data scarcity by leveraging simulated sensors from biomechanical models to form a stepping-stone domain through which easily accessible data can be translated into data-limited domains. We developed and optimized a deep domain adaptation network that replaces costly, device-specific, labeled data with open-source datasets and unlabeled exoskeleton data. Using our network, we trained a hip and knee joint moment estimator with performance comparable to a best-case model trained with a complete, device-specific dataset [incurring only an 11 to 20%, 0.019 to 0.028 newton-meters per kilogram (Nm/kg) increase in error for a semisupervised model and 20 to 44%, 0.033 to 0.062 Nm/kg for an unsupervised model]. Our network significantly outperformed counterpart networks without domain adaptation (which incurred errors of 36 to 45% semisupervised and 50 to 60% unsupervised). Deploying our models in the real-time control loop of a hip/knee exoskeleton ( N = 8) demonstrated estimator performance similar to offline results while augmenting user performance based on those estimated moments (9.5 to 14.6% metabolic cost reductions compared with no exoskeleton). Our framework enables researchers to train real-time deployable deep learning, task-agnostic models with limited or no access to labeled, device-specific data.

ABSTRACT Motor module analysis is an important tool in the study of movement, particularly in people with impaired neural control. The most common method for computing motor modules is non-negative matrix factorization (NMF), which identifies a matrix of motor modules and their corresponding time-series activity from electromyography data. NMF has several limitations, including dependence of the muscle weightings on the number of modules selected. Approaches for selecting the number of modules vary between studies, making it difficult to compare and reproduce results. Some metrics of motor control complexity use the variance accounted for when extracting a single motor module (VAF 1 ), yet that module’s structure offers little biomechanical interpretability. In this work, we present a method for computing motor modules using an autoencoder, a neural network architecture that can find latent representations of data. Using a single layer autoencoder, we extracted motor modules from data in able-bodied and individuals post-stroke. The structure of autoencoder-computed modules were significantly less sensitive to selected module number. With the autoencoder-computed modules, increasing the number of modules added new information, instead of splitting previous modules. Autoencoder-computed modules, especially at low module counts, had more distinct and interpretable biomechanical functions. Lastly, the autoencoder-computed modules are consistent with previous NMF studies in persons with stroke, which found fewer modules needed to explain the muscle activity of paretic limbs. Our autoencoder-based method offers a new approach for computing motor modules, with advantages of better stability in module structure across module counts, and a more biomechanically relevant interpretation of VAF 1 . NEW &amp; NOTEWORTHY This work presents an approach for computing motor modules using an autoencoder and comprehensively compares the in stability of motor module structure, functional significance at low module counts, and interpretation of VAF 1 to the current state of the art. The AE-computed module structures were more stable at different module counts. The AE has the potential to improve confidence in module structure and make analysis less dependent on the specific number of modules selected.

Epically Powerful is an open-source robotics infrastructure that streamlines the underlying framework of wearable robotic systems - managing communication protocols, clocking, actuator commands, visualization, sensor data acquisition, data logging, and more - while also providing comprehensive guides for hardware selection, system assembly, and controller implementation. Epically Powerful contains a code base enabling simplified user implementation via Python that seamlessly interfaces with various commercial state-of-the-art quasi-direct drive (QDD) actuators, single-board computers, and common sensors, provides example controllers, and enables real-time visualization. To further support device development, the package also includes a recommended parts list and compatibility guide and detailed documentation on hardware and software implementation. The goal of Epically Powerful is to lower the barrier to developing and deploying custom wearable robotic systems without a pre-specified form factor, enabling researchers to go from raw hardware to modular, robust devices quickly and effectively. Though originally designed with wearable robotics in mind, Epically Powerful is broadly applicable to other robotic domains that utilize QDD actuators, single-board computers, and sensors for closed-loop control.

BACKGROUND: Previous studies on microprocessor-controlled prosthetic knees (MPKs) often investigate benefits of MPKs as a class of knees rather than clinically relevant differences between specific knees, despite their distinct features. OBJECTIVES: To systematically evaluate and report outcomes associated with three commercially available MPKs following a standardized real-world use period. DESIGN: Randomized crossover study. SETTING: Research laboratory and community environment. PARTICIPANTS: Ten patients with transfemoral amputation. INTERVENTIONS: Three MPKs were fitted, trained, and worn for a 1-week period including C-Leg 4.0 (Ottobock, Duderstadt, Germany), Rheo Knee-Model RM7 (Össur, Reykjavik, Iceland), and Power Knee-PKA01 (Össur, Reykjavik, Iceland). MAIN OUTCOME MEASURES: Primary outcomes were the 10-meter walk test (10-mwt), the 2-minute walk test (2-mwt), and the Prosthesis Evaluation Questionnaire (PEQ). Secondary outcomes were stance time asymmetry, physiological cost index, stair and ramp speeds, the narrowing beam walking test, and community ambulation monitoring. RESULTS: Participants walked 11% faster in Rheo than Power Knee during the 10-mwt (95% confidence interval [CI]: 0.046-0.184, p = .015). In the 2-mwt, participants walked 12% faster in C-Leg (95% CI: 0.034-0.241, p = .003) and 9% faster in Rheo (95% CI: 0.031, 0.163, p = .027) than in Power Knee. On the PEQ, participants reported greater satisfaction with C-Leg compared to Power Knee (p = .006). Ramp ascent speed was 8% faster in Rheo than Power Knee (95% CI: 0.026-0.130, p = .024). No significant differences were found for other secondary outcomes. Notably, 10 of 12 outcomes showed individuals performing their best by a defined difference on an MPK different from the cohort's best-performing MPK. CONCLUSIONS: Participants walked faster in C-Leg and Rheo than Power Knee and reported greater satisfaction with C-Leg. Consideration of patient needs and characteristics may allow more individualized MPK prescription and thereby improve rehabilitation outcomes. DATABASE REGISTRATION: NCT06399471.

Accurate walking speed estimation in lower-limb prostheses is crucial for delivering biomechanically appropriate assistance across varying speeds. However, training robust models requires extensive domain-specific, user-dependent (DEP) data, which is impractical for every new prosthesis user. This study presents a transfer learning framework to simplify and enhance the training process. Convolutional neural networks were pre-trained on publicly available datasets from able-bodied (AB) individuals and transfemoral amputees using the Open Source Leg (OSL) knee-ankle prosthesis, then fine-tuned with data from a transfemoral amputee using the Power Knee (PK) prosthesis. The fine-tuned models, AB-PK and OSL-PK were trained with varying data amounts and evaluated across constant and variable walking speed trials, with performance compared to DEP models trained from scratch on PK data. Training and testing were conducted on a per-subject basis, with performance averaged across subjects (N=7). The lowest post-fine-tuning error was observed in AB-PK, with RMSE values of 0.041 m/s for constant speeds, 0.072 m/s for variable speeds, and 0.088 m/s for novel speeds not included in the original training data. Significant error reductions were observed in both fine-tuned models compared to DEP when fewer than 30 gait cycles per speed of training data were available. Notably, AB datasets appeared highly viable for this application and may even outperform OSL datasets in transfer learning for walking speed estimation, perhaps due to the much larger original training dataset. This approach highlights the potential of transfer learning across different subject populations and devices, offering insights into the data needed to achieve state-of-the-art speed estimation.

Abstract Recent advancements in data-driven methods have enabled real-time estimation of biomechanical states for exoskeleton control. While biological joint moments can be directly used to scale exoskeleton assistance, this approach is often suboptimal. An optimized mapping between biological joint moments and exoskeleton assistance could enhance end-to-end controllers based on the user’s physiological state. We introduce a flexible parametrization of biological moment-based control using delay, scaling, and shaping terms to transform joint moment estimates into commanded torque. We performed human-in-the-loop optimization, using metabolic cost to evaluate each iteration’s controller parameters, for 9 subjects across three ambulation modes: level walking at 1.1 m/s, 1.5 m/s, and 5° inclined walking. We evaluated three methods of exoskeleton control: 1. Personalized/Task Dependent, 2. Task Dependent/Non-personalized, and 3. Task Agnostic/Non-personalized. On average, our personalized approach provided the greatest benefit of 18.3% reduction in metabolic cost compared to walking without the exoskeleton, with the task dependent and task agnostic controllers producing similar reductions of 8.6% and 8.4%, respectively. Our results show that while generalizable, task agnostic control parameters can improve user energetics across cyclic tasks, fully personalized exoskeleton control parameters yield larger metabolic reductions, highlighting the value of personalizing exoskeleton assistance to users across many diverse tasks.

Lower limb exoskeletons show promise to assist human movement, but their utility is limited by controllers designed for discrete, predefined actions in controlled environments, restricting their real-world applicability. We present an uncertainty-aware control framework that enables ankle exoskeletons to operate safely across diverse scenarios by automatically disengaging when encountering unfamiliar movements. Our approach uses an uncertainty estimator to classify movements as similar (in-distribution) or different (out-of-distribution) relative to actions in the training set. We evaluated three architectures (model ensembles, autoencoders, and generative adversarial networks) on an offline dataset and tested the strongest performing architecture (ensemble of gait phase estimators) online. The online test demonstrated the ability of our uncertainty estimator to turn assistance on and off as the user transitioned between in-distribution and out-of-distribution tasks (F1: 89.2). This new framework provides a path for exoskeletons to safely and autonomously support human movement in unstructured, everyday environments.

A primary challenge in continual learning (CL) for wearable robotics, especially prosthetics, is balancing the need to retain learned knowledge (stability) with the necessity to adapt to new information (plasticity). This balance is crucial for online adaptation, enabling systems to transition between tasks without losing prior knowledge. In this paper, we introduce a novel online optimizer-based framework designed to manage the stability-plasticity balance through strategic datapoint replay and learning-rate adjustments of a deep neural network. We applied this framework to speed estimation systems for transfemoral prostheses (TFA users), conducting offline validation tests using data from 10 individuals with TFA, and online tests with three TFA and six able-bodied (AB) participants. Our results demonstrate statistically significant improvements: in offline settings, our method showed a 39.2% increase in stability and a 35.2% boost in plasticity over traditional CL approaches during leave-one-subject-out validation. Similarly, in real-time trials with AB participants, we observed statistically significant gains in handling both previously encountered and new walking speeds. Finally, trials with individuals with TFA showed that the system improved the plasticity of the baseline model by 67.45% and the stability of the traditional CL approach by 31.36%; reducing overall average walking speed estimation error by 19.47%.

OBJECTIVE: Real-time determination of human kinematics and kinetics could advance biomechanics research and enable valuable applications of biofeedback and generalizable exoskeleton control. This work aims to investigate a task-independent, user-independent method for obtaining precise real-time joint state estimation across lower-body joints during a wide variety of tasks. METHODS: We developed a generalizable sensing approach using a suit comprised of inertial measurement units (IMUs) and pressure insoles. With the suit, we collected a dataset of 33 tasks commonly performed during construction and hazardous waste cleanup (N = 10). We then trained deep learning user-independent, task-agnostic models to estimate joint lower-body kinematics and dynamics using only worn sensor data. We likewise computed joint kinematics and dynamics analytically from sensor data to serve as a comparison tool for model results. RESULTS: Our models achieved overall angle estimation root-mean-squared-errors (RMSE) of 6.56±.92°, 8.60±1.01°, 7.58±.89°, and 6.00±.73° compared to 13.9±.1.3°, 15.31±1.0°, 10.76±.70°, and 7.56±.48° via analytical methods at the lower back, hip, knee, and ankle, respectively. Likewise, our models achieved overall normalized moment estimation RMSEs of .207±.069 Nm/kg, .242±.044 Nm/kg, .202±.038 Nm/kg, and .193±.034 Nm/kg compared to .306±.036 Nm/kg, .407±.021 Nm/kg, 1.18 ±.022 Nm/kg, and 1.73±.071 Nm/kg via analytical methods at the lower back, hip, knee, and ankle, respectively. CONCLUSION: These results are comparable to other state-of-the-art wearable sensing systems, establishing deep learning as a viable sensing approach that generalizes to new users and tasks. SIGNIFICANCE: This work shows promise for enabling accurate real-world biomechanical data collection and enhancement of biofeedback systems and wearable robot control.

3D surface reconstruction is essential across applications of virtual reality, robotics, and mobile scanning. However, RGB-based reconstruction often fails in low-texture, low-light, and low-albedo scenes. Handheld LiDARs, now common on mobile devices, aim to address these challenges by capturing depth information from time-of-flight measurements of a coarse grid of projected dots. Yet, these sparse LiDARs struggle with scene coverage on limited input views, leaving large gaps in depth information. In this work, we propose using an alternative class of "blurred" LiDAR that emits a diffuse flash, greatly improving scene coverage but introducing spatial ambiguity from mixed time-of-flight measurements across a wide field of view. To handle these ambiguities, we propose leveraging the complementary strengths of diffuse LiDAR with RGB. We introduce a Gaussian surfel-based rendering framework with a scene-adaptive loss function that dynamically balances RGB and diffuse LiDAR signals. We demonstrate that, surprisingly, diffuse LiDAR can outperform traditional sparse LiDAR, enabling robust 3D scanning with accurate color and geometry estimation in challenging environments.

BACKGROUND: Current processes for identifying the best microprocessor-controlled prosthetic knee (MPK) for individuals with transfemoral amputations are subjective, nonscientific, and sometimes fail to consider unique patient needs. Inaccurate prescriptions may hinder a patient's ability to make a speedy rehab. OBJECTIVES: We developed a clinical decision equation that outputs MPK recommendation scores for 3 commercially available MPKs (Power Knee, C-Leg 4.0, Rheo Knee) based on easily acquirable user evaluation data. STUDY DESIGN: Participants wore each of the study MPKs at home for a 1-week acclimation period. On the experiment day, participants completed a set of functional tasks including a 10-m walk test, stair and ramp ambulation tasks, a 2-minute walk test, and a narrow beam walking test. Performance outcome measures were collected. METHODS: Microprocessor-controlled prosthetic knees were scored relatively to the best performing knee based on their performance in 5 areas of interest: agility, community ambulation, energy, stability, and gait quality. The relative importance of each of these areas was computed based on a quantitative prediction of a user's functional needs from features including age, body mass index (BMI), AMPnoPRO score, and likelihood of stairs/ramps. We describe the algorithm-suggested optimal patient profiles for each device. RESULTS: We developed an application that allows clinicians to obtain instant recommendations. Clinicians can further adjust the relative importance of each area of interest based on patient needs. CONCLUSIONS: This algorithm represents a transparent, experimentally backed clinical decision-making aid with the potential to streamline the prosthesis fitting process. Future studies are required to evaluate the effectiveness of the algorithm.

OBJECTIVE: Accurate, non-invasive methods for estimating joint and muscle physiological states have the potential to greatly enhance control of wearable devices during real-world ambulation. Traditional modeling approaches and current estimation methods used to predict muscle dynamics often rely on complex equipment or computationally intensive simulations and have difficulty estimating across a broad spectrum of tasks or subjects. METHODS: Our approach used deep learning (DL) models trained on kinematic inputs to estimate internal physiological states at the knee, including moment, power, velocity, and force. We assessed each model's performance against ground truth labels from both a commonly used, standard OpenSim musculoskeletal model without EMG (static optimization) and an EMG-informed method (CEINMS), across 28 different cyclic and noncyclic tasks. RESULTS: EMG provided no benefit for joint moment/power estimation (e.g., biological moment), but was critical for estimating muscle states. Models trained with EMG-informed labels but without EMG as an input to the DL system significantly outperformed models trained without EMG (e.g., 33.7% improvement for muscle moment estimation) (p < 0.05). Models that included EMG-informed labels and EMG as a model input demonstrated even higher performance (49.7% improvement for muscle moment estimation) (p < 0.05), but require the availability of EMG during model deployment, which may be impractical. CONCLUSION/SIGNIFICANCE: While EMG information is not necessary for estimating joint level states, there is a clear benefit during muscle level state estimation. Our results demonstrate excellent tracking of these states with EMG included only during training, highlighting the practicality of real-time deployment of this approach.

Imaging system design is a complex, time-consuming, and largely manual process; LiDAR design, ubiquitous in mobile devices, autonomous vehicles, and aerial imaging platforms, adds further complexity through unique spatial and temporal sampling requirements. In this work, we propose a framework for automated, task-driven LiDAR system design under arbitrary constraints. To achieve this, we represent LiDAR configurations in a continuous six-dimensional design space and learn task-specific implicit densities in this space via flow-based generative modeling. We then synthesize new LiDAR systems by modeling sensors as parametric distributions in 6D space and fitting these distributions to our learned implicit density using expectation-maximization, enabling efficient, constraint-aware LiDAR system design. We validate our method on diverse tasks in 3D vision, enabling automated LiDAR system design across real-world-inspired applications in face scanning, robotic tracking, and object detection.

Robust autonomous navigation in environments with limited visibility remains a critical challenge in robotics. We present a novel approach that leverages Non-Line-of-Sight (NLOS) sensing using single-photon LiDAR to improve visibility and enhance autonomous navigation. Our method enables mobile robots to “see around corners” by utilizing multi-bounce light information, effectively expanding their perceptual range without additional infrastructure. We propose a three-module pipeline: (1) Sensing, which captures multi-bounce histograms using SPAD-based LiDAR; (2) Perception, which estimates occupancy maps of hidden regions from these histograms using a convolutional neural network; and (3) Control, which allows a robot to follow safe paths based on the estimated occupancy. We evaluate our approach through simulations and real-world experiments on a mobile robot navigating an L-shaped corridor with hidden obstacles. Our work represents the first experimental demonstration of NLOS imaging for autonomous navigation, paving the way for safer and more efficient robotic systems operating in complex environments. We also contribute a novel dynamics-integrated transient rendering framework for simulating NLOS scenarios, facilitating future research in this domain.

Robotic and biofeedback-assisted interventions are promising alternatives to surgical intervention and supplements for traditional physical therapy for children with gait impairments. This work utilizes a human-in-the-loop optimization strategy to adaptively modulate parameters for a lightweight robotic knee exoskeleton and biofeedback video game to maximize learning potential following the challenge point framework. We tested our approach on three able-bodied participants and one pediatric patient with genu recurvatum, a common walking pattern in children with neurological injuries. We implement a Covariance Matrix Adaptation-Evolutionary Strategy (CMA-ES) optimizer to enforce a target success rate of 70 % by continuously adjusting visual biofeedback and exoskeleton assistance parameters. Our experimental results demonstrate the system's ability to maintain the target challenge level for the pediatric participant. Stance hyperextension decreased significantly from pre- to post-training trials on day $2\left(9.2^{\circ}\right)$ and $3\left(3.2^{\circ}\right)$ of the case study. Swing flexion approached the clinical target of 65° by the end of the third day. The promising optimizer performance and changes in gait kinematics validate the feasibility of autonomous parameter tuning to maximize learning potential in pediatric gait rehabilitation.

Powered and passive knee-and-ankle prostheses can restore mobility for individuals with transfemoral amputation (TFA), but their effects on biological joints remain underexplored. Overuse of biological joints with prostheses may lead to chronic pain. This study compared biological joint work during ramp and stair ascent and descent for nine individuals with TFA using the powered prosthesis compared to the passive prosthesis. We hypothesized that the powered prosthesis would reduce positive mechanical work in ascent due to active knee extension and the negative mechanical work in descent due to controlled energy dissipation. In ascent, the powered prosthetic knee generated more positive work (p < 0.05), reducing sound-side hip joint work by 29.3 % (CI: [1.5 %, 57.1 %]; p = 0.041) on ramps and 22.8 % (CI: [7.2 %, 38.3 %]; p = 0.019) on stairs. The powered prosthesis reduced biological joint work by 50.6 % (CI: [2.7 %, 98.4 %]; p = 0.041) during swing phase on ramp ascent. In descent, the powered prosthetic ankle absorbed twice the negative work on ramps (CI: [164.9 %, 269.9 %]; p = 0.001) and 2.5 times on stairs (CI: [-73.5 %, 372.9 %]; p = 0.145) by acting as a virtual rotational damper instead of a spring. No significant reductions in biological work were seen in descent tasks, though magnitudes were generally lower. Overall, the powered knee provided biomechanical benefits in ramp and stair ascent, while the powered ankle provided mild benefits in ramp and stair descent. However, the intact joint work remains elevated compared to able-bodied individuals, highlighting the need for further prosthetic improvements.

Ankle exoskeletons have the potential to augment mobility, but control strategies have largely failed to seamlessly adapt to changes in the locomotion task. Here, we introduce a multi-headed network that predicts gait speed, ground incline, stance/swing transitions, and percent stance. These predictions are mapped to exoskeleton torque using typical biological torques as a guide. The model was trained on 9 subjects walking/jogging for 12 minutes across a range of speeds and inclines. The controller was validated on 4 subjects, and achieved stance phase prediction error of 3.4% across a range of speeds and inclines, both inside and outside the training set distribution. A secondary analysis showed similar accuracy could have been obtained with only 10% of the collected data, suggesting researchers may need fewer total strides of training data, provided the data is sufficiently diverse across users and tasks. Metabolic cost was improved during running compared to wearing the exoskeleton powered off, but was beneficial for only one subject during level walking and ramp ascent when compared to no exoskeleton. Overall, our controller smoothly adapted to time-varying inclines and walking/jogging speeds, and achieved high accuracy with a reduced training dataset, though larger torque magnitudes may be required to see metabolic benefit.

Fall incidents due to slips are some of the most common causes of injuries for industry workers and older adults, motivating research to assist balance recovery following slips. To assist balance recovery during a slip, a detection algorithm that can work with an assistive device, such as an exoskeleton, needs to be able to detect slips rapidly after onset, which remains a critical gap in the field. Here, we compared the ability of linear discriminant analysis (LDA), extreme gradient boosting (XGBoost), and convolutional neural networks (CNN) to detect slip using only native sensors on a hip exoskeleton. We trained and evaluated user-independent models on early-stance (ES) and late-stance (LS) slips of various magnitudes collected through treadmill-based slips. All models, except LDA with LS slips, detected slips with >90% accuracy. Overall, he best model was XGBoost, with its fastest results achieving average detection times and median accuracies of 155.06 ms at 96.25% for ES slips and 228.88 ms at 93.75% for LS slips, while also achieving 100% sensitivity at 195.64 ms (ES) and 266.24 ms (LS). Our results indicate a promising direction for further research into designing a generalizable model for balance recovery during slip perturbations using robotic hip exoskeletons.

BACKGROUND Individuals with transfemoral amputation often report difficulty with ambulating on inclined surfaces. Conventional prosthetic control strategies typically apply a level walking controller in incline walking modes, which may not be biomechanically optimal. Able-bodied individuals modulate knee stance pre-flexion substantially during incline walking, which is absent in most prosthetic level walking controllers. However, the biomechanical effects of stance pre-flexion for users with robotic microprocessor-controlled knees are not well-explored during inclines. METHODS In this study (n = 10), we investigated the joint kinematics/kinetics/power, biological joint level work and metabolic energy cost to evaluate the biomechanical effects of stance pre-flexion on a 7.5o incline walking using a commercially available robotic prosthetic knee, the Össur Power Knee, and a passive foot, the Össur Pro-Flex LP. We ran a Bradley-Terry model to rank user preferences on stance pre-flexion conditions. FINDINGS We found that a 16.7 % reduction on the contralateral biological ankle joint positive work during stance phase when stance pre-flexion increased (p < 0.01). However, there was no significant difference in metabolic energy cost. Survey data revealed participants preferred higher stance pre-flexion angles (12o -18o) compared to lower angles (0o - 6o), indicating consistent preference towards increased stance pre-flexion on inclines. INTERPRETATION Our results indicate that reduction in biological joint work associated with stance pre-flexion emphasizes the need to implement stance pre-flexion adjustments in prosthesis controllers, as opposed to using a level-walking controller.

This study introduces a novel continual learning algorithm that incrementally improves the performance of deep-learning-based walking speed estimators during level-ground walking with a powered knee-ankle prosthesis. While user-dependent (DEP) estimators generally outperform user-independent (IND) estimators, they require the pre-collection of DEP training data. In contrast, our real-time algorithm adapts IND estimators to self-labeled DEP data generated during walking, eliminating the need for pre-collected datasets. The algorithm also features a biomimetic scaling mechanism that adjusts prosthetic assistance based on speed estimates. We evaluated our algorithm on novel subjects (N=10) with unilateral above-knee amputations during treadmill and overground walking. For treadmill trials, when adapted with estimated and ground truth labels, estimators achieved mean absolute errors (MAEs) of 0.074 [0.023] (mean, [standard deviation]) and 0.074 [0.018] m/s, respectively, reflecting a significant 28% (p ¡ 0.05) reduction in MAE compared to non-adapted estimators. For overground trials, treadmill-adapted estimators demonstrated a significant 18% (p ¡ 0.05) reduction in MAE compared to non-adapted estimators. Our algorithm significantly reduced speed estimation errors within one minute of walking and delivered biomimetic assistance (r <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">${=}0.91$ </tex-math></inline-formula>) across speeds. This approach allows off-the-shelf powered prostheses to seamlessly adapt to new users, delivering biomimetic assistance through precise, real-time walking speed estimation.

Human-robot teams in dynamic environments have the potential to leverage robot sensing and intelligence to augment human performance through motion suggestions. More specifically, we examine how humans can use external sensors (fixed or robotic) and a motion planning algorithm to help them navigate environments with dynamic obstacles. The novel human-centric velocity-obstacle (HCVO) algorithm suggests a feasible goal-oriented action while avoiding obstacles. Participants were placed in a custom virtual reality (VR) environment and tasked to follow a dynamic goal while avoiding collisions. We demonstrate, over N = 10 participants, that the HCVO algorithm’s guidance significantly improves safety compared to a base VO algorithm. We then examine the performance of N = 15 participants in three conditions: (1) no assistance/control, (2) a top-down drone-view of the entire environment, and 3) motion planner-informed suggestions. The core contributions of this research include (1) introducing and tuning of a human-centric velocity-obstacle (HCVO) algorithm, (2) demonstrating the benefits of the HCVO algorithm compared to a base VO algorithm, (3) demonstrating the benefits of the HCVO algorithm compared with a standard overhead drone view. Long term, the deployment of effective human-centric motion planners can make people safer from workplace to warzone. Code: https://github.com/ROAMR-GT/HCVO-Game . Video: https://www.youtube.com/watch?v=9ITD1GBBz24 .

OBJECTIVE: This study applies human-in-the-loop optimization to identify optimal hip exoskeleton assistance patterns for stair climbing. METHODS: Ten participants underwent optimization to individualize hip flexion and extension assistance, followed by a validation comparing optimized assistance (OPT) to biological hip moment-based assistance (BIO), no assistance (No-Assist), and no exoskeleton (No-Exo) conditions. RESULTS: OPT reduced metabolic cost by 4.5% compared to No-Exo, 11.44% compared to No-Assist, and 5.02% compared to BIO, demonstrating the effectiveness of the optimization approach. Statistical analysis revealed distinct characteristics in optimal assistance timing and magnitude that deviated systematically from biological hip moment patterns. Compared to BIO, OPT exhibited later peak flexion timing (76.4 $\pm$ 3.7% vs 65.0%), shorter flexion duration (29.2 $\pm$ 3.6% vs 40.0%), later peak extension timing (26.7 $\pm$ 3.8% vs 20.0% of gait cycle), and higher peak flexion magnitude (11.1 $\pm$ 1.5 Nm vs 10.0 Nm). While individual optimal assistance profiles varied across participants, comparison between individually optimized parameters and the best subject-independent parameters identified through post-hoc analysis showed consistency. On average, metabolic rate convergence was achieved after 18 iterations, while most exoskeleton control parameters did not reach our convergence criteria within 20 iterations. CONCLUSION: These findings demonstrate that human-in-the-loop optimization can effectively identify task-specific assistance patterns for stair climbing, while the consistency between individual and subject-independent parameters suggests the potential for developing generalized assistance strategies. The systematic differences between optimized and biological moment-based assistance underscore the fundamental distinctions between biological torque-based control and optimal control for exoskeleton assistance.

Deploying machine vision for wearable robot control faces challenges in terms of usability, reliability, privacy, and costs.

Traditional robotic knee-ankle prostheses categorize ambulation modes such as level walking, ramps, and stairs. However, human movement scales continuously across various states rather than discretely, making traditional mode classifiers inadequate for accurate intent recognition. This paper proposes a mode-unified intent recognition strategy that continuously estimates terrain slopes across five modes: level ground, ramp ascent/descent, and stair ascent/descent. Locomotion data from 16 individuals with transfemoral amputation were utilized to train slope estimation and mode classification models based on deep temporal convolutional networks. The proposed method was compared to the traditional mode classifier via offline test, using leave-one-subject-out validations for the user-independent performance. The mode-unified slope estimator achieved an MAE of 1.68 ± 0.60 degrees, outperforming the mode classifier's MAE of 1.94 ± 0.97 degrees (p<0.05). The lower slope estimation errors resulted in higher accuracy in replicating knee kinematics of able-bodied subjects, with the proposed system achieving an average MAE of 5.13 ± 2.00 degrees for knee clearance and 6.74 ± 2.97 degrees for knee contact angle, compared to the traditional classifier's 12.10 ± 5.20 degrees and 13.80 ± 3.28 degrees (p<0.01), respectively, in stair ascent. These results suggest that our mode-unified approach can enable continuous adjustment to terrains without mode classification.

Humans operating in dynamic environments with limited visibility are susceptible to collisions with moving objects, occupational hazards, and/or other agents, which can result in personal injuries or fatalities. Most existing research has focused on using vibrotactile cues to address this challenge. In this work, we propose a fundamentally new approach that utilizes variable impedance on an active exoskeleton to guide humans away from hazards and towards safe areas. This framework combines artificial potential fields with current impedance-based theories of exoskeleton control to provide a comprehensive navigational system that is intuitive for human operators. First, we present the mathematical framework to encode information about the locations of obstacles and the safest direction in which to move. Next, we optimize controller parameters in a series of human-subject experiments. Finally, we evaluate the framework in virtual reality on a set of randomly generated obstacle fields in environments where vision is either fully or partially occluded. Our results suggest that the exoskeleton provides significant separation from obstacles and reduced collisions compared to vision alone in conditions where visibility was limited to less than 1.3 m. Our work demonstrates that force-feedback in parallel with a human can improve overall navigation ability in low visibility conditions.

Vision systems in nature show remarkable diversity, from simple light-sensitive patches to complex camera eyes with lenses. While natural selection has produced these eyes through countless mutations over millions of years, they represent just one set of realized evolutionary paths. Testing hypotheses about how environmental pressures shaped eye evolution remains challenging since we cannot experimentally isolate individual factors. Computational evolution offers a way to systematically explore alternative trajectories. Here we show how environmental demands drive three fundamental aspects of visual evolution through an artificial evolution framework that co-evolves both physical eye structure and neural processing in embodied agents. First, we demonstrate computational evidence that task specific selection drives bifurcation in eye evolution - orientation tasks like navigation in a maze leads to distributed compound-type eyes while an object discrimination task leads to the emergence of high-acuity camera-type eyes. Second, we reveal how optical innovations like lenses naturally emerge to resolve fundamental tradeoffs between light collection and spatial precision. Third, we uncover systematic scaling laws between visual acuity and neural processing, showing how task complexity drives coordinated evolution of sensory and computational capabilities. Our work introduces a novel paradigm that illuminates evolutionary principles shaping vision by creating targeted single-player games where embodied agents must simultaneously evolve visual systems and learn complex behaviors. Through our unified genetic encoding framework, these embodied agents serve as next-generation hypothesis testing machines while providing a foundation for designing manufacturable bio-inspired vision systems. Website: http://eyes.mit.edu/

Repetitive manual labor tasks involving twisting, bending, and lifting commonly lead to lower back and knee injuries in the workplace. To identify tasks with high injury risk, we recruited N = 9 participants to perform industry-relevant, 2-handed lifts with a 11-kg weight. These included symmetrical/asymmetrical, ascending/descending lifts that varied in start-to-end heights (knee-to-waist and waist-to-shoulder). We used a data-driven musculoskeletal model that combined force and motion data with a muscle activation-informed solver (OpenSim, CEINMS) to estimate 3-dimensional internal joint contact forces (JCFs) in the lower back (L5/S1) and knee. Symmetrical lifting resulted in larger peak JCFs than asymmetrical lifting in both the L5/S1 (+20.2% normal [P < .01], +20.3% shear [P = .001], +20.6% total [P < .01]) and the knee (+39.2% shear [P = .001]), and there were no differences in peak JCFs between ascending versus descending motions. Below-the-waist lifting generated significantly greater JCFs in the L5/S1 and knee than above-the-waist lifts (P < .01). We found a positive correlation between knee and L5/S1 peak total JCFs (R2 = .60, P < .01) across the task space, suggesting motor coordination that favors sharing of load distribution across the trunk and legs during lifting.

Robotic exoskeletons can transform mobility for individuals with lower-limb disabilities. However, their widespread adoption is limited by controller degradation caused by varying gait dynamics across different users and environments. Here, we propose an online adaptation framework that leverages real-time data streams to continuously update the user state estimator model. This approach allows the exoskeleton to learn the user-specific gait patterns, effectively customizing the model for each new user. Additionally, we demonstrate a sensor signal transformation technique that enables model transfer across different exoskeleton hardware (from a research-grade exoskeleton to a commercial device). With less than one minute of adaptation, our framework improved gait phase estimation, which directly affects assistance timing, by 40.9% for able-bodied subjects and 65.9% for stroke survivors (p<0.05), and reduced torque profile error by 32.7% compared to the baseline model (p<0.05). Furthermore, in a pilot test, we applied our adaptation framework with human-in-the-loop optimization for control tuning. In a single stroke survivor, this approach led to a 21.8% increase in walking speed and a 6.5% reduction in metabolic cost compared to walking without exoskeleton. While preliminary, these results suggest the potential for personalized exoskeleton assistance in clinical populations.

Exoskeletons offer promising solutions for improving human mobility, but a key challenge is ensuring the controller adapts to changing walking conditions. We present an artificial intelligence (AI)-driven universal exoskeleton system that dynamically switches assistance types between walking modes, modulates assistance levels corresponding to the ground slope, and delivers assistance timely based on the current gait phase in real-time. During treadmill validation, AI-based assistance reduced metabolic cost by 6.5% compared to 3.5% for conventional assistance. We expanded testing the controller in real-world walking, where AI-based assistance showed effective modulation and higher user preference compared to conventional assistance. Leveraging the AI-based approach and a comprehensive dataset, the controller achieved superior performance in environment- and user-state estimations. This approach does not require a separate mode classifier and operates on a user-independent basis, enabling immediate deployment across diverse conditions. This study highlights the potential of AI-driven exoskeletons in facilitating human locomotion in real-world ambulation.

Post-stroke gait is slow, energetically costly, and unstable. Rehabilitation is necessary to encourage, retrain, and assist proper gait mechanics in stroke survivors. Evidence indicates robotic ankle exoskeletons can improve gait outcomes in stroke survivors, however challenges remain with proper lower limb positioning for optimal receipt of the assistance. Biofeedback can be used to improve positioning of the limb for receipt of robotic ankle exoskeleton assistance. In this study, four stroke survivors used bilateral powered robotic ankle exoskeletons (Dephy Exoboots) and an innovative, custom-designed vibrotactile-audio biofeedback interface targeting trailing limb angle to test the hypotheses that each intervention alone improves gait outcomes over baseline, and when combined they improve outcomes over either intervention alone. Compared to baseline, we found increases in average paretic propulsive impulse during the biofeedback-only and exoskeleton-plus-biofeedback conditions. Biofeedback alone induced the greatest increase on average self-selected walking speed, and the combination of exoskeleton assistance and biofeedback increased speed more compared to the robotic exoskeleton-only condition. Our preliminary results indicate that biofeedback in combination with a robotic exoskeleton produces greater synergistic benefits on gait performance than the use of an exoskeleton alone.

Community ambulation is essential for maintaining a healthy lifestyle, but it poses significant challenges for individuals with limb loss due to complex task demands. In wearable robotics, particularly powered prostheses, there is a critical need to accurately estimate environmental context, such as walking speed and slope, to offer intuitive and seamless assistance during varied ambulation tasks. We developed a user-independent and multicontext, intent recognition system that was deployed in real-time on an Open Source Leg (OSL). We recruited 11 individuals with transfemoral amputation, with seven participants used for real-time validation. Our findings revealed two main conclusions: (1) the user-independent (IND) performance across speed and slope was not statistically different from user-dependent (DEP) models in real-time and did not degrade compared to its offline counterparts, and (2) IND walking speed estimates showed ∼0.09 m/s mean absolute error (MAE) and slope estimates showed ∼0.95 deg MAE across multicontext scenarios. Additionally, we provide an open-source dataset to facilitate further research in accurately estimating speed and slope using an IND approach in real-world walking tasks on a powered prosthesis.

3D surface reconstruction is essential across applications of virtual reality, robotics, and mobile scanning. However, RGB-based reconstruction often fails in low-texture, low-light, and low-albedo scenes. Handheld LiDARs, now common on mobile devices, aim to address these challenges by capturing depth information from time-of-flight measurements of a coarse grid of projected dots. Yet, these sparse LiDARs struggle with scene coverage on limited input views, leaving large gaps in depth information. In this work, we propose using an alternative class of "blurred" LiDAR that emits a diffuse flash, greatly improving scene coverage but introducing spatial ambiguity from mixed time-of-flight measurements across a wide field of view. To handle these ambiguities, we propose leveraging the complementary strengths of diffuse LiDAR with RGB. We introduce a Gaussian surfel-based rendering framework with a scene-adaptive loss function that dynamically balances RGB and diffuse LiDAR signals. We demonstrate that, surprisingly, diffuse LiDAR can outperform traditional sparse LiDAR, enabling robust 3D scanning with accurate color and geometry estimation in challenging environments.

Robust autonomous navigation in environments with limited visibility remains a critical challenge in robotics. We present a novel approach that leverages Non-Line-of-Sight (NLOS) sensing using single-photon LiDAR to improve visibility and enhance autonomous navigation. Our method enables mobile robots to "see around corners" by utilizing multi-bounce light information, effectively expanding their perceptual range without additional infrastructure. We propose a three-module pipeline: (1) Sensing, which captures multi-bounce histograms using SPAD-based LiDAR; (2) Perception, which estimates occupancy maps of hidden regions from these histograms using a convolutional neural network; and (3) Control, which allows a robot to follow safe paths based on the estimated occupancy. We evaluate our approach through simulations and real-world experiments on a mobile robot navigating an L-shaped corridor with hidden obstacles. Our work represents the first experimental demonstration of NLOS imaging for autonomous navigation, paving the way for safer and more efficient robotic systems operating in complex environments. We also contribute a novel dynamics-integrated transient rendering framework for simulating NLOS scenarios, facilitating future research in this domain.

The incredible diversity of visual systems in the animal kingdom is a result of millions of years of coevolution between eyes and brains, adapting to process visual information efficiently in different environments. We introduce the generative design of visual intelligence (GenVI), which leverages computational methods and generative artificial intelligence to explore a vast design space of potential visual systems and cognitive capabilities. By cogenerating artificial eyes and brains that can sense, perceive, and enable interaction with the environment, GenVI enables the study of the evolutionary progression of vision in nature and the development of novel and efficient artificial visual systems. We anticipate that GenVI will provide a powerful tool for vision scientists to test hypotheses and gain new insights into the evolution of visual intelligence while also enabling engineers to create unconventional, task-specific artificial vision systems that rival their biological counterparts in terms of performance and efficiency.

Crouch gait is one of the most common compensatory walking patterns found in individuals with neurological disorders, often accompanied by their limited physical capacity. Notable kinematic characteristics of crouch gait are excessive knee flexion during stance and reduced range of motion during swing. Knee exoskeletons have the potential to improve crouch gait by providing precisely controlled torque assistance directly to the knee joint. In this study, we implemented a finite-state machine-based impedance controller for a powered knee exoskeleton to provide assistance during both stance and swing phases for five children and young adults who exhibit chronic crouch gait. The assistance provided a strong orthotic effect, increasing stance phase knee extension by an average of 12 deg. Additionally, the knee range of motion during swing was increased by an average of 15 deg. Changes to spatiotemporal outcomes, such as preferred walking speed and percent stance phase, were inconsistent across subjects and indicative of the underlying intricacies of user response to assistance. This study demonstrates the potential of knee exoskeletons operating in impedance control to mitigate the negative kinematic characteristics of crouch gait during both stance and swing phases of gait.

Reducing medical costs is one of the three strategic aims of U.S. health care reform. This research quantifies potential expense reductions of U.S. health insurers and examines portfolio effects on expenses of differential health insurance mixes. We use the coronavirus disease 2019 (COVID-19) pandemic as a natural stress test for potential savings in medical services. We employ a two-stage residual inclusion generalized linear model and uncover differential expense reductions for four major health insurance markets: individual, group, Medicaid managed care, and Medicare Advantage. Our results could serve as a benchmark for reducing medical costs through redesigning insurance coverages following the most effective group insurance model. The results also provide policy implications on restructuring health insurance markets to increase efficiency. We empirically examine mutual impacts among major health insurance markets, and document explicitly three optimal health insurance portfolios in reducing expenses of health care: Medicaid managed care and individual plans, exclusively group plans, and Medicaid managed care and Medicare Advantage plans.

Photoelasticity enables full-field stress analysis in transparent objects through stress-induced birefringence. Existing techniques are limited to 2D slices and require destructively slicing the object. Recovering the internal 3D stress distribution of the entire object is challenging as it involves solving a tensor tomography problem and handling phase wrapping ambiguities. We introduce NeST, an analysis-by-synthesis approach for reconstructing 3D stress tensor fields as neural implicit representations from polarization measurements. Our key insight is to jointly handle phase unwrapping and tensor tomography using a differentiable forward model based on Jones calculus. Our non-linear model faithfully matches real captures, unlike prior linear approximations. We develop an experimental multi-axis polariscope setup to capture 3D photoelasticity and experimentally demonstrate that NeST reconstructs the internal stress distribution for objects with varying shape and force conditions. Additionally, we showcase novel applications in stress analysis, such as visualizing photoelastic fringes by virtually slicing the object and viewing photoelastic fringes from unseen viewpoints. NeST paves the way for scalable non-destructive 3D photoelastic analysis.