Powered exoskeletons have the potential to revolutionize the labor workplace across many disciplines, from manufacturing to agriculture. However, there are still many barriers to adoption and widespread implementation of exoskeletons. One major research gap of powered exoskeletons currently is the development of a control framework to best cooperate with the user. This limitation is first in understanding the physical and cognitive interaction between the user and exoskeleton, and then in designing a controller that addresses this interaction in a way that provides both physical assistance towards completing a task, and a decrease in the cognitive demand of operating the device. This work demonstrates that multi-objective, optimization-based control can be used to provide a coincident implementation of autonomous robot control, and human-input driven control. A parameter called 'acceptance' can be added to the weights of the cost functions to allow for an automatic trade-off in control priority between the user and robot objectives. This is paired with an update function that allows for the exoskeleton control objectives to track the user objectives over time. This results in a cooperative, powered exoskeleton controller that is responsive to user input, dynamically adjusting control autonomy to allow the user to act to complete a task, learn the control objective, and then offload all effort required to complete the task to the autonomous controller. This reduction in effort is physical assistance directly towards completing the task, and should reduce the cognitive load the user experiences when completing the task. To test the hypothesis of whether high task assistance lowers the cognitive load of the user, a study is designed and conducted to test the effect of the shared autonomy controller on the user's experience operating the robot. The user operates the robot under zero-, full-, and shared-autonomy control cases. Physical workload, measured through the force they exert to complete the task, and cognitive workload, measured through pupil dilation, are evaluated to significantly show that high-assistance operation can lower the cognitive load experienced by a user alongside the physical assistance provided. Automatic adjustment in autonomy works to allow this assistance while allowing the user to be responsive to changing objectives and disturbances. The controller does not remove all mental effort from operation, but shows that high acceptance does lead to less mental effort. When implementing this control beyond the simple reaching task used in the study, however, the controller must be able to both track to the user's desired objective and converge to a high-assistance state to lead to the reduction in cognitive load. To achieve this behavior, first is presented a method to design and enforce Lyapunov stability conditions of individual tasks within a multi-objective controller. Then, with an assumption on the form of the input the user will provide to accomplish their intended task, it is shown that the exoskeleton can stably track an acceptance-weighted combination of the user and robot desired objectives. This guarantee of following the proper trajectory at corresponding autonomy levels results in comparable accuracy in tracking a simulated objective as the base shared autonomy approach, but with a much higher acceptance level, indicating a better match between the user and exoskeleton control objectives, as well as a greater decrease in cognitive load. This process of enforcing stability conditions to shape human-exoskeleton system behavior is shown to be applicable to more tasks, and is in preparation for validation with further user studies. / Doctor of Philosophy / Powered exoskeletons are robots that can be worn by users to physically aid them in accomplishing tasks. These robots differ in scale, from single-joint devices like powered ankle supports or lower-back braces for lifting, to large, multi-joint devices with a broad range of capabilities and potential applications. These multi-joint exoskeletons have been used in many applications such as medical rehabilitation robots, and labor-assisting devices for enhancing strength and avoiding injury. Broader use and adoption in industry could have a great positive impact on the experience of workers performing any heavy-labor tasks. There are still barriers to widespread adoption, however. When closely interacting with machinery like a powered exoskeleton, workers want guarantees of saftey, trust, and cooperation that current exoskeletons have not been able to provide. In fact, studies have shown that industrial devices capable of providing significant assistive force when accomplishing a task, also tend to impart additional, uncomfortable disturbance forces on the user. For example, a lower-body exoskeleton meant to help in lifting tasks might make the simple act of walking more difficult, both physically and mentally. There is a need for exoskeletons that are intuitively cooperative, and can provide both physical assistance towards completing a task and cognitive assistance that makes coordinating with the human user easier. In this dissertation we examine the control problem of powered exoskeletons. In the past, many powered exoskeleton controllers are direct, scripted controllers with exact objectives, or actions tied only to human input. To go beyond this, we leverage "multi-objective-control", originally designed for humanoid robots, which is capable of controlling the robot to accomplish multiple goals at the same time. This approach is the base on which a more complex controller can be created. We show first that the multi-objective control can be used to achieve human desired actions and robot autonomous control tasks at the same time, with a parameter to trade-off which actor, the human or the robot, has the priority control at that time. This framework has the capacity to allow the human to instruct the robot in tasks to accomplish, and then robot can fully mimic the user, offloading the physical effort required to accomplish the task. It is proposed that this offloading of effort from the user will also lower the cognitive load the user is under when actively commanding the exoskeleton. To test this hypothesis, a user study is conducted where human operators work with an upper-body powered exoskeleton to complete a simple reaching task. This study shows that on average, the more assistance the exoskeleton provides to the user, the lower their mental demand is. Additionally, when responding to new challenges or sudden disturbances, the robot can easily cooperate, balancing its own autonomy with the user's to allow the user to respond as they need to their changing environment, then resume active assistance when the change is resolved. Finally, to guarantee that the exoskeleton responds quickly and accurately to the user's intentions, a new strategy is derived to update the robot's internal objectives to match the users' goals. This strategy is based on the assumption that the exoskeleton knows what type of task the user is trying to complete. If this is true, then the exoskeleton can estimate the users objectives from the actions they task, and ensure assistance towards completing the task. This control design is proven in simulation, and in preparation for followup studies to evaluate the user experience of this improved strategy.

This thesis presents both a control scheme for naturalistic control of an exoskeleton glove and a glove design. Exoskeleton development has been focused primarily on design, improving soft actuator and cable-driven systems, with only limited focus on intelligent control. There is a need for control that is not limited to position or force reference signals and is user-driven. By implementing a motion amplification controller to increase weak movements of an impaired individual, a finger joint trajectory can be observed and used to predict their grasping intention. The motion amplification functions off of a virtual dynamical system that safely enforces the range of motion of the finger joints and ensures stability. Three grasp prediction algorithms are developed with improved levels of accuracy: regression, trajectory, and deep learning based. These algorithms were tested on published finger joint trajectories. The fusion of the amplification and prediction could be used to achieve naturalistic, user-guided control of an exoskeleton glove. The key to accomplishing this is series elastic actuators to move the finger joints, thereby allowing the wearer to deflect against the glove and inform the controller of their intention. These actuators are used to move the fingers in a nine degree of freedom exoskeleton that is capable of achieving all the grasps used most frequently in daily life. The controllers and exoskeleton presented here are the basis for improved exoskeleton glove control that can be used to assist or rehabilitate impaired individuals. / Master of Science / Millions of Americans report difficulty holding small or even lightweight objects. In many of these cases, their difficulty stems from a condition such as a stroke or arthritis, requiring either rehabilitation or assistance. For both treatments, exoskeleton gloves are a potential solution; however, widespread deployment of exoskeletons in the treatment of hand conditions requires significant advancement. Towards that end, the research community has devoted itself to improving the design of exoskeletons. Systems that use soft actuation or are driven by artificial tendons have merit in that they are comfortable to the wearer, but lack the rigidity required for monitoring the state of the hand and controlling it. Electromyography sensors are also a commonly explored technology for determining motion intention; however, only primitive conclusions can be drawn when using these sensors on the muscles that control the human hand. This thesis proposes a system that does not rely on soft actuation but rather a deflectable exoskeleton that can be used in rehabilitation or assistance. By using series elastic actuators to move the exoskeleton, the wearer of the glove can exert their influence over the machine. Additionally, more intelligent control is needed in the exoskeleton. The approach taken here is twofold. First, a motion amplification controller increases the finger movements of the wearer. Second, the amplified motion is processed using machine learning algorithms to predict what type of grasp the user is attempting. The controller would then be able to fuse the two, the amplification and prediction, to control the glove naturalistically.

The soft robotics field is developing rapidly and is poised to have a wide impact in a variety of applications. Soft robots have intrinsic compliance, offering a number of benefits as compared to traditional rigid robots. Compliance can provide compatibility with biological systems such as the human body and can provide some benefits for human safety and control. Further research into soft robots can be advanced by further development of pneumatic actuators. Pneumatic actuators are a good fit for exoskeleton robots because of their light weight, small size, and flexible materials. This is because a wearable robot should be human friendly, therefore, it should be light weight, slim, powerful, and simple. In this paper, a novel pneumatic artificial muscle using soft materials including integrated electronics for wearable exoskeletons is proposed. We describe the design, fabrication, and evaluation of the actuator, as well as the manufacturing process used to create it. Compared to traditional pneumatic muscle actuators such as the McKibben actuator and new soft actuators that were recently proposed, the novel actuator overcomes shortcomings of prior work. This is due to the actuator's very high contraction ratio that can be controlled by the manufacturing process. In this paper, we describe the design, fabrication, and evaluation of a novel pneumatic actuator that can accommodate integrated electronics for displacement and pressure measurements used for data analysis and control. The desired performance characteristics for the actuator were 100 ~ 400N at between 35kPa and 105kPa, and upon testing we found almost 120 ~ 300N which confirms that these actuators may be suitable in soft exoskeleton applications with power requirements comparable to rigid exoskeletons. Furthermore, a novel soft pneumatic elbow exoskeleton based on the pneumatic actuator concept and manufacturing process is presented. Each structure is designed and manufactured with all fabric. The distally-worn structure is only 300g, which is light weight for an arm exoskeleton, and the design is simple, leading to a low materials cost. / Master of Science / The soft robotics field is developing rapidly and is poised to have a wide impact in a variety of applications. The soft robotics is the specific field of robotics, which deals with flexible materials and different geometry in contrast to general robots made by rigid materials. Therefore, soft robots have intrinsic compliance, offering a number of benefits as compared to traditional rigid robots. Compliance can provide compatibility with biological systems such as the human body and can provide some benefits for human safety and control. Further research into soft robots can be advanced by the further development of pneumatic actuators. Pneumatic actuators are a good fit for exoskeleton robots because of their light weight, small size, and flexible materials. This is because a wearable robot should be human-friendly, therefore, it should be light weight, slim, powerful, and simple. In this paper, a novel pneumatic artificial muscle using soft materials for wearable exoskeletons is proposed. We describe the design, fabrication, and evaluation of the actuator, as well as the manufacturing process used to create it and its electronic system for data analysis and control. Furthermore, a novel soft pneumatic elbow exoskeleton based on the pneumatic actuator concept and manufacturing process is presented. Each structure is designed and manufactured with all fabric. The distally-worn structure is only 300g, which is light weight for an arm exoskeleton, and the design is simple, leading to a low materials cost.

This thesis presents the development and implementation of a machine learning prediction model for concurrently aggregating interval linear step distance predictions before future foot placement. Specifically, on-body inertial measurement units consisting of accelerometers, gyroscopes, and magnetometers, through integrated development by Xsens, are used for measuring human walking behavior in real-time. The data collection process involves measuring activity from two subject participants who travel an intended course consisting of flat, stair, and sloped walking elements. This work discusses the formulation of the ensemble machine learning prediction algorithm, real-time application design considerations, feature extraction and selection, and experimental testing under which this system performed several different test case conditions. It was found that the system was able to predict the linear step distances for 47.2% of 1060 steps within 7.6cm accuracy, 67.5% of 1060 steps within 15.2cm accuracy, and 75.8% of 1060 steps within 23cm. For separated flat walking, it was found that 93% of the 1060 steps have less than 25% error, and 75% of the 1060 steps have less than 10% error which is an improvement over the commingled data set. Future applications and work to expand upon from this system are discussed for improving the results discovered from this work. / Master of Science / This thesis presents the development and implementation of a machine learning prediction model for determining the stepping distance of future steps in real-time walking before their placement occurs. Specialized sensor units for measuring human motion activity are worn on the body for collecting and characterizing human walking behavior in real-time. Two subject participants are asked to walk a planned course consisting of flat, stair, and sloped walking elements. This work discusses the prediction algorithm voting scheme, real-time application design considerations, descriptive data elements for the algorithm, and experimental testing under which this system performed several different test case conditions. Detailed experimental tests are concluded in order to fully understand the extent of the system’s performance and the behaviors it exhibits throughout. The approach explored in this work enables researchers and roboticists to develop improvements and construct variations which may become superior to this method.

No description available.

Work-related musculoskeletal disorders (WMSDs) due to overexertion and consequently the low back pain (LBP) are one of the most prevalent sources of nonfatal occupational injuries and illnesses in all over the world. In the past several years, the industrial exoskeletons especially the passive ones have been proposed as alternative intervention and assistive devices, which are capable of reducing the risk of WMSDs and LBP. However, more research is warranted to validate the applicability of these exoskeletons. In addition, because the majority of previous studies have been limited to specific lifting tasks using only one type of lift assistive exoskeleton, more research is needed to examine the effect of alteration of different lift-assistive exoskeletons on reducing the activity of back muscles and metabolic reduction. The main objective of this dissertation is to render an overview of three studies that attempt to improve the literature by providing comprehensive biomechanical evaluations and metabolic assessments of three passive lift-assistive exoskeletons (VT-Lowe's Exoskeleton (developed in ARLab at VT), Laevo and SuitX). This dissertation has been composed of three related studies. The first study aimed to investigate and examine the capability of a novel lift assistive exoskeleton, VT-Lowe's exoskeleton, in reducing the peak and mean activity of back and leg muscles. Findings revealed that the exoskeleton significantly decreased the peak and mean activity of back muscles (IL(iliocostalis lumborum) and LT(longissimus thoracis)) by 31.5% and 29.3% respectively for symmetric lifts, and by 28.2% and 29.5% respectively for asymmetric lifts. Furthermore, the peak and mean EMG of leg muscles were significantly reduced by 19.1% and 14.1% during symmetric lifts, and 17.4% and 14.6% during asymmetric lifts. Interestingly, the VT-Lowe's exoskeleton showed higher reduction in activity of back and leg muscles compared to other passive lift-assistive exoskeletons available in the literatures. In the second study, the metabolic cost reduction associated with the use of VT-Lowe's exoskeleton during freestyle lifting was theoretically modelled, validated and corresponding metabolic savings were reported. The metabolic cost and the oxygen consumption results supported the hypothesis that the VT-Lowe's exoskeleton could significantly reduce the metabolic demands (~7.9% on average) and oxygen uptake (~8.7% on average) during freestyle lifting. Additionally, we presented a prediction model for the metabolic cost of exoskeleton during repetitive freestyle lifting tasks. The prediction models were very accurate as the absolute prediction errors were small for both 0% (< 1.4%) and 20% (< 0.7%) of body weight. In the third study, the biomechanical evaluation, energy expenditure and subjective assessments of two passive back-support exoskeletons (Laevo and SuitX) were examined in the context of repetitive lifting tasks. The experimental lifting tasks in this study were simulated in a laboratory environment for two different levels of lifting symmetry (symmetric vs. asymmetric) and lifting posture (standing vs. kneeling). Results of this study demonstrated that using both exoskeletons during dynamic lifting tasks could significantly lower the peak activity of trunk extensor muscles by ~10-28%. In addition, using both exoskeletons could save the energy expenditure by ~4-13% in all conditions tested by partially offsetting the weight of the torso. Such reductions were, though, task-dependent and differed between the two tested exoskeletons. Overall, the results of all three studies in this dissertation showed the capability of passive lift-assistive exoskeletons in reducing the activity of back and leg muscles and providing metabolic savings during repetitive lifting tasks. / Doctor of Philosophy / Low back pain (LBP) due to overexertion is known as one of the most important sources of nonfatal occupational injuries especially for the workers or manual material handlers who are involved in frequent or repetitive lifting tasks. Every year, many workers are temporarily or permanently disabled due to overuse injuries at workplace. In the past several years, industrial exoskeletons have gained growing interest among biomechanist, roboticist, and other human factor researchers as potential assistive devices to reduce the risk of LBP. In general, the industrial exoskeletons are either “passive or “active”; Active exoskeletons are powered by mechanical/electrical motors and actuators, however, the passive exoskeletons often work using cheaper devices such as gas or metal springs, elastic elements, etc. The exoskeletons discussed in this dissertation are categorized as passive rigid lower-back exoskeletons and they function by storing energy in a spring when the wearer bends and returning the stored energy when the wearer lifts. This dissertation consists of three studies that attempt to provide comprehensive biomechanical evaluations and metabolic assessments of three passive lift-assistive exoskeletons (i.e., VT-Lowe’s Exoskeleton, Laevo and SuitX). The first study examined the efficacy of a novel lift-assistive exoskeleton, VT-Lowe’s exoskeleton, in reducing the peak and mean activity of back and leg muscles. The results of this study demonstrated that the exoskeleton reduced the peak and mean activity of back and leg muscles for symmetric and asymmetric lifting tasks. VT-Lowe’s exoskeleton also showed higher reduction in activity of back muscles compared to other passive lift-assistive exoskeletons available in the literature. In the second study, the metabolic cost reduction with VT-Lowe’s exoskeleton was theoretically modeled and the modeling outcomes were compared to metabolic costs measurements when the exoskeleton was worn. The experimental findings of this study supported the applicability of the exoskeleton by significantly reducing the metabolic cost and oxygen uptake during the freestyle repetitive lifting tasks. Moreover, the prediction metabolic cost model of the exoskeleton showed high accuracy as the absolute prediction errors were within 1.5%. In the third study, the biomechanical evaluation, energy expenditure and subjective assessments of two passive back-support exoskeletons (Laevo and SuitX) were examined in repetitive lifting tasks. The lifting tasks of this study were simulated in a laboratory environment for two different levels of lifting symmetry (symmetric vs. asymmetric) and lifting posture (standing vs. kneeling). Findings of this study showed that both exoskeleton significantly lowered the peak activity of back muscles during the dynamic lifting tasks. Moreover, using both exoskeletons provided metabolic cost savings in all of the studies conditions. Overall, results obtained from the three studies in this dissertation verified the capability of these passive lift- vi assistive exoskeleton in reducing the activity of back and leg muscles and providing the metabolic savings during repetitive lifting tasks.

Wearable assistive robotics have the potential to address an unmet medical need of reducing disability in individuals with chronic hand impairments due to neurological trauma. Despite myriad prior works, few patients have seen the benefits of such devices. Following application experience with tendon-actuated soft robotic gloves and a collaborator's orthosis with novel flat-spring actuators, we identified two common assumptions regarding hand orthosis design. The first was reliance on incomplete studies of grasping forces during activities of daily living as a basis for design criteria, leading to poor optimization. The second was a neglect of increases in muscle tone following neurological trauma, rendering most devices non-applicable to a large subset of the population. To address these gaps, we measured joint torques during activities of daily living with able-bodied subjects using dexterity representative of orthosis-aided motion. Next, we measured assistive torques needed to extend the fingers of individuals with increased flexor tone following TBI. Finally, we applied this knowledge to design a cable actuated orthosis for assisting finger extension, providing a basis for future work focused on an under-represented subgroup of patients.

The human spine is an integral part of the human body. Its functions include mobilizing the torso, controlling postural stability, and transferring loads from upper body to lower body, all of which are essential for the activities of daily living. However, the many complex tasks of the spine leave it vulnerable to damage from a variety of sources. Prolonged walking with a heavy backpack can cause spinal injuries. Spinal diseases, such as scoliosis, can make the spine abnormally deform. Neurological disorders, such as cerebral palsy, can lead to a loss of torso control. External torso support has been used in these cases to mitigate the risk of spinal injuries, to halt the progression of spinal deformities, and to support the torso. However, current torso support designs are limited by rigid, passive, and non-sensorized structures. These limitations were the motivations for this work in developing the science for design of torso exoskeletons that can improve the effectiveness of current external torso support solutions. Central features to the design of these exoskeletons were the abilities to sense and actively control the motion of or the forces applied to the torso. Two applications of external torso support are the main focus in this study, backpack load carriage and correction of spine deformities. The goal was to develop torso exoskeletons for these two applications, evaluate their effectiveness, and exploit novel assistive and/or treatment paradigms. With regard to backpack load carriage, current torso support solutions are limited and do not provide any means to measure and/or adjust the load distribution between the shoulders and the pelvis, or to reduce dynamic loads induced by walking. Because of these limitations, determining the effects of modulating these loads between the shoulders and the pelvis has not been possible. Hence, the first scientific question that this work aims to address is What are the biomechanical and physiological effects of distributing the load and reducing the dynamic load of a backpack on human body during backpack load carriage? Concerning the correction of spinal deformities, the most common treatment is the use of a spine brace. This method has been shown to effectively slow down the progression of spinal deformity. However , a limitation in the effectiveness of this treatment is the lack of knowledge of the stiffness characteristics of the human torso. Previously, there has been no means to measure the stiffness of human torso. An improved understanding of this subject would directly affect treatment outcomes by better informing the appropriate external forces (or displacements) to apply in order to achieve the desired correction of the spine. Hence, the second scientific question that this work aims to address is How can we characterize three dimensional stiffness of the human torso for quantifiable assessment and targeted treatment of spinal deformities? In this work, a torso exoskeleton called the Wearable upper Body Suit (WEBS) was developed to address the first question. The WEBS distributes the backpack load between the shoulders and the pelvis, senses the vertical motion of the pelvis, and provides gait synchronized compensatory forces to reduce dynamic loads of a backpack during walking. It was hypothesized that during typical backpack load carriage, load distribution and dynamic load compensation reduce gait and postural adaptations, the user’s overall effort and metabolic cost. This hypothesis was supported by biomechanical and physiological measurements taken from twelve healthy male subjects while they walked on a treadmill with a 25 percent body weight backpack. In terms of load distribution and dynamic load compensation, the results showed reductions in gait and postural adaptations, muscle activity, vertical and braking ground reaction forces, and metabolic cost. Based on these results, it was concluded that the wearable upper body suit can potentially reduce the risk of musculoskeletal injuries and muscle fatigue associated with carrying heavy backpack loads, as well as reducing the metabolic cost of loaded walking. To address the second question, the Robotic Spine Exoskeleton (ROSE) was developed. The ROSE consists of two parallel robot platforms connected in series that can adjust to fit snugly at different levels of the human torso and dynamically modulate either the posture of the torso or the forces exerted on the torso. An experimental evaluation of the ROSE was performed with ten healthy male subjects that validated its efficacy in controlling three dimensional corrective forces exerted on the torso while providing flexibility for a wide range of torso motions. The feasibility of characterizing the three dimensional stiffness of the human torso was also validated using the ROSE. Based on these results, it was concluded that the ROSE may alleviate some of the limitations in current brace technology and treatment methods for spine deformities, and offer a means to explore new treatment approaches to potentially improve the therapeutic outcomes of the brace treatment.

Manufacturing and construction workers undertake physically strenuous activities increasing the risk of health problems, disability, and sick leave, leading to lower job attractiveness and job candidate scarcity. In the EU, up to 44 million workers are affected by workplace-related musculoskeletal disorders (MSDs), representing a total annual cost of more than €240 billion. Exoskeleton use could alleviate muscle peak loads and reduce the risks of injury of workers. This work is related to the INTERREG's project "EXSCALLERATE" which aimed to accelerate the adoption of exoskeletons among SMEs. This research presents a comparative study of using exoskeletons by workers while performing different tasks related to their job. The tests evaluate the advantages of using exoskeletons in reducing human muscle activity, thereby, reducing the fatigue and tiredness. The study uses two commercially available exoskeletons, (1) upper body exoskeleton known as Eksovest and (2) lower body exoskeleton known as LegX. For upper body, the study performed drilling tasks at shoulder height and roof drilling positions, whereas, for the lower body, virtual chair position and squatting positions are tested which involved frequent bending of knees. Besides, the experiments based on accuracies of the data collection techniques and compare three volunteer’s body muscle data acquired by EMG sensor. From these comparisons, it is found that the muscle activity can be reduced up to 60% by using these exoskeletons, hence, increasing the work life of the workforce. The results of this study will help create awareness among SMEs towards the adoption of exoskeletons.

As a large generation ages, the collective financial and ethical responsibility to prevent egregious bodily harm through fall prevention and gait assistant exoskeleton devices increases. Risk for falls increases with age and the severity of the fall does as well. To support this elderly population, motorized exoskeletons can both increase stability as well as respond faster to fall scenarios, but current models do not more around the existing biological framework. Giving participants a range of motion in key pelvic areas can closely approximate synchronous rotation around the femoral head, while limiting an increase in their sagittal profile. Utilizing 3D printed components while incorporating existing orthic methods provide short production times on modular designs. Although primarily mechanically based, these designs consider electronic requirements and are capable for supporting movement for a 200 lbs. user at a brisk walking pace for 1 hour.