<p> The purpose of this thesis is to ascertain the feasibility of using strain gauges attached to a Unmanned Aerial Vehicle (UAV) gripper to determine, upon impact, the hardness of a landing site. We design and fabricate a four finger gripper that uses a rotary component to convert the rotational motion of a servo to the linear motion of the finger assemblies. We functionally test a gripper prototype made from rapid-prototype material. We conduct three experiments to test the gripper's functionality. The first experiment tests the gripper's ability to grasp, lift, and release a centered payload, and the gripper performed with overall success rates of 91%, 100%, and 87% respectively. The second experiment tests the gripper's ability to self-align, lift and release the payload and the gripper performed with overall success rates of 99%, 100%, and 96% respectively. The third experiment tests the functional durability of the gripper, and it performed without error for 5000 open/close cycles. </p>

Human standing posture depends on sensory inputs from vision, from touch and from the vestibular system. Humans use these inputs to determine position in the environment and to compensate for external forces resulting in constant swaying. This thesis investigated the creation of a four-degree-of-freedom robotic model of human upright stance using only hip actuation. In this model, the ankles are free so that there is no reaction torque from the ground and balance is provided by torques at the hip joint, resulting in an underactuated system. The mathematical model of the system was first derived using the Euler-Lagrange formulation. Using this model, simulations on the system were performed with different inputs to control the actuators. The robot was then designed and built. This was followed by tests done on the robot to examine its utility as a robotic model of human stance. From these tests, a new design with possible improvements was then proposed.

No description available.

<p> Integration of Unmanned Aircraft Systems (UAS) in the National Airspace System will require UAS to meet the standards expected of conventional manned aircraft, including interactions with air traffic controllers (ATCo). To study the effect UAS delays have on ATCo acceptance of UAS operations eight ATCos managed traffic scenarios with conventional manned aircraft and one UAS. To mimic the potential latencies of UAS operations 1.5 or 5 second delays were added to the UAS pilot's verbal and execution initiation responses. Delays were either constant or variable within each scenario. While ATCos were tolerant of UAS delays, the duration and consistency of verbal and execution delays did affect communications and ATCos' experiences managing traffic. Limitations and recommendations for future research are discussed.</p>

<p> This work demonstrates the first low cost, all-elastomer capacitive tactile arrays compatible with roll-to-roll manufacturing. A new manufacturing process has been developed in which elastomer sheets are covered with a stencil, spray coated with conductive elastomer on one or both sides, and stacked to create the sensor array. These arrays are highly flexible and can withstand large strains. Sensor costs are below $0.12/sensor in small quantities. Some variants in the fabrication process result in a slightly curved sensor so the change in capacitance is highly nonlinear at low pressures, but approaches theoretical sensitivities at higher pressures. The sensors have been determined to be highly sensitive, with a sensor resolution of 0.5 Pa and reveal a repeatable response from 1 kPa up to 120 kPa. A variety of materials ranging in modulus, thickness and texture were investigated for static, dynamic, and spatial location testing.</p>

<p> Myoelectric control systems (MECs) remain the technological bottleneck in the development of advanced prosthetic hands. MECs should provide a human machine interface that deciphers user intent in real-time and operates effectively in daily life. Current MECs like finite state machines and pattern recognition systems require physiologically inappropriate commands to indicate intent and/or lack effectiveness in a clinical setting. The work of this dissertation aims to develop and validate a novel MEC architecture, namely postural control, in order to supplant the current state of the art MECs and recreate more of the characteristics of the intact limb. Specifically, the development of the postural control systems builds upon previous work based on principal component analysis of human grasping. Novel attributes of the postural control system were then added to the MEC, empirically tested, and validated with able limbed subjects using a virtual hand interface. Further investigation of the postural controller was performed by comparing it to state of the art commercial and research MECs with able limbed subjects using a physical prosthesis during activities of daily living. The dissertation concludes by verifying the increased effectiveness and robustness of the postural controller compared to other MECs when used by persons with transradial limb loss to perform activities of daily living with a physical prosthesis.</p>

<p> Automation has been known to provide both costs and benefits to experienced humans engaged in a wide variety of operational endeavors. Its influence on skill acquisition for novice trainees, however, is poorly understood. Some previous research has identified impoverished learning as a potential cost of employing automation in training. One prospective mechanism for any such deficits can be identified from related literature that highlights automation's role in reducing cognitive workload in the form of perceived task difficulty and mental effort. However three experiments using a combination of subjective self-report and EEG based neurophysiological instruments to measure mental workload failed to find any evidence that link the presence of automation to workload or to performance deficits resulting from its previous use. Rather the results in this study implicate engagement as an underlying basis for the inadequate mental models associated with automation-induced training deficits. The conclusion from examining these various states of cognition is that automation-induced training deficits observed in novice unmanned systems operators are primarily associated with distraction and disengagement effects, not an undesirable reduction in difficulty as previous research might suggest. These findings are consistent with automation's potential to push humans too far "out of the loop" in training. The implications of these findings are discussed.</p>

<p> A software prototype for autonomous 3D scanning of uncooperatively rotating orbital debris using a point cloud sensor is designed and tested. The software successfully generated 3D models under conditions that simulate some on-orbit orbit challenges including relative motion between observer and target, inconsistent target visibility and a target with more than one plane of symmetry. The model scanning software performed well against an irregular object with one plane of symmetry but was weak against objects with 2 planes of symmetry. </p><p> The suitability of point cloud sensors and algorithms for space is examined. Terrestrial Graph SLAM is adapted for an uncooperatively rotating orbital debris scanning scenario. A joint EKF attitude estimate and shape similiarity loop closure heuristic for orbital debris is derived and experimentally tested. The binary Extended Fast Point Feature Histogram (EFPFH) is defined and analyzed as a binary quantization of the floating point EFPFH. Both the binary and floating point EPFH are experimentally tested and compared as part of the joint loop closure heuristic.</p>

<p> We present the background and motivation for ground vehicle autonomy, and focus on uses for space-exploration. Using a simple design example of an autonomous ground vehicle we derive the equations of motion. After providing the mathematical background for nonlinear systems and control we present two common methods for exactly linearizing nonlinear systems, feedback linearization and backstepping. We use these in combination with three adaptive control methods: model reference adaptive control, adaptive sliding mode control, and extremum-seeking model reference adaptive control. We show the performances of each combination through several simulation results. We then consider disturbances in the system, and design nonlinear disturbance observers for both single-input-single-output and multi-input-multi-output systems. Finally, we show the performance of these observers with simulation results.</p>

<p> The human knee joint exhibits a spring-type behavior during the stance phase of walking at the preferred speed, which is both subject-specific and gait-specific. This observation led us to hypothesize that the human knee joint could partially adapt to an externally-applied tuned mechanical stiffness during the stance phase leading to reduced muscle involvement and energy expenditure. We also hypothesized that a spring, which is tuned to the body size and gait speed, in parallel with an impaired knee joint during the stance phase can partially restore the natural spring-type behavior of the knee joint. Three experimental and theoretical steps were taken to test these hypotheses.</p><p> First, a series of statistical models were developed that can closely characterize the moment-angle behavior of the knee joint using a set of measurable parameters including body weight and height, gait speed, and joint excursion. It is explained that these models can be used to tune the components of knee exoskeletons/orthoses and prostheses to the body size and gait speed of users, as well as general applications in understanding gait biomechanics. The statistical models of the knee joint were used in the next steps of this research to tune the stiffness of the experimental exoskeletal devices throughout the experimental sessions.</p><p> To experimentally test the first main hypothesis, a pair of quasi-passive knee exoskeletons was developed. When worn on a healthy subject, each exoskeleton implements an interchangeable spring in parallel with the knee joint during the stance and allows free rotation during the swing phase. The exoskeletons with a range of stiffness were used in a series of experiments on healthy individuals to study the mechanics and energetics of human gait in interaction with exoskeletal impedances in parallel with the knee joint. Healthy lower extremity joints showed substantial adaptation to the exoskeleton stiffness/assistance suggesting that replicating the natural behavior of a joint could be a viable method for the design of lower extremity exoskeletons to reduce muscle involvement and energy expenditure. It was also observed that a healthy knee joint can fully accommodate external assistance only to a certain level, above which the knee joint adaptation saturates and biarticular effects emerge.</p><p> To test the second hypothesis, a compliant stance control orthosis was developed that implements a spring in parallel with an impaired knee joint during the stance and allows free rotation during the swing phase. It was found that a compliant stance control orthosis can restore the natural spring-type behavior of an impaired knee joint during the stance phase. The compliant stance control orthosis showed higher gait speed and more natural kinematic patterns when compared with the state-of-the-art stance control orthoses that rigidly lock the knee during the stance phase. The findings of this research also showed that a friction-based latching mechanism can be a viable option in the design of lower extremity assistive devices that require engagement and disengagement of passive components.</p>