Arvid Q. L. Keemink

Mechanical design of a manipulation system for unmanned aerial vehicles

In this paper, we present the mechanical design and modeling of a manipulation system for unmanned aerial vehicles, which have to physically interact with environments and perform ultrasonic non-destructive testing experiments and other versatile tasks at unreachable locations for humans. The innovation of the prototype lies in the use of a three degrees of freedom Delta robotic manipulator together with a nondestructive testing end-effector, realized by a Cardan gimbal that allows the ultrasonic sensor to compliantly interact with the remote environment. The Cardan gimbal is endowed with a small actuator for the roll motion of the end-effector, a compliant element in the direction of interaction and two passive rotational degrees of freedom with defined equilibria to overcome gravity and to define a stable zero reference. Simulation results of a ducted-fan unmanned aerial vehicle interacting with a wall validate the overall mechanical design.

Developing an aerial manipulator prototype: physical interaction with the environment

This article focuses on the design, modeling, and control of an aerial manipulator prototype, i.e., an innovative configuration consisting of a miniature quadrotor helicopter endowed with a robotic manipulator. The overall system is designed to accomplish operations that require physical interaction with the surrounding environment while remaining airborne. To investigate the dynamical model of the aerial manipulator, a simple planar benchmark is used to analyze the interactions between the quadrotor, the robotic manipulator, and the environment. A control strategy for the planar system is designed to guarantee robustness in the presence or absence of contacts. Experiments on a real setup validate the control in the two different scenarios in which the aerial manipulator is either freely flying or physically interacting with the environment.

Evaluation of EMG, force and joystick as control interfaces for active arm supports

BackgroundThe performance capabilities and limitations of control interfaces for the operation of active movement-assistive devices remain unclear. Selecting an optimal interface for an application requires a thorough understanding of the performance of multiple control interfaces.MethodsIn this study the performance of EMG-, force- and joystick-based control interfaces were assessed in healthy volunteers with a screen-based one-dimensional position-tracking task. The participants had to track a target that was moving according to a multisine signal with a bandwidth of 3 Hz. The velocity of the cursor was proportional to the interface signal. The performance of the control interfaces were evaluated in terms of tracking error, gain margin crossover frequency, information transmission rate and effort.ResultsNone of the evaluated interfaces was superior in all four performance descriptors. The EMG-based interface was superior in tracking error and gain margin crossover frequency compared to the force- and the joystick-based interfaces. The force-based interface provided higher information transmission rate and lower effort than the EMG-based interface. The joystick-based interface did not present any significant difference with the force-based interface for any of the four performance descriptors. We found that significant differences in terms of tracking error and information transmission rate were present beyond 0.9 and 1.4 Hz respectively.ConclusionsDespite the fact that the EMG-based interface is far from the natural way of interacting with the environment, while the force-based interface is closer, the EMG-based interface presented very similar and for some descriptors even a better performance than the force-based interface for frequencies below 1.4 Hz. The classical joystick presented a similar performance to the force-based interface and holds the advantage of being a well established interface for the control of many assistive devices. From these findings we concluded that all the control interfaces considered in this study can be regarded as a candidate interface for the control of an active arm support.

Design and control of an experimental active elbow support for adult Duchenne Muscular Dystrophy patients

Currently, a considerable group of adult Duchenne Muscular Dystrophy patients lives with severe physical impairments and strong dependency on care. Active arm supports can improve their quality of life by augmenting their arms residual motor capabilities. This paper presents the design and control of an experimental active elbow support specially made to investigate different control interfaces with adult DMD patients. The system can be controlled either with EMG or force signals which are used as inputs for an admittance-based controller. A preliminary test with a 22-year-old DMD patient with no arm function left, shows that the system is capable of successfully supporting the elbow flexion-extension movements using the low-amplitude EMG and force signals that still remained measurable.

Adaptive gravity and joint stiffness compensation methods for force-controlled arm supports

People with muscular weakness can benefit from arm supports that compensate the weight of their arms. Due to the disuse of the arms, passive joint stiffness increases and providing only gravity compensation becomes insufficient to support the arm function. Hence, joint stiffness compensation is also required, for which the use of active arm supports is essential. Force-based control interfaces are a solution for the operation of arm supports. A critical aspect of force-based interfaces, to properly detect the movement intention of the user, is the ability to distinguish the voluntary forces from any other force, such as gravity or joint stiffness forces. Model- and calibration-based strategies for the estimation of gravity and joint stiffness forces lack adaptability and are time consuming since they are measurement dependent. We propose two simple, effective and adaptive methods for the compensation of forces resulting from gravity and joint stiffness. The compensation methods are based on the estimation of the compensation force using a low-pass filter, and switching of control parameters using a finite state machine. The compensation methods were evaluated with an adult man suffering from Duchenne muscular dystrophy with very limited arm function. The results show that when gravity and joint stiffness forces were adaptively compensated the reachable workspace of the user was increased more than 50% compared to the workspace reached when only constant gravity compensation was provided.

Switching proportional EMG control of a 3D endpoint arm support for people with duchenne muscular dystrophy

Duchenne muscular dystrophy (DMD) is a disease resulting in progressive muscle degeneration. Active arm supports can improve the quality of life for people with DMD by augmenting the residual motor capabilities of their arm. As an extension of our previous study, this research aims at developing a EMG-based control interface to detect the users movement intention required to control more than 1-DOF. The interface switches between two horizontal and one vertical translations. Translations are proportionally controlled by EMG. The passive interaction torques measured between the arm and the active arm support, are used to make the robots endpoint resemble a gimbal mechanism. Hence decreasing the endpoints DOF from six to three by actively reducing the impedance of the rotational DOF. A preliminary evaluation of the control method has been carried out with one healthy subject, within a series of 2-D horizontal tracing and 3-D horizontal-vertical reaching tasks. A pilot study was also conducted with a boy with DMD controlling the device in a 2-D horizontal tracing task. Performance was evaluated in terms of path efficiency, smoothness, task completion rate and time. The results indicate that the control method is able to successfully detect the intention of the user and translate it into the intended movement. Furthermore, the reduction of the endpoints DOF, results in a simple yet functional controller able to support natural movements of the arm.

Design and control of the Active A-Gear: A wearable 5 DOF arm exoskeleton for adults with Duchenne muscular dystrophy

Adults with Duchenne muscular dystrophy (DMD) can benefit from active arm supports that augment the residual motion capabilities of their arms. While there is a considerable number of active exoskeletons commercially available for rehabilitation purposes, no active exoskeletons for supporting the arm function during the performance of activities of daily living (ADL) are available. In this paper, we present the Active A-gear, a wearable five degree-of freedom exoskeleton that has been specially developed to assist people with DMD for the performance of ADL. The Active A-Gear is built upon our previous design of the passive A-Gear, thus combining active and passive support. The Active A-Gear can be controlled using force-based (admittance) control. We report the results of the system verification, which included endpoint position bandwidth, workspace, weight, joint speed, endpoint force and joint backlash measurements. A pilot evaluation was carried out with one healthy subject (27 years-old) performing ADL. The results of the verification and evaluation tests indicate that with some modifications the Active A-Gear is a suitable platform to test the functional performance and feasibility of a wearable and active support for adults with DMD.

Haptic Human-Human Interaction Through a Compliant Connection Does Not Improve Motor Learning in a Force Field

Humans have a natural ability to haptically interact with other humans, for instance while physically assisting a child to learn how to ride a bicycle. A recent study has shown that haptic human-human interaction can improve individual motor performance and motor learning rate while learning to track a continuously moving target with a visuomotor rotation. In this work we investigated whether these benefits of haptic interaction on motor learning generalize to a task in which the interacting partners track a target while they learn novel dynamics, represented by a force field. Pairs performed the tracking task and were intermittently connected to each other through a virtual spring. Motor learning was assessed by comparing each partner’s individual performance during trials in which they were not connected to the performance of participants who learned the task alone. We found that haptic interaction through a compliant spring does not lead to improved individual motor performance or an increase in motor learning rate. Performance during interaction was significantly better than when the partners were not interacting, even when connected to a worse partner.

Admittance control for physical human–robot interaction:

This paper presents an overview of admittance control as a method of physical interaction control between machines and humans. We present an admittance controller framework and elaborate control scheme that can be used for controller design and development. Within this framework, we analyze the influence of feed-forward control, post-sensor inertia compensation, force signal filtering, additional phase lead on the motion reference, internal robot flexibility, which also relates to series elastic control, motion loop bandwidth, and the addition of virtual damping on the stability, passivity, and performance of minimal inertia rendering admittance control. We present seven design guidelines for achieving high-performance admittance controlled devices that can render low inertia, while aspiring coupled stability and proper disturbance rejection.

Haptic Physical Human Assistance

This dissertation covers three aspects of upper-extremity exoskeleton design: 1) Kinematics & motion: How to support the full range of motion of the human shoulder? We present a 2D visualization method that can show coupling between the range of motion (ROM) of rotations of the glenohumeral joint. This visualization helps in communication, comparison, design and analysis of human and assistive device ROM. We furthermore provide a conceptual design and differential inverse kinematics method for a redundant 4 degree of freedom (DOF) shoulder-exoskeleton. The extra DOF allows for movement redundancy to steer away from body collisions and kinematic singularity. 2) Haptics & Control: How to get devices such robots or exoskeletons to behave as some defined impedance in a stable manner when interacting with human users; how to implement stable admittance control with inertia reduction? We analyze the energetic behavior of the control method ‘admittance control’. During admittance control an interaction force with a human user is measured, which is used in a dynamical mechanical model that prescribes a motion for the exoskeleton to follow. Such a method is inherently active (i.e. it generates energy that can result in coupled instability) when it is used to reduce the apparent inertia of the exoskeleton. We provide insight into why this energetically active behavior occurs, and provide guidelines to design a controller that is (close to) passive and is therefore (almost) always stable when in contact with a human limb. 3) Human Factors: How do humans respond to dissipative shared control forces? Passive and active exoskeletons can apply forces to the human user to steer or help the person and share control authority. A passive force that only dissipates energy is a damping force. We investigate how position dependent damping forces around reaching targets influence human reaching time and kinematics. Results show that humans increase their accelerations and decrease their reaching time when assisted in this manner. We pose the hypothesis that damping forces attenuate neural activation dependent motor noise. Without the damping, this higher noise for higher accelerations would have had too much of a negative effect on the required task accuracy.