While modeling multi-contact manipulation as a quasi-static mechanical process transitioning between different contact equilibria, we propose formulating it as a planning and optimization problem, explicitly evaluating (i) contact stability and (ii) robustness to sensor noise. Specifically, we conduct a comprehensive study on multi-manipulator control strategies, focusing on dual-arm execution in a planar peg-in-hole task and extending it to the Multi-Manipulator Multiple Peg-in-Hole (MMPiH) problem to explore increased task complexity. Our framework employs Dynamic Movement Primitives (DMPs) to parameterize desired trajectories and Black-Box Optimization (BBO) with a comprehensive cost function incorporating friction cone constraints, squeeze forces, and stability considerations. By integrating parallel scenario training, we enhance the robustness of the learned policies. To evaluate the friction cone cost in experiments, we test the optimal trajectories computed for various contact surfaces, i.e., with different coefficients of friction. The stability cost is analytical explained and tested its necessity in simulation. The robustness performance is quantified through variations of hole pose and chamfer size in simulation and experiment. Results demonstrate that our approach achieves consistently high success rates in both the single peg-in-hole and multiple peg-in-hole tasks, confirming its effectiveness and generalizability. The video can be found at https://youtu.be/IU0pdnSd4tE.

In this work, we present a simplified quasi-static model of a two-point contact gripper turning a circular valve to predict the reaction torques produced at the base of the valve as a function of the axis misalignment. Specifically, we learned that geometric features such as (i) the misalignment vector being tangent to the reaction torques, (ii) length of the misalignment vector being directly proportional to the magnitude of the reaction torques and (iii) small axial misalignments resulting in well defined ‘double-loops’ in the 2D reaction torques space are indicative of axis misalignment.

In this paper we present a methodology for optimising the design of a metamaterial structure with one degree of freedom that is able to simultaneously bend and stretch. The structure is intended for assisting flexion-extension of the wrist joint. The metamaterial is comprised of serially connected, individually designed cells. The design parameters can be chosen to optimally fit a desired planar curve such as the curvature of the skin on a plane normal to the flexion/extension axis of the wrist joint. A tool for the optimised design is described and the experimental validation of the output is conducted, to show the ability of the mechanism to conform to a 2D curve, also exhibiting a change in length, which is desirable for reducing sliding and shear on the skin. The design tool allows the generation of metamaterials optimised for a multitude of other applications where actuated mechanisms must be worn by a user for rehabilitation or assistance purposes.

The domain of robotics is challenging to apply deep reinforcement learning due to the need for large amounts of data and for ensuring safety during learning. Curriculum learning has shown good performance in terms of sample-efficient deep learning. In this paper, we propose an algorithm (named GloCAL) that creates a curriculum for an agent to learn multiple discrete tasks, based on clustering tasks according to their evaluation scores. From the highest-performing cluster, a global task representative of the cluster is identified for learning a global policy that transfers to subsequently formed new clusters, while remaining tasks in the cluster are learnt as local policies. The efficacy and efficiency of our GloCAL algorithm are compared with other approaches in the domain of grasp learning for 49 objects with varied object complexity and grasp difficulty from the EGAD! dataset. The results show that GloCAL is able to learn to grasp 100% of the objects, whereas other approaches achieve at most 86% despite being given 1.5× longer training time.

The applicability of deep reinforcement learning algorithms to the domain of robotics is limited by the issue of sample inefficiency. As in most machine learning methods, more samples generally mean better learning effectiveness. Sample collection for robotics application is a time-consuming process in addition to safety issues for both the robot itself and the environment surrounding it that come into play for real-world scenarios. Because of these limitations, sample efficiency plays a very vital role in the field of robotic learning. To deal with this, curriculum learning offers a methodology that allows robots to suffer less from the sample collection burden required, trying to keep it at a minimum. This study aims to tackle the sample inefficiency that deep reinforcement learning algorithms face in the domain of robotics by designing a curriculum. We propose an algorithm which decides on the sequence of tasks that the agent must learn to enable the transfer of knowledge in a sample-efficient manner towards the target task. Our algorithm performs a parameter-space task representation for the purpose of deciding on the difficultiness of the tasks. Once the difficulty level of each is determined, easy tasks are learned first before the final target task. We perform a study on a double inverted pendulum setup. Simulation results showed that transfer of knowledge via curriculum is more sample efficient than a direct transfer.

In this work, we introduce a novel, end-to-end motion planner for quadrotor navigation. Informed by a rough path to goal in partially unknown environments, our method creates desirable motion plans using raw depth images from a front-facing camera. It exploits correlations between local spatial portions of these images to generate desirable motion primitive sequences on the fly without conducting explicit sensing-reconstructing-planning. We evaluate our method through an extensive comparison with three competitor algorithms over ten different environments in AirSim simulations. Our method outperforms its competitors in terms of safe navigation distance, navigation time, and crash rate over 50 flights. We also deploy our method for real flight tests with DJI F330 Quadrotor equipped with Intel RealSense D435, and demonstrate its real-time ap-plicability. Our method successfully performs 15 real flights in three different environment settings with increasing complexity. The experiments can be found at https://youtu.be/hw0sxNwliqs.

Current robotic rehabilitation devices have a high cost-to-benefit ratio, which prevents their large scale adoption by the clinical rehabilitation community. This paper first presents H-Man, a low cost planar robot, as a quantitative assessment and training tool. This is followed by a preliminary study to investigate baseline performance measures for motor assessment during reaching tasks as a step toward replacing conventional ordinal scales with continuous quantitative scales. Thirteen healthy and one participant with upper limb motor impairment participated in the study and performed reaching tasks with their dominant and non-dominant hands in three directions. The results from healthy subjects indicate no significant difference between different directions for both limbs and also between corresponding directions of dominant and non-dominant limbs (p > 0.05, all cases). However, differences in measures can be observed for the impaired subject.

The majority of industrial robots come with position/velocity control architectures which allow precise positioning and accurate path-following, but they fail when performing contact tasks for finishing stages. To address this issue, impedance and force control have been proposed as an alternative to the position/velocity control paradigm. However, while the former requires torque control and hence cannot be applied to most of the position/velocity-based industrial robots, the latter comes with issues such as limited bandwidth and instability. A more recent approach consists of mounting an active pneumatic tool on the robot end-effector with the aim of regulating the contact force independently from the robot control architecture. Nevertheless, such active tools are bulky and hence require robots with high payload. In this paper we present a prototype of a novel agonist-antagonist (A-A) cable-driven active tool that is light-weight and it has been conceived to control the impedance at the contact interface. In order to estimate and analyze the bandwidth, the A-A system will reproduce the behavior of elastic band which have a perfect bandwidth with a known stiffness. Preliminary experiments and results are presented in order to evaluate the goodness of the rendered impedance.

Robotic platforms developed to assist conventional motor therapy and to investigate human motor control have received an increasing interest over the past decades. However, for most of the proposed solutions, bulkiness and expensiveness have limited the use of such devices to specialized clinics that can afford their cost. This paper presents the H-Man, a two degree-of-freedom planar device designed according to three main principles: cost-effectiveness, portability and ease of control. The key component of the device is a planar H-shaped cable differential mechanism which ensures a constant Jacobian and homogeneous perceived inertia over the entire workspace. The paper presents the mechanical design as well as the performance evaluation in terms of perceived impedance.

This study presents the design of a novel minimalist liftoff-capable flapping-wing microaerial vehicle. Two wings are each directly driven by a geared pager motor by utilizing an elastic element for energy recovery, resulting in a maximum lift-to-weight ratio of 1.4 at 10 Hz for the 2.7 g system. Separate directly driven wings allow the system to both resonate and control individual wing flapping angle, reducing necessary power consumption, as well as allowing the production of roll and pitch body torques. With a series of varied prototypes, system performance is examined with change in wing offset from center of rotation and elastic element stiffness. Prototype liftoff is demonstrated with open loop driving a tethered prototype without guide wires. A dynamic model of the system is adapted and compared with the prototype experimental results for later use in prototype optimization.

In previous work, a proof-of-concept artificial flapper was devised by Campolo et al. to demonstrate that DC motors, in concert with compliant mechanisms, would be able to directly flap wings at relatively high frequencies and large angles without exceeding their operational limits. The prototype makes use of a pair of relatively long elastic strings as the compliant structures. In this article, we experimentally analyze the wing kinematics and efficiency of a more compact prototype, where small helical springs are implemented instead of elastic strings. Since the proof-of-concept prototype validated the quasi-sinusoidal assumption in spite of nonlinear aerodynamic damping, incorporating instantaneous wing kinematics into the analysis is not necessary, simplifying experiments and data processing. Along with wing kinematics and system efficiency, the possibility of controlling the wing pair to flap independently is evaluated as well.

This work investigates how to design a comfortable wrist exoskeleton which complies with the natural coordination mechanisms in the redundant wrist. Human sensorimotor control is known to impose intrinsic kinematic constraints to solve redundant motor tasks. To this end, the effect of an exoskeleton on natural motor strategies was assessed during pointing tasks performed with the wrist. The exoskeleton was designed based on the kinematic model of one specific subject. Then wrist orientation was observed during pointing tasks with the exoskeleton in the following conditions: i) optimal alignment between human and exoskeleton joints; ii) varying degrees of misalignment between human and exoskeleton joints; iii) optimal alignment while the PS axis was locked (i.e. no redundancy). The results exhibited a modification of the natural coordination mechanisms characterized by a subject-specific Koenderink shape index. Kruskal-Wallis pairwise analyses demonstrated a significant variation between test conditions indicating a change of intrinsic constraints with misalignment and locked PS. The assessment methodologies presented in this paper can be used to test for ergonomic constraints and can guide the design of robotic systems performing kinematically redundant tasks.

This work introduces a coordinate-independent method to analyse movement variability of tasks performed with hand-held tools, such as a pen or a surgical scalpel. We extend the classical uncontrolled manifold (UCM) approach by exploiting the geometry of rigid body motions, used to describe tool configurations. In particular, we analyse variability during a static pointing task with a hand-held tool, where subjects are asked to keep the tool tip in steady contact with another object. In this case the tool is redundant with respect to the task, as subjects control position/orientation of the tool, i.e. 6 degrees-of-freedom (dof), to maintain the tool tip position (3dof) steady. To test the new method, subjects performed a pointing task with and without arm support. The additional dof introduced in the unsupported condition, injecting more variability into the system, represented a resource to minimise variability in the task space via coordinated motion. The results show that all of the seven subjects channeled more variability along directions not directly affecting the task (UCM), consistent with previous literature but now shown in a coordinate-independent way. Variability in the unsupported condition was only slightly larger at the endpoint but much larger in the UCM.

This work focuses on anthropomorphic exoskeletons for the human wrist. We consider a 2 dof model for the human wrist with non intersecting joints and a similar model for the exoskeleton. We assume a viscoelastic attachment between the human hand and the handle of the exoskeleton which on one side allows the different kinematics of the exoskeleton to follow the human wrist and, on the other side, induces reaction forces at all joints, in particular causing discomfort. We quantify discomfort as the amount of potential energy stored in the deformation of the viscoelastic attachment. For a specific exoskeleton implementation, based on kinematic simulations, we report the kinematic mismatch (i.e. differences between the human joints and the corresponding exoskeleton joints) as well as the reaction forces arising when the human joints assume postures throughout their physiological range of motion. Considering a typical distribution of joint offset for humans (derived from literature) and the asymmetry in the discomfort function (derived from our simulations) we address the “one-size-fits-all” problem and propose an optimal joint offset for the exoskeleton, based on the minimization of the aggregate loss function.

This paper reports on an autonomous mobile robot embodying a simple instantiation of theDouble Hybrid Control Architecture, with a living animal in the control loop for improving robot navigational capabilities. In particular, a Carassius auratus (common Goldfish) receives a natural visual feedback from the environment, while its motor reactions, acquired by a digital camera and adequately processed, are mapped into motor commands, which are fed as inputs to the robot controller. Both the hardware and the software required for robot control is presented, as well as preliminary experimental data demonstrating the viability of the proposed approach.

This work proposes a mechatronic solution to increase the back-drivability of a state-of-the art robot for wrist neurorehabilitation. The final goal is to reduce robot mechanical impedance in order to cope with intrinsic kinematic constraints, which are adopted by the human brain to solve redundancy during pointing tasks with the wrist. The handle of the robot has been provided with a load cell and a direct force control scheme has been implemented to minimize the interaction forces/torques between the user and the robot. To this aim gravity, inertia and friction of the more proximal DOF of the robot (relative to Pronation/Supination (PS) movements) have been estimated and compensated for. The proposed solution resulted in a 70% reduction of the end-point perceived inertia in PS DOF as well as in a decrease of torques exerted by the user during both 1-DOF and 3-DOFs tasks. The average reduction of interaction torques is around 81% and 78% respectively. This work constitutes an important starting point for the analysis of the effect that different levels of robot transparency could have on the human neural constraints adopted during redundant tasks, such as pointing movements with the wrist.

In this work a multimodal head-mounted device for the assessment of social orienting behavior in children between 12 and 24 months is presented. The device is specifically designed to be used in poorly structured and uncontrolled environments such as day-care centers. Accordingly, a calibration procedure is described which fully exploits the multimodal approach and which is particularly suitable for an ecological assessment.

Ideally, robots used for motor rehabilitation, in particular, during assessment, should minimally perturb the voluntary movements of a subject. In this paper, we show how a state-of-the-art back-drivable robot, i.e., a robot that can be moved by the user with a low perceived mechanical impedance, when used for assessment can still perturb the voluntary movements of a subject. In particular, we show that, despite its low mechanical impedance, a robot may still not comply with the intrinsic kinematic constraints, which are of neural origin and are adopted by the human brain to solve redundancy in motor tasks. Specifically, the redundant task under consideration is the 2-D pointing task, which is performed by a subject with the sole use of the wrist [3 degree of freedom (DOF) kinematics]. Wrist orientations during pointing tasks are assessed in two different scenarios. In the first experiment, a lightweight handheld device is used, which introduces no loading effect. In the second experiment, similar pointing tasks are performed with the subject interacting with a state-of-the-art robot for wrist rehabilitation. In the first case, intrinsic kinematic constraints arise as 2-D surfaces embedded in the 3-D space of wrist configuration. Such surfaces are typically subject-dependent and reveal personal motor strategies. In the second case, a strong influence of the robot is remarked. In particular, 2-D surfaces still arise but are similar for all subjects and are referable to a mechanical origin (excessive loading by the robot). The assessment approach described in this paper, including both the experimental apparatus and data-analysis method, can be used as a test for the degree of back-drivability of mechanisms and robots in relation to constraints of neural origin, thus allowing the design of robots that can actually cope with such constraints. The clinical potential impact is also discussed.

In this work, the kinematics of the human wrist during pointing tasks is assessed and discussed, especially in relation to the use of wrist robots. First, the existence of intrinsic kinematics constraints (or Donders' law, similarly to oculomotor system) for the human wrist during pointing tasks is verified. To this end, a novel approach based on a hand-held device is presented which allow assessing wrist movements without any mechanical loading. Second, similar pointing tasks are assessed by means of a state-of-the-art wrist robot, typically used in robot-mediated therapy as well as assessment tool. By comparing experimental results relative to the pointing task performed by healthy subjects with the hand-held device and with the robot, a loading effect on the performance due to the mechanisms of the robot is remarked. A functional requirement for the next generation of rehabilitation robots is thus provided.

Tremor analysis in human or animal model plays a fundamental role for understanding the physiopathology of human disorders and to test new pharmacological treatments. Mechatronic systems for automatic and quantitative behavioural analysis are now of current use for neuroscientists to improve the results of their research. Most of these devices are portable and need to be used out of engineering labs by personnel with no technical expertise and without specific equipment. The calibration of the devices, that should be performed before each experimental session, is a typical issue to be faced. This paper deals with a new calibration method that is fast, simple and doesn't need other external measurement systems. It is based on the parallel use of an accelerometer and an optical sensor. The two signals are processed and compared to obtain the final calibration curve of the device.

In this paper, we study sensor fusion for the attitude estimation of micro aerial vehicles (MAVs), in particular mechanical flying insects. First, following a geometric approach, a dynamic observer is proposed which estimates attitude based on kinematic data available from different and redundant bio-inspired sensors such as halteres, ocelli, gravitometers, magnetic compass and light polarization compass. In particular, the traditional structure of complementary filters, suitable for multiple sensor fusion, is specialized to the Lie group of rigid body rotations SO(3). Then, a numerical implementation of the filter is provided for the specific case of inertial/magnetic navigation, i.e. when gravitometers, magnetometer and gyroscopes are available. Finally, the filter performance is experimentally tested via a 3 degrees-of-freedom robotic flapper and a custom-made set of inertial/magnetic sensors. Experimental results show good agreement, upon proper tuning of the filter, between the actual kinematics of the robotic flapper and the kinematics reconstructed from the inertial/magnetic sensors via the proposed filter.

There are recent evidence which show how brain development is strictly linked to the action. Movements shape and are, in turn, shaped by cortical and sub-cortical areas. In particular spontaneous movements of newborn infants matter for developing the capability of generating voluntary skill movements. Therefore studying spontaneous infants' movements can be useful to understand the main developmental milestones achieved by humans from birth onward. This work focuses on the design and development of a mechatronic wearable device for ecological movement analysis called WAMS (Wrist and Ankle Movement Sensor). The design and assembling of the device is presented, as well as the communication protocol and the synchronization with other marker-based optical movement analysis systems.

This paper presents the design of a novel sensor for slip detection. It consists of an easily fabricated miniaturized thermal probe that senses the additional convective heat transfer associated with the occurrence of mechanical slip. The fabrication procedures and the operating principle for the device are described in detail. A simple experimental setup was used to test the effectiveness of the proposed device. Tests were performed with varying velocities on four materials of differing thermal properties and surface roughnesses. The results show that slip can be effectively detected by the proposed sensor with a response times which can be as low as 6.3 ms. The performance of the device can be further improved when used in conjunction with a separate pressure sensor and by using more accurate methods of electrical resistance measurement.

Neuro-developmental engineering is a new interdisciplinary research area at the intersection of developmental neuroscience and bioengineering. Applications can be found in early detection of neuro-developmental disorders via a new generation of mechatronic toys for assessing the regular development of perceptual and motor skills in infants, in particular coordination of mobile and multiple frames of reference during manipulation. This paper focuses on the design of a novel mechatronic toy, shaped as a 5 cm (diameter) ball, i.e. small enough to be grasped with a single hand by a 1 year old child. The sensorized ball is designed to embed a kinematics sensing unit, able to sense both the orientation in 3D space and linear accelerations, as well as a force sensing unit, to detect grasping patterns during manipulation. Dimensioning of batteries able to operate for 1 hour during experimental sessions as well as a wireless communication unit are also included in the design.

Great amounts of raw data acquired with the use of an innovative mechatronic platform during an extensive clinical trial in a neuro-rehabilitation setting needs an analysis and interpretation. The platform records data from eight 6 DOF force-torque sensors during an isometric functional assessment of post-stroke patients. The identification of preprocessing parameters and onset detection methods, developed thanks to the close collaboration between biomedical engineers and clinicians, is presented in the paper. The present work presents also the implementation and testing of the software for the data pre-processing.

In this work the problem of orientation tracking based on inertial/magnetic sensors is restated in geometric terms, in particular an intrinsic observer, i.e. an observer whose performance does not depend on a specific choice of coordinates, is derived on the Lie group of rigid body rotations SO(3). Measurements of the gravitational and geomagnetic fields are used to estimate orientation errors. A coordinate-free control law is defined on the Lie algebra and fed back in terms of angular velocity that steers the observer towards the correct attitude. A proof of stability for the proposed estimator is provided which relies on the natural (bi-invariant) metric of SO(3). The observer results stable for almost the whole configuration space. Presence of unstable equilibria as a limitation for global stability is also discussed. Based on the proposed intrinsic control law, a filter is designed which implements the observer. Simulations are presented that test the numerical implementation of the proposed observer

This paper presents a new biphasic electrofluidic rotary actuator relying on electrochemical mechanisms. Gas (H2and O2) is generated at two platinum (Pt) electrodes, immersed in an electrolytic solution, when a low voltage, V, is applied to them. The generated gas pressurizes a closed elastomeric chamber, that deforms under the action of the applied pressure. The chamber is shaped so that no structural failure occurs and the desired output motion is produced, without any need for additional mechanical means such as motion converters. Simple thermodynamic considerations allow estimating the electromechanical coupling factors for the most common external loads.

Movement and behavior analysis is a key research area in the domain of biomedical engineering and in many other medical research domains aiming at the understanding of physiological motor and cognitive basic mechanisms. The systematic application of robotic and mechatronic technologies to realize new tools and measurement methods for quantitatively assessing motor and cognitive functions in humans as well as in animal models is gaining an increasing popularity. This work represents a first step towards the development of a sensorised environment for behavioral phenotyping of animal models. In particular, this paper focuses on tremor analysis in reeler mice, an emerging potential animal model for anatomical and behavioral traits observed in autism. Ground reaction force (GRF) sensing is indeed the most direct means of measuring tremor. Although force platforms have extensively been used for large size animals, only few attempts have been made to measure GRF at a single paw for animals as small as mice or rats. Under the hypothesis that in-plane GRF components are directly connected to tremor, a small size, low-cost, 2-axis force sensor for measuring the in-plane components of GRF was designed and developed. Special care was paid to design a structure that would allow self-aligned assembly, for repeatability, and modularity for combining multiple platforms for a sensorised floor. Preliminarily testing was performed with both reeler and wildtype mice. Fourier analysis was deployed to extract information due to tremor, validating the hypothesis of a direct connection between tremor and in-plane GRFs.

This paper presents the fabrication and the testing of piezoelectric unimorph actuators with embedded piezoelectric sensors which are meant to be used for the actuation of the Micromechanical Flying Insect (MFI). First the fabrication process of a piezoelectric bending actuator comprising a standard unimorph and a rigid extension is described together with the advantages of adding such an extension. Then the convenience of obtaining an embedded piezoelectric sensor by a simple and inexpensive variation of the fabrication process is pointed out. A model for the sensor embedded into a unimorph actuator with rigid extension is derived together with its flat response band limits. Calibration steps are also outlined which allow, despite residual parasitic actuator-sensor coupling, the use of the actuator with the embedded sensor for measuring position and inertial forces when external mechanical structures are driven. An experiment is carried out which validates the model for the actuator/sensor device under desired operating conditions. Preliminary application of the fabricated device to the MFI is also presented where the mechanical power fed into the wing is estimated.

This paper presents an analysis of the major mechanical component (the thorax) of the micromechanical flying insect (MFI), a centimeter sized aerial vehicle currently in development at UC Berkeley. We present a description of the kinematics of the mechanism which converts piezoelectric actuation into complex 3D wing motion. A complete non-linear modeling of the system based on the Lagrangian energy technique is presented. A design methodology is presented in order to achieve optimal matching conditions. Two kinds of sensors which are presently utilized on the MFI are described. Experimental results are presented which validate some of the modeled non-linear aspects of the mechanism.

This paper focuses on the design, fabrication and characterization of unimorph actuators for a microaerial flapping mechanism. PZT-SH and PZN-PT are investigated as piezoelectric layers in the unimorph actuators. Design issues for microaerial flapping actuators are discussed, and criteria for the optimal dimensions of actuators are determined. For low power consumption actuation, a square wave based electronic driving circuit is proposed. Fabricated piezoelectric unimorphs are characterized by an optical measurement system in quasi-static and dynamic mode. Experimental performance of PZT-5H and PZIV-PT based unimorphs is compared with desired design specifications. A 1-DOF flapping mechanism with a PZT-SH unimorph is constructed, and 180/spl deg/ stroke motion at 95 Hz is achieved. Thus, it is shown that unimorphs could be promising flapping mechanism actuators.