In this paper we propose a method for online motion planning of constrained nonlinear systems. The method consists of three steps: the offline generation of a library of parametric trajectories via direct trajectory optimization, the online search in the library for the best candidate solution to the optimal control problem we aim to solve, and the online refinement of this trajectory. The last phase of this process takes advantage of a sensitivity-like analysis and guarantees to comply with the first-order approximation of the constraints even in case of active set changes. Efficiency of the trajectory generation process is discussed and a valid strategy to minimize online computations is proposed; together with this, an effective procedure for searching the candidate trajectory is also presented. As a case study, we examine optimal control of a planar soft manipulator performing a pick-and-place task: through simulations and experiments, we show how crucial online computation times are to achieve considerable energy savings in the presence of variability of the task to perform.
An effective robotic wrist represents a key en- abling element in robotic manipulation, especially in prosthetics. In this paper, we propose an under-actuated wrist system, which is also adaptable and allows to implement different under- actuation schemes. Our approach leverages upon the idea of soft synergies - in particular the design method of adaptive synergies - as it derives from the field of robot hand design. First we intro- duce the design principle and its implementation and function in a configurable test bench prototype, which can be used to demonstrate the feasibility of our idea. Furthermore, we report on results from preliminary experiments with humans, aiming to identify the most probable wrist pose during the pre-grasp phase in activities of daily living. Based on these outcomes, we calibrate our wrist prototype accordingly and demonstrate its effectiveness to accomplish grasping and manipulation tasks.
Human hands are capable of a variety of movements, thanks to their extraordinary biomechanical structure and rely- ing on the richness of human tactile information. Recently, soft robotic hands have opened exciting possibilities and, at the same time, new issues related to planning and control. In this work, we propose to study human strategies in environmental constraint exploitation to grasp objects from a table. We have considered both the case where participants’ fingertips were free and with a rigid shell worn on them to understand the role of cutaneous touch. Main kinematic strategies were quantified and classified in an unsupervised manner. The principal strategies appear to be consistent in both experimental conditions, although cluster cardinality differs. Furthermore, as expected, tactile feedback improves both grasp precision and quality performance. Results opens interesting perspective for sensing and control of soft manipulators.
This work has been awarded with the "Best Student Paper Award" and the "Best Paper in Session - Robotics"
The rich variety of human upper limb movements requires an extraordinary coordination of different joints according to specific spatio-temporal patterns. However, unvealing these motor schemes is a challenging task. Principal components have been often used for analogous purposes, but such an approach relies on hypothesis of temporal uncorrelation of upper limb poses in time. To overcome these limitations, in this work, we leverage on functional principal component analysis (fPCA). We carried out experiments with 7 subjects performing a set of most significant human actions, selected considering state-of-the-art grasp taxonomies and human kinematic workspace. fPCA results show that human upper limb trajectories can be reconstructed by a linear combination of few principal time-dependent functions, with a first component alone explaining around 60/70% of the observed behaviors. This allows to infer that in daily living activities humans reduce the complexity of movement by modulating their motions through a reduced set of few principal patterns. Finally, we discuss how this approach could be profitably applied in robotics and bioengineering, opening fascinating perspectives to advance the state of the art of artificial systems, as it was the case of hand synergies.
Current prosthetic hands are frequently rejected in part due to limited functionality and versatility. We assessed the feasibility of a novel prosthetic hand, the SoftHand Pro (SHP), whose design combines soft robotics and hand postural synergies. Able-bodied subjects (n = 23) tracked cursor motion by opening and closing the SHP and performed a grasp-lift-hold-release (GLHR) task with a sensorized cylindrical object of variable weight. The SHP control was driven by electromyographic (EMG) signals from two antagonistic muscles. Although the time to perform the GLHR task was longer for the SHP than native hand for the first few trials (10.2 ± 1.4 s and 2.13 ± 0.09 s, respectively), performance was much faster on subsequent trials (~5 s). The SHP steady-state grip force was significantly modulated as a function of object weight (p <; 0.001). For the native hand, however, peak and steady-state grip forces were modulated to a greater extent (+68% and +91%, respectively). These changes were mediated by the modulation of EMG amplitude and co-contraction. These data suggest that the SHP has a promise for prosthetic applications and point-to-design modifications that could improve the SHP.
Research reported in this publication was supported by the Grainger Foundation, the Eunice Kennedy Shriver National Institute Of Child Health and Human Development of the National Institutes of Health (NIH) under Award Number R21HD081938, and the European Commission within Horizon 2020 2015.ICT.24a Robot- ics RIA, with Grant no. 688857 “SOFTPRO -Synergy-based Open-source Foundations and Technologies for Prosthetics and RehabilitatiOn.”
This work was supported by the European Commission project (Horizon 2020 research program) SOFTPRO 688857, the European Research Council under the Advanced Grant SoftHands “A Theory of Soft Synergies for a New Generation of Artificial Hands,” ERC-291166, and the Proof of Concept Project SoftHand Pro-H, ERC-2016-PoC 727536.
Electrical Impedance Tomography (EIT) is a medical imaging technique that has been recently used to realize stretchable pressure sensors. In this method, voltage measurements are taken at electrodes placed at the boundary of the sensor and are used to reconstruct an image of the applied touch pressure points. The drawback with EIT-based sensors, however, is their low spatial resolution due to the ill-posed nature of the EIT reconstruction. In this paper, we show our performance evaluation of different EIT drive patterns, specifically strategies for electrode selection when performing current injection and voltage measurements. We compare voltage data with Signal-to-Noise Ratio (SNR) and Boundary Voltage Changes (BVC), and study image quality with Size Error (SE), Position Error (PE) and Ringing (RNG) parameters, in the case of one-point and two-point simultaneous contact locations. The study shows that, in order to improve the performance of EIT based sensors, the electrode selection strategies should dynamically change correspondingly to the location of the input stimuli. In fact, the selection of one drive pattern over another can improve the target size detection and position accuracy up to 0.04. and 0.18, respectively.
The past thirty years have seen increasingly rapid advances in the field of laparoscopic surgery, in part because of the use of robots. A well-known example is the da Vinci surgical system. However, far too little attention has been paid to Hand Assisted Laparoscopic Surgery (HALS), a surgery in which the surgeon introduces the non-dominant hand into the abdomen of the patient. The risk of collision between the hand of the surgeon and the tool moved by the robot is the reason why these robots for laparoscopic surgery are not appropriate for HALS. On the other hand, in recent years, there has been an increasing interest in wearables, which have been introduced in our daily life. This interest and the lack of surgery robots for HALS are the reasons to develop a sensing glove which co-works whit a collaborative robot in this kind of surgery. The aim of this paper is to study the use of a sensing glove which will provide information of the movements of the surgeon‚Äôs hand to the collaborative robot. This information determinates the actions that the robot will carry on. The first step was to define different movements of the hand which could be identified. An algorithm identifies these movements using the data given by the sensing glove. For the purpose of algorithm accuracy measurement, 4 persons wearing the sensing glove made a sequence with different movements. The evidence from this study suggests that a sensing glove can be used to send information of the movements of the surgeon‚Äôs hand to a collaborative robot during a HALS.
In this paper we present the design of a one degree of freedom assistive platform to augment the strength of upper limbs. The core element is a variable stiffness actuator, closely reproducing the behavior of a pair of antagonistic muscles. The novelty introduced by this device is the analogy of its control parameters with those of the human muscle system, the threshold lengths. The analogy can be obtained from a proper tuning of the mechanical system parameters. Based on this, the idea is to control inputs by directly mapping the estimation of the muscle activations, e.g. via ElectroMyoGraphic(EMG) sensors, on the exoskeleton. The control policy resulting from this mapping acts in feedforward in a way to exploit the muscle-like dynamics of the mechanical device. Thanks to the particular structure of the actuator, the exoskeleton joint stiffness naturally results from that mapping. The platform as well as the novel control idea have been experimentally validated and the results show a substantial reduction of the subject muscle effort.
Low stiffness elements have a number of applications in Soft Robotics, from Series Elastic Actuators (SEA) to torque sensors for compliant systems. In its general formulation, the design problem of elastic components is complex and depends on several variables: material properties, load range, shape factor and size constraints. Consequently, most of the spring designs presented in literature are based on heuristics or are optimized for specific working conditions. This work presents the design study and characterization of a scalable spoked elastic element with hinge tip constraints. We compared the proposed design with three existing spring principles, showing that the spoked solution is the convenient option for low-stiffness and low shape factor elastic elements. Therefore, a design analysis on the main scaling parameters of the spoked spring, namely number of spokes and type of constraints, is presented. Finally, an experimental characterization has been conducted on physical prototypes. The agreement among simulations and experimental results demonstrates the effectiveness of the proposed concept.
Humans are able to intuitively exploit the shape of an object and environmental constraints to achieve stable grasps and perform dexterous manipulations. In doing that, a vast range of kinematic strategies can be observed. However, in this work we formulate the hypothesis that such ability can be described in terms of a synergistic behavior in the generation of hand postures, i.e., using a reduced set of commonly used kinematic patterns. This is in analogy with previous studies showing the presence of such behavior in different tasks, such as grasping. We investigated this hypothesis in experiments performed by six subjects, who were asked to grasp objects from a flat surface. We quantitatively characterized hand posture behavior from a kinematic perspective, i.e., the hand joint angles, in both pre-shaping and during the interaction with the environment. To determine the role of tactile feedback, we repeated the same experiments but with subjects wearing a rigid shell on the fingertips to reduce cutaneous afferent inputs. Results show the persistence of at least two postural synergies in all the considered experimental conditions and phases. Tactile impairment does not alter significantly the first two synergies, and contact with the environment generates a change only for higher order Principal Components. A good match also arises between the first synergy found in our analysis and the first synergy of grasping as quantified by previous work. The present study is motivated by the interest of learning from the human example, extracting lessons that can be applied in robot design and control. Thus, we conclude with a discussion on implications for robotics of our findings.
Robotic hands embedding human motor control principles in their mechanical design are getting increasing interest thanks to their simplicity and robustness, combined with good performance. Another key aspect of these hands is that humans can use them very effectively thanks to the similarity of their behavior with real hands. Nevertheless, controlling more than one degree of actuation remains a challenging task. In this paper, we take advantage of these characteristics in a multi-synergistic prosthesis. We propose an integrated setup composed of Pisa/IIT SoftHand 2 and a control strategy which simultaneously and proportionally maps the human hand movements to the robotic hand. The control technique is based on a combination of non-negative matrix factorization and linear regression algorithms. It also features a real-time continuous posture compensation of the electromyographic signals based on an IMU. The algorithm is tested on five healthy subjects through an experiment in a virtual environment. In a separate experiment, the efficacy of the posture compensation strategy is evaluated on five healthy subjects and, finally, the whole setup is successfully tested in performing realistic daily life activities.
This paper presents an approach to achieve adaptive grasp of unknown objects whose position is only approximately known via point-cloud data. We exploit the adaptability of a soft robotic hand which can autonomously conform to the shape of a grasped object if properly approached. Once a grasp approach has been preliminarily planned based only on rough estimates of the object position, the hand is shaped to a pregrasp configuration. Before closing the hand, a sensor-based algorithm is applied that corrects the relative hand-object posture so as to enhance the probability that the object is uniformly approached by all fingers, thus avoiding undesired premature contacts. The algorithm minimizes the distance between the hand's fingerpads and the object by continuously controlling both the wrist pose and orientation and the hand closure. Experimental studies with a Kuka-LWR arm and a Pisa/IIT Softhand illustrate the benefit of the developed technique and the improvement in grasping performance with respect to open-loop execution of grasps planned on the basis of prior RGB-D cues only.
In humanoids and other redundant robots interacting with the environment, one can often choose between different configurations and control parameters to achieve a given task. A classic tool to describe specifications of the desired force/displacement behavior in such problems is the stiffness ellipsoid, whose geometry is affected by the choice of parameters in both joint control and redundancy resolution—namely, gains and angles. As is well known, impedance control techniques can regulate gains to realize any desired shape of the Cartesian stiffness ellipsoid at the end-effector, so that robot geometry selection could appear secondary. However, humans do not use this possibility: To control the stiffness of our arms, we predominantly use arm configurations. Why is that, and does it makes sense to do the same in robots? To understand this discrepancy, we provide a more complete analysis of the task-space force/deformation behavior of compliant redundant arms to illustrate why the arm geometry plays a dominant role in interaction capabilities of robots. We introduce the notion of allowable Cartesian force/displacement (“stiffness feasibility”) regions (SFR) for compliant robots with given torque boundaries. We show that different robot configurations modify such regions and explore the role of robot geometry in achieving an appropriate SFR for the task at hand. The novel concepts and definitions are first illustrated in simulations. Experimental results are then provided to verify the effectiveness of the proposed Cartesian force and stiffness control.
The paper presents the first simulative results and algorithmic developments of the task-priority based control applied to a distributed sampling network in an area coverage or adaptive sampling mission scenario. The proposed approach allowing the fulfilment of a chain of tasks with decreasing priority each of which directly related to both operability and safety aspects of the entire mission. The task-priority control is presented both in the centralized and decentralized implementations showing a comparison of performance. Finally simulations of the area coverage mission scenario are provided showing the effectiveness of the proposed approach.
The cerebellum is a crucial brain structure in enabling precise motor control in animals. Recent advances suggest that the timing of the plasticity rule of Purkinje cells, the main cells of the cerebellum, is matched to behavioral function.Simultaneously, counter-factual predictive control (CFPC), a cerebellar-based control scheme, has shown that the optimal rule for learning feed-forward action in an adaptive filter playing the role of the cerebellum must include a forward model of the system controlled. Here we show how the same learning rule obtained in CFPC, which we term as Model-enhanced least mean squares (ME-LMS), emerges in the problem of learning the gains of a feedback controller. To that end, we frame a model-reference adaptive control (MRAC) problem and derive an adaptive control scheme treating the gains of a feedback controller as if they were the weights of an adaptive linear unit. Our results demonstrate that the approach of controlling plasticity with a forward model of the subsystem controlled can provide a solution to a wide set of adaptive control problems
The endpoint stiffness of the human arm has been long recognized as a key component ensuring the quasi-static stability of the arm physical interactions with the external world. Similarly, the understanding of the joint stiffness behavior can provide complementary insights, e.g., on the underlying stiffness regulation principles across different joints including the nullspace stiffness profiles. Traditionally, the experimental modeling and estimation of the human arm joint stiffness is achieved by the transformation of the identified arm endpoint stiffness to the joint coordinates. Due to the underlying kinematic redundancy, the obtained joint stiffness matrix is rank-deficient which implies that the information in the joint stiffness matrix is incomplete. While in robotics applications this issue can be addressed by designing a desired nullspace stiffness behavior through appropriate projections, the use of a similar technique in the identification of human joint stiffness profile is meaningless. Hence, the first objective of this work is to address this issue by developing a novel technique to identify the complete and physiologically meaningful joint stiffness of human arm. Second, we present a model-based online estimation technique to estimate the seven-dimensional complete joint stiffness in various arm poses and activation levels of the two dominant arm muscles that correspond to the geometric and volume modifications of the joint stiffness profile, respectively.
In this work, we present WALK-MAN, a humanoid platform that has been developed to operate in realistic unstructured environment, and demonstrate new skills including powerful manipulation, robust balanced locomotion, high-strength capabilities, and physical sturdiness. To enable these capabilities, WALK-MAN design and actuation are based on the most recent advancements of series elastic actuator drives with unique performance features that differentiate the robot from previous state-of-the-art compliant actuated robots. Physical interaction performance is benefited by both active and passive adaptation, thanks to WALK-MAN actuation that combines customized high-performance modules with tuned torque/velocity curves and transmission elasticity for high-speed adaptation response and motion reactions to disturbances. WALK-MAN design also includes innovative design optimization features that consider the selection of kinematic structure and the placement of the actuators with the body structure to maximize the robot performance. Physical robustness is ensured with the integration of elastic transmission, proprioceptive sensing, and control. The WALK-MAN hardware was designed and built in 11 months, and the prototype of the robot was ready four months before DARPA Robotics Challenge (DRC) Finals. The motion generation of WALK-MAN is based on the unified motion-generation framework of whole-body locomotion and manipulation (termed loco-manipulation). WALK-MAN is able to execute simple loco-manipulation behaviors synthesized by combining different primitives defining the behavior of the center of gravity, the motion of the hands, legs, and head, the body attitude and posture, and the constrained body parts such as joint limits and contacts. The motion-generation framework including the specific motion modules and software architecture is discussed in detail. A rich perception system allows the robot to perceive and generate 3D representations of the environment as well as detect contacts and sense physical interaction force and moments. The operator station that pilots use to control the robot provides a rich pilot interface with different control modes and a number of teleoperated or semiautonomous command features. The capability of the robot and the performance of the individual motion control and perception modules were validated during the DRC in which the robot was able to demonstrate exceptional physical resilience and execute some of the tasks during the competition.
Decellularized human livers are considered the perfect extracellular matrix (ECM) surrogate because both three-dimensional architecture and biological features of the hepatic microenvironment are thought to be preserved. However, donor human livers are in chronically short supply, both for transplantation or as decellularized scaffolds, and will become even scarcer as life expectancy increases. It is hence of interest to determine the structural and biochemical properties of human hepatic ECM to derive design criteria for engineering biomimetic scaffolds. The intention of this work was to obtain quantitative design specifications for fabricating scaffolds for hepatic tissue engineering using human livers as a template. To this end, hepatic samples from five patients scheduled for hepatic resection were decellularized using a protocol shown to reproducibly conserve matrix composition and microstructure in porcine livers. The decellularization outcome was evaluated through histological and quantitative image analyses to evaluate cell removal, protein, and glycosaminoglycan content per unit area. Applying the same decellularization protocol to human liver samples obtained from five different patients yielded five different outcomes. Only one liver out of five was completely decellularized, while the other four showed different levels of remaining cells and matrix. Moreover, protein and glycosaminoglycan content per unit area after decellularization were also found to be patient- (or donor-) dependent. This donor-to-donor variability of human livers thus precludes their use as templates for engineering a generic "one-size fits all" ECM-mimic hepatic scaffold.
Touch provides an important cue to perceive the physical properties of the external objects. Recent studies showed that tactile sensation also contributes to our sense of hand position and displacement in perceptual tasks. In this study, we tested the hypothesis that, sliding our hand over a stationary surface, tactile motion may provide a feedback for guiding hand trajectory. We asked participants to touch a plate having parallel ridges at different orientations and to perform a self-paced, straight movement of the hand. In our daily-life experience, tactile slip motion is equal and opposite to hand motion. Here, we used a well-established perceptual illusion to dissociate, in a controlled manner, the two motion estimates. According to previous studies, this stimulus produces a bias in the perceived direction of tactile motion, predicted by tactile flow model. We showed a systematic deviation in the movement of the hand towards a direction opposite to the one predicted by tactile flow, supporting the hypothesis that touch contributes to motor control of the hand. We suggested a model where the perceived hand motion is equal to a weighted sum of the estimate from classical proprioceptive cues (e.g., from musculoskeletal system) and the estimate from tactile slip.
This work is supported in part by the European Research Council under the Advanced Grant SoftHands “A Theory of Soft Synergies for a New Generation of Artificial Hands” no. ERC-291166, by the EU H2020 project “SOFTPRO: Synergy-based Open-source Foundations and Technologies for Prosthetics and RehabilitatiOn” (no. 688857) and by the EU FP7 project (no. 601165), “WEARable HAPtics for Humans and Robots (WEARHAP)”. We thank Priscilla Balestrucci and Colleen P. Ryan for helpful comments and suggestions.
Myoelectric prostheses have seen increased application in clinical practice and research, due to their potential for good functionality and versatility. Yet, myoelectric prostheses still suffer from a lack of intuitive control and haptic feedback, which can frustrate users and lead to abandonment. To address this problem, we propose to convey proprioceptive information for a prosthetic hand with skin stretch using the Rice Haptic Rocker. This device was integrated with the myo-controlled version of Pisa/IIT SoftHand and a size discrimination test with 18 able bodied subjects was performed to evaluate the effectiveness of the proposed approach. Results show that the Rice Haptic Rocker can be successfully used to convey proprioceptive information. A Likert survey was also presented to the experiment participants, who evaluated the integrated setup as easy to use and effective in conveying proprioception.
The authors gratefully acknowledge Matteo Rossi for his valuable advice and Mikaela Juzswik for her unique contribution in the physical realization of some of the equipment used in the experiments. This work was partially supported by the European Community funded project WEARHAP (contract 601165), by the European Commission project (Horizon 2020 research program) SOFTPRO (no. 688857), by the ERC Advanced Grant no. 291166 SoftHands and by the NSF grant IIS-1065497.
Soft robots (SRs) represent one of the most significant recent evolutions in robotics. Designed to embody safe and natural behaviors, they rely on compliant physical structures purposefully designed to embody desirable and sometimes variable impedance characteristics. This article discusses the problem of controlling SRs. We start by observing that most of the standard methods of robotic control—e.g., high-gain robust control, feedback linearization, backstepping, and active impedance control—effectively fight against or even completely cancel the physical dynamics of the system, replacing them with a desired model. This defeats the purpose of introducing physical compliance. After all, what is the point of building soft actuators if we then make them stiff by control? An alternative to such approaches can be conceived by observing humans, who can obtain good motion accuracy and repeatability while maintaining the intrinsic softness of their bodies. In this article, we show that an anticipative model of human motor control, using a feedforward action combined with low-gain feedback, can be used to achieve human-like behavior. We present an implementation of such an idea that uses iterative learning control. Finally, we present the experimental results of the application of such learned anticipative control to a physically compliant robot. The control application achieves the desired behavior much better than a classical feedback controller used for comparison.
This work is supported by European Commission grant H2020-ICT-645599 (“SOMA”: SOft Manipulation) and European Research Council Advanced grant 291166 (“SoftHands”).