Search

With the ongoing Industry 4.0 (I4.0) revolution, plant management and supervision play a key role in the development and (re-)design of industrial plants. In the arising scenarios, the need to coordinate human workers and autonomous systems, sharing the same environment, teaming together, becomes a fundamental requirement. Indeed, even though automation in standard assembly lines has reached high efficiency and reliability, for complex and new applications, a certain amount of failure must be considered for future addressing. This paper presents a framework for the flexible coordination of such a complex and heterogeneous cyber-physical system. A digital Ttwin mirrors in real-time the plant system, while a dashboard displays plant status, providing the human operators with fundamental tools for supervision and prompt intervention in case of failure. The framework was developed and tested in an industrially relevant environment, specifically for the assembly of the interior of an aircraft fuselage.

Series Elastic Actuators (SEAs) are multi-degree-of-freedom spring-mass-damper systems that employ elastic transmission elements to reduce mechanical impedance and buffer energy from motor (input) to load (output). This work deals with the mathematical modeling of SEAs and their sizing for given application requirements. In designing traditional actuation systems a common approach is to optimize the speed ratio of the transmission with respect to a given figure of merit, such as the motor rated torque or power, once the output load torque and motion trajectories are given. Taking inspiration from existing techniques developed for the sizing of traditional actuators having just one degree of freedom, a new approach is presented here for two degrees of freedom SEAs, which enables the designer to optimally select at the same time both the ratio and the compliance of the elastic transmission.

The recovery of symmetric and efficient walking is one of the key goals of a rehabilitation program in patients with stroke. The use of overground exoskeletons alongside conventional gait training might help foster rhythmic muscle activation in the gait cycle toward a more efficient gait. About twenty-nine patients with subacute stroke have been recruited and underwent either conventional gait training or experimental training, including overground gait training using a wearable powered exoskeleton alongside conventional therapy. Before and after the rehabilitation treatment, we assessed: (i) gait functionality by means of clinical scales combined to obtain a Capacity Score, and (ii) gait neuromuscular lower limbs pattern using superficial EMG signals. Both groups improved their ability to walk in terms of functional gait, as detected by the Capacity Score. However, only the group treated with the robotic exoskeleton regained a controlled rhythmic neuromuscular pattern in the proximal lower limb muscles, as observed by the muscular activation analysis. Coherence analysis suggested that the control group (CG) improvement was mediated mainly by spinal cord control, while experimental group improvements were mediated by cortical-driven control. In subacute stroke patients, we hypothesize that exoskeleton multijoint powered fine control overground gait training, alongside conventional care, may lead to a more fine-tuned and efficient gait pattern.

Introduction: In advanced Non-Small Cell Lung Cancer (NSCLC), Programmed Death Ligand 1 (PD-L1) remains the only used biomarker to candidate patients to immunotherapy (IO) with many limits. Given the complex dynamics of the immune system it is improbable that a single biomarker could be able to profile prediction with high accuracy. A promising solution cope with this complexity is provided by Artificial Intelligence (AI) and Machine Learning (ML), which are techniques able to analyse and interpret big multifactorial data. The present study aims at using AI tools to improve response and efficacy prediction in NSCLC patients treated with IO.MethodsReal world data (clinical data, PD-L1, histology, molecular, lab tests) and the blood microRNA signature classifier (MSC), which include 24 different microRNAs, were used. Patients were divided into responders (R), who obtained a complete or partial response or stable disease as best response, and non-responders (NR), who experienced progressive or hyperprogressive disease and those who died before the first radiologic evaluation. Moreover, we used the same data to determine if the overall survival of the patients was likely to be shorter or longer than 24 months from baseline IO. For A literature review and forward feature selection technique was used to extract a specific subset of the patients’ data. To develop the final predictive model, different ML methods have been tested, i.e., Feedforward Neural Network (FFNN), Logistic Regression (LR), K-nearest neighbours (K-NN), Support Vector Machines (SVM), and Random Forest (RF).Results 200 patients were included. 164 out of 200 (i.e., only those patients with PD-L1 data available) were considered in the model, 73 (44.5%) were R and 91 (55.5%) NR. Overall, the best model was the LR and included 5 features: 2 clinical features including the ECOG performance status and IO-line of therapy; 1 tissue feature such as PD-L1 tumour expression; and 2 blood features including the MSC test and the neutrophil-to-lymphocyte ratio (NLR). The model predicting R/NR of the patient achieves accuracy ACC= 0.756, F1 score F1=0.722, and Area Under the ROC Curve AUC=0.82. The use of the PD-L1 alone has an ACC=0.655. The accuracy of the ML models excluding some of the features from the model were as follow: without PD-L1 value (ACC=0.726), MSC (ACC=0.750), and both PD-L1 and MSC (ACC=0.707), i.e., considering only clinical features. At data cut-off (Nov 2020), median Overall Survival (mOS) for R was 38.5 months (m) (95%IC 23.9 - 53.1) vs 3.8 m (95%IC 2.8 - 4.7) for NR, with pConclusionsThe results suggest that the integration of multifactorial data provided by ML techniques is a useful tool to improve personalized selection of NSCLC patients candidates to IO. In particular, compared to PD-L1 alone the expected improvement was around 10%. In particular, the model shows that the higher the ECOG, NLR value, IO-line, and MSC test level the lower the response, and the higher PD-L1 the higher the response. Considering the difference in survival among R and NR groups, these results suggest that the model can also be used to indirectly predict survival. Moreover, a second model was able to predict long survival patients with good accuracy.

It is common for animals to use self-generated movements to actively sense the surrounding environment. For instance, rodents rhythmically move their whiskers to explore the space close to their body. The mouse whisker system has become a standard model to study active sensing and sensorimotor integration through feedback loops. In this work, we developed a bioinspired spiking neural network model of the sensorimotor peripheral whisker system, modelling trigeminal ganglion, trigeminal nuclei, facial nuclei, and central pattern generator neuronal populations. This network was embedded in a virtual mouse robot, exploiting the Neurorobotics Platform, a simulation platform offering a virtual environment to develop and test robots driven by brain-inspired controllers. Eventually, the peripheral whisker system was properly connected to an adaptive cerebellar network controller. The whole system was able to drive active whisking with learning capability, matching neural correlates of behaviour experimentally recorded in mice.

Dealing with the physical interaction between humans and robots, Series Elastic Actuators (SEAs) are identified as one solution to overcome many limits, such as reducing contact forces or detect collisions. Nevertheless, the low-damping dynamic of a SEA can lead to undesired behaviours, especially during particular applications where a high level of precision is required. In this paper, a linear control architecture to enhance the damping performance of a SEA is presented. The proposed structure consists in a cascade control where loops are regulated using three types of controllers: PI, PD and a generalized controller specifically designed to damp oscillations. A frequency-domain approach with related constraints could not satisfy the time-domain goal in term of oscillation damping, for this reason an optimization problem able to consider them both is taken into account. A robust design is mandatory to the model mismatch introduced by neglecting coupling between motor. Therefore, robustness constraints are introduced in the optimization procedure. Indeed, the effectiveness of the control architecture is tested on a real compliant robot with six degrees of freedom equipped with as many SEAs. Each test aims to highlight the damping performance of the controlled system while the robot performs various tasks or it is subject to external disturbances.

Background: Appropriate training modalities for post-stroke upper-limb rehabilitation are key features for effective recovery after the acute event. This work presents a novel human-robot cooperative control framework that promotes compliant motion and renders different high-level human-robot interaction rehabilitation modalities under a unified low-level control scheme. Methods: The presented control law is based on a loadcell-based impedance controller provided with positive-feedback compensation terms for disturbances rejection and dynamics compensation. We developed an elbow flexion-extension experimental setup, and we conducted experiments to evaluate the controller performances. Seven high-level modalities, characterized by different levels of (i) impedance-based corrective assistance, (ii) weight counterbalance assistance, and (iii) resistance, have been defined and tested with 14 healthy volunteers.Results: The unified controller demonstrated suitability to promote good transparency and render compliant and high-impedance behavior at the joint. Superficial electromyography results showed different muscular activation patterns according to the rehabilitation modalities. Results suggested to avoid weight counterbalance assistance, since it could induce different motor relearning with respect to purely impedance-based corrective strategies. Conclusion: We proved that the proposed control framework could implement different physical human-robot interaction modalities and promote the assist-as-needed paradigm, helping the user to accomplish the task, while maintaining physiological muscular activation patterns. Future insights involve the extension to multiple degrees of freedom robots and the investigation of an adaptation control law that makes the controller learn and adapt in a therapist-like manner.

International audience; Robotized assembly and manufacturing often require to modify the robot motion at runtime. When the primary constraint is to preserve the geometrical path as much as possible, it is convenient to scale the nominal trajectory in time to meet the robot constraints. Look-ahead techniques are computationally heavy, while non-look-ahead ones usually show poor performance in critical circumstances. This paper proposes a novel technique that can be embedded in non-look-ahead scaling algorithms to improve their performance. The proposed method takes into account the robot velocity, acceleration, and torque limits and modifies the velocity profile based on an approximated look-ahead criterion. To do this, it considers only the last point of a look-ahead window and, by linearizing the problem, it computes the maximum admissible robot velocity. The technique can be applied to existing trajectory scaling algorithms to confer look-ahead properties on them. Simulation and experimental results on a 6-degree-of-freedom manipulator show that the proposed method significantly reduces path-following errors.

Background Early detection of gait impairments in older adults allows the early uncovering of fall risk and/or cognitive deficits, resulting in timely interventions. Dual-task paradigms have been shown to be more sensitive than single-task conditions for the detection of subtle yet relevant gait impairments. Research question Can a system - encompassing a pair of instrumented insoles and a customized mobile app - transparently and accurately study ecological walking activities in single- and dual-task conditions, with the aim of detecting early and subtle age-related alterations of gait? Methods The system was tested on 19 older adults during outdoor walking (two identical single-task trials and two motor-cognitive dual-task trials with the user engaged in a simple phone call and in a cognitive-demanding phone call). A single-task cognitive trial was included. Relative reliability of the gait parameters provided by the insoles during single-task walking was investigated (Intraclass Correlation Coefficient). The effect of dual tasking on both motor (Friedman test) and cognitive (Wilcoxon signed-rank test) domains was studied. To study usability, the system was tested on 5 older adults in real-life environment over 3 months. Results Most of the parameters showed excellent reliability. Independently from the cognitive demand, walking while talking resulted in increased gait cycle and step time, with a prolonged stance phase due to an augmented double-support. Variability of gait cycle and stance phase increased only during the most demanding dual-task. Dual tasking resulted in a reduced cognitive score. Usability feedback were excellent, with users reporting to understand the usefulness of the devised system and to feel at ease when using the system and the insoles. Significance This work paves the way toward fruitful applications of the devised system to achieve accurate and ecological monitoring of daily-life walking activities, with the final aim of detecting early and subtle alterations of gait.

In human-robot collaboration, the robot is required to properly react to the operator's movements to avoid collisions and interference. In this abstract, we present a path re-planning strategy to change the robot path online when a human is on the robot's way. The approach exploits a set of pre-computed paths to speed up the search for a new feasible solution and to improve the trajectory's readability. In addition, the algorithm continuously optimizes the current path in an anytime fashion to deal with strict computing time requirements. Experimental tests demonstrate the effectiveness of the proposed framework with respect to industry best practices.

SUMMARYThe V-groove joint of thick wall intersecting pipes must be filled by multi-layer weld. The welding path of intersecting pipes is complicated, and hence multi-layer welds increase the complexity of the problem. This paper proposes a methodology for path planning of multi-layer weld of thick wall intersecting pipes. The methodology is based on measuring the electrode pose located in both side and front views of intersecting pipes. In order to compensate for the path deviation around the pipe circumference, the measured values are used to interpolate the path of each pass between two views. The methodology has been applied in a case study. Simulation results approve that multi-layer weld appropriately fills the V-groove joint space around the pipe circumference. In addition, collision avoidance between welding torch and pipes is considered by introducing a safety ring. While the robot wrist moves inside the safety ring, no collision occurs. Simulation results show the robustness of the proposed path planning method, introduced for collision avoidance.

The cerebellum is essential for motor learning and adaptation since it can change the relationship between sensory input and motor output. Cerebellar activity contributes to the motor execution with the aim of optimizing the resulting performance. This optimization is driven by the error information, that the cerebellum receives through the sensory feedback. In this study, we aim at studying the cerebellar contribution in the saccadic adaptation, fine and rapid control of eye movements. We set up a virtual experiment where an iCub robot, controlled by a cerebellar-inspired spiking neural network, was challenged in a saccadic adaptation task. The virtual robot, at first, made inaccurate saccades and therefore attempted to adjust them, trial after trial, based on sensory error feedback. The spiking neural network model was equipped with plasticity mechanisms, that allowed the robot to change and adapt the amplitude of the saccade movement after a few trials. Interestingly, we observed that for this task a single plasticity mechanism was sufficient to produce a fast adaptation, while in other cerebellar-driven tasks, the addition of more plasticities granted an improvement of the adaptation performances.

A key challenge in human-robot shared workspace is defining the decision criterion to select the next task for a fluent, efficient and safe collaboration. While working with robots in an industrial environment, tasks may comply with precedence constraints to be executed. A typical example of precedence constraint in industry occurs at an assembly station when the human cannot perform a task before the robot ends on its own. This paper presents a methodology based on the Maximum Entropy Inverse Optimal Control for the identification of a probability distribution over the human goals, packed into a software tool for human-robot shared-workspace collaboration. The software analyzes the human goal and the goal precedence constraints, and it is able to identify the best robot goal along with the relative motion plan. The approach used is, an algorithm for the management of goal precedence constraints and the Partially Observable Markov Decision Process (POMDP) for the selection of the next robot action. A comparison study with 15 participants was carried out in a real world assembly station. The experiment focused on evaluating the task fluency, the task efficiency and the human satisfaction. The presented model displayed reduction in robot idle time and increased human satisfaction.

While the research interest for exoskeletons has been rising in the last decades, missing standards for their rigorous evaluation are potentially limiting their adoption in the industrial field. In this context, exoskeletons for worker support have the aim to reduce the physical effort required by humans, with dramatic social and economic impact. Indeed, exoskeletons can reduce the occurrence and the entity of work-related musculoskeletal disorders that often cause absence from work, resulting in an eventual productivity loss. This very urgent and multifaceted issue is starting to be acknowledged by researchers. This article provides a systematic review of the state of the art for functional performance evaluation of low-back exoskeletons for industrial workers. We report the state-of-the-art evaluation criteria and metrics used for such a purpose, highlighting the lack of a standard for this practice. Very few studies carried out a rigorous evaluation of the assistance provided by the device. To address also this topic, the article ends with a proposed framework for the functional validation of low-back exoskeletons for the industry, with the aim to pave the way for the definition of rigorous industrial standards.

Nowadays, clinicians have multiple tools that they can use to stimulate the brain, by means of electric or magnetic fields that can interfere with the bio-electrical behaviour of neurons. However, it is still unclear which are the neural mechanisms that are involved and how the external stimulation changes the neural responses at network-level. In this paper, we have exploited the simulations carried out using a spiking neural network model, which reconstructed the cerebellar system, to shed light on the underlying mechanisms of cerebellar Transcranial Magnetic Stimulation affecting specific task behaviour. Namely, two computational studies have been merged and compared. The two studies employed a very similar experimental protocol: a first session of Pavlovian associative conditioning, the administration of the TMS (effective or sham), a washout period, and a second session of Pavlovian associative conditioning. In one study, the washout period between the two sessions was long (1 week), while the other study foresaw a very short washout (15 min). Computational models suggested a mechanistic explanation for the TMS effect on the cerebellum. In this work, we have found that the duration of the washout strongly changes the modification of plasticity mechanisms in the cerebellar network, then reflected in the learning behaviour.

Dependability of robot co-workers plays an important role in increasing the effectiveness of human-robot interaction in manufacturing. Our goal is to understand the role of motion planning parameters in human-robot collaboration and to provide guidelines for the selection of the most suitable motion planner. The human factors analysis provided in this paper highlights that repeatability of the motion and predictability of the robot timing affect the quality of human-robot collaboration. Copyright (C) 2020 The Authors.

Combining task and motion planning efficientlyin human-robot collaboration (HRC) entails several challengesbecause of the uncertainty conveyed by the human behavior.Tasks plan execution should be continuously monitored andupdated based on the actual behavior of the human andthe robot to maintain productivity and safety. We proposecontrol-based approach based on two layers, i.e., task planningand action planning. Each layer reasons at a different levelof abstraction: task planning considers high-level operationswithout taking into account their motion properties; actionplanning optimizes the execution of high-level operations basedon current human state and geometric reasoning. The result is ahierarchical framework where the bottom layer gives feedbackto top layer about the feasibility of each task, and the top layeruses this feedback to (re)optimize the process plan. The methodis applied to an industrial case study in which a robot and ahuman worker cooperate to assemble a mosaic.

The analysis of the writing gesture has been successfully investigated in the diagnosis of age-related diseases, but the current technologies and methods still do not allow the ecological daily monitoring of handwriting, mostly because they rely on standardized writing protocols. In this study, we first designed and validated a novel electronic ink pen, equipped with motion and writing force sensing, for the ecological daily-life monitoring of handwriting in uncontrolled environments. We used the pen to acquire writing activities from healthy adults, from which we computed useful handwriting and tremor indicators. We evaluated the reliability of our measurements by computing the intraclass correlation coefficients (ICC) and the minimal detectable changes (MDC). Moderate to excellent reliability were obtained for all the handwriting indicators computed in two different writing tasks. MDC values can be used as reference to discriminate a real change in the handwriting parameters from a measurement error in longitudinal studies. These results pave the way towards the use of the pen for daily life handwriting monitoring.

Within the project EURECA (https://cordis.europa.eu/project/id/738039) we investigated innovative control strategies to assist human operators. Human-robot cooperation is increasingly demanded in industrial applications. Many tasks require the robot to enhance the capabilities of humans, allowing them to execute onerous tasks or improving their functionalities. Besides wearable robotics, standard industrial manipulators are common solutions adopted to empower humans. With this aim, the paper describes a fuzzy-impedance control approach for assisting the human operator in onerous industrial applications. The developed method allows setting in real-time the set-point of the impedance controller based on human intentions. Two fuzzy membership functions have been defined, respectively, on the basis of interaction force derivative and robot velocity signals, allowing to calculate the needed assistance level for the human operator. The effectiveness of the proposed approach is verified on KUKA iiwa 14 R820 in an experimental procedure where human subjects evaluate the robot’s performance in a collaborative lifting task.

Complex parts can be successfully manufactured by means of Additive Manufacturing (AM) techniques based on thermoplastic polymer extrusion, whose use for mass production is restricted by their slow printing speed. To address this limitation, a flexible AM platform for big plastic objects has been realized mounting an industrial screw-based extruder on an anthropomorphic robot. An experimental campaign has been performed to set a suitable range of relevant process parameters, with the aim of ensuring a regular deposited layer geometry. Moreover, a closed-loop control strategy has been developed to correct the robot height based on data measured during the material deposition, thus further improving the process parameter setting and compensating the material shrinkage or other unexpected defects. Eventually, an online re-slicing algorithm has been implemented to preserve the desired height of the manufactured object, despite the layer height changes. The proposed approach allows a deposition flow rate up to 1250 cm3/h within a building volume limited only by the robot workspace.

This paper deals with the control of a collaborative robot manipulator with series elastic actuators. In particular, robust tuning rules for cascade control of the joints are presented. Both the motor velocity and link position control loops are considered. The proposed tuning rules allow the online computation of robust control parameters to cope with the different link reflected inertia. Experimental results show the effectiveness of the method in real applications. (C) 2018, IFAC (International Federation of Automatic Control) Hosting by Elsevier Ltd. All rights reserved.

Conventional material removal systems, like CNC milling machines, have proven to be very effective and accurate. Their major drawbacks are related to machine cost, restricted working area and limitations on the allowed workpiece geometry. In principle, industrial robots (IRs) could be an excellent alternative for machining, being both flexible and cheap (if compared to a CNC machine). On the contrary, IRs do not offer sufficient positioning accuracy and they are prone to vibration onset due to a low capacity of disturbance rejection. These critical aspects currently limit the use of robots in typical machining applications. In this paper, the cutting capability of a serial IR, for light milling operation on aluminum alloy, has been analyzed with respect to the multiple poses allowed by its kinematics redundancy. Firstly, an experimental modal analysis has allowed to identify the variability of the robot static and dynamic compliance over different poses, which influence both tool deflection and vibration onset during milling. The static stiffness of the robot end-effector (EE), which causes tool deflection, has been analyzed by means of an elastic model with lumped parameters. It is shown that the usual static compliance criterion for the optimal pose selection does not guarantee the best performance in terms of vibration level and, thus, surface roughness. This claim has been supported by a set of cutting tests in two significant cases. The results are explained considering the pose-dependent static and the dynamics properties.

While multi-robot cells are being used more often in industry, the problem of work-piece position optimization is still solved using heuristics and the human experience and, in most industrial cases, even a feasible solution takes a considerable amount of trials to be found. Indeed, the optimization of a generic performance index along a path is complex, due to the dimension of the feasible-configuration space. This work faces this challenge by proposing an iterative layered-optimization method that integrates a Whale Optimization and an Ant Colony Optimization algorithm, the method allows the optimization of a user-defined objective function, along a working path, in order to achieve a quasi-optimal, collision free solution in the feasible-configuration space.

Task time is essential information for the optimal planning and scheduling of industrial scenarios, such as assembly cells. In Human-Robot Collaboration (HRC), the robot execution time, i.e., the robot task time, depends on the task the human is executing simultaneously to the robot and on the human movements. Indeed, the robot may be requested to modify its speed along a predefined path (i.e., to slow down or to stop its motion) in order to avoid possible collisions with the human. This paper presents an approach for the estimation of the robot execution time, when the robot path and the human task are assigned. Specifically, a workspace segmentation is performed considering the volume occupied by the human and the robot during their motion. Then, this segmentation is exploited for the definition of a set of Markov chains modeling human-robot interaction and allowing the estimation of the robot execution time. Simulated and real test beds are presented and discussed.

In the context of sensor-based human-robot interaction, a particularly promising solution is represented by myoelectric control schemes based on synergy-derived signals. We developed and tested on healthy subjects a synergy-based control to achieve simultaneous, continuous actuation of three degrees of freedom of a humanoid robot, while performing functional reach-to-grasp movements. The control scheme exploits subject-specific muscle synergies extracted from eleven upper limb muscles through an easy semi-supervised calibration phase, and computes online activation coefficients to actuate the robot joints. The humanoid robot was able to well reproduce the subjects’ motion, which consisted in free multi-degree-of-freedom reach-to-grasp movements at self-paced speeds. Furthermore, the synergy-based online control significantly outperformed a traditional muscle-pair approach (that uses a pair of antagonist muscles for each joint), in terms of decreased error, increased correlation, and peak correlation between the subjects’ and the robot’s joint angles. The delay introduced by the two algorithms was comparable. This work is a proof-of-concept for an intuitive and robust myocontrol interface, without the need for any training and practice. It has several potential applications, especially for functional assistive engaging devices in children with social and motor impairments.

Aquatic exercises have been demonstrated to benefit the facilitation of motor recovery and the enhancement of well-being in middle-aged adults and the elderly. Personalization, immersiveness, and biofeedback are key for amplifying and accelerating any rehabilitation process in neurological and orthopedic patients. However, a therapist can neither properly visualize nor monitor rehabilitation exercises executed under water, nor can he/she measure them. Therefore, this paper aims to provide adaptive biofeedback during aquatic exercises in order to enhance the training’s effectiveness. A wearable biofeedback suit equipped with wearable underwater-resistant sensor nodes has been designed, produced, and tested. A dedicated algorithm for quantitatively extracting joint angles has been developed and validated against the optical tracking system. Multiple biofeedback modalities are proposed based on visual feedback: amplitude control with set target angles; velocity-amplitude control with set target angles and angular velocity; and velocity tutor with set target angles, a frequency value, and a rest period. Joint angles estimated using the sensor network are compared to those estimated using an optical tracking system with the root-mean-squared angle error between the two systems ranging from 4.0° to 6.3° and a significant correlation coefficient that is always greater than 0.99. Pilot tests during aquatic exercises executed in a thermal environment demonstrate the feasibility and usability of the complete system in the final working environment. The relevant angles are correctly calculated and monitored online during the exercises, and the tested subjects understand the implemented biofeedback modalities easily and follow them well as the SUS evaluation indicates.

Handling and assembling applications with small batch size and high production mix require requires high adaptability, reconfigurability and flexibility. Thus, human-robot collaboration could be an effective solution to ensure production performance and operator satisfaction. This scenario requires human-awareness in different levels of the software framework, from the robot control to the task planning. The goal is to assign high added value activities to the human as much as possible, while the robot has to be able to substitute the human when needed. Team PIROS faces this goal by designing a IEC 61499/ROS-based architecture which integrate safety assessment, advanced force control, human-aware motion planning, gesture recognition, and task scheduling.

International audience; Redundant manipulators are usually required to perform tasks in the operational space, but collision-free path planning is computed in the configuration space. Limiting the deviation with respect to the collision-free configuration-space trajectory may allow the robot to avoid collisions without modifying the primary task. This paper proposes a method to guarantee that the solution of the inverse kinematic problem deviates from the nominal joint-space trajectory less than a desired threshold. The excursion limitation is ensured by means of linear constraints and the automatic regulation of the weights of secondary tasks. Numerical and experimental results prove the validness of the proposed approach.

We present the development of a novel smart ink pen instrumented with force and motion sensors designed for the quantitative and ecological assessment of daily-life handwriting. This work aims at testing the pen’s sensors and algorithms and the use of the smartpen to detect age-related changes in writing and tremor parameters during daily-life handwriting. A comparison against reference instruments was carried out to validate the pen tip force during static and dynamic conditions, the pen’s tilt angle, and the algorithm for the segmentation of the force signal into strokes. The smartpen was tested on 43 healthy adults divided into three age groups (young, middle-old, and old) during unconstrained handwriting on paper. The validation of the pen’s sensors and algorithms reported excellent results. A solid test–retest reliability was found in the writing and tremor indicators extracted from both young and old adults. As for the age-related analysis, the older age groups were characterized by an increase of temporal writing measures, a more uniform writing pressure, and more repetitive and predictable tremor oscillation components. The greatest accomplishment of our smart ink pen is the ability to combine the advantages of the digitizing tablet technology with the natural “feel,” the ease of use, and the ecological validity of the traditional pen-and-paper approach. The proposed solution represents an important step toward a simple, ecologically valid, reliable quantitative assessment of daily-life handwriting. Our smartpen is particularly suitable for daily-life telemonitoring applications in a number of important health-related fields.

Within the Industry 4.0 context, industrial robots need to show increasing autonomy. The manipulator has to be able to react to uncertainties/changes in the working environment, displaying a robust behavior. In this paper, a control framework is proposed to perform industrial interaction tasks in uncertain working scenes. The proposed methodology relies on two components: i) a 6D pose estimation algorithm aiming to recognize large and featureless parts; ii) a variable damping impedance controller (inner loop) enhanced by an adaptive saturation PI (outer loop) for high accuracy force control (i.e., zero steady-state force error and force overshoots avoidance). The proposed methodology allows to be robust w.r.t. task uncertainties (i.e. , positioning errors and interaction dynamics). The proposed approach has been evaluated in an assembly task of a side-wall panel to be installed inside the aircraft cabin. As a test platform, the KUKA iiwa 14 R820 has been used together with the Microsoft Kinect 2.0 as RGB-D sensor. Experiments show the reliability in the 6D pose estimation and the high-performance in the force-tracking task, avoiding force overshoots while achieving the tracking of the reference force.

Industry 4.0 is taking human-robot collaboration at the center of the production environment. Collaborative robots enhance productivity and flexibility while reducing human’s fatigue and the risk of injuries, exploiting advanced control methodologies. However, there is a lack of real-time model-based controllers accounting for the complex human-robot interaction dynamics. With this aim, this paper proposes a Model-Based Reinforcement Learning (MBRL) variable impedance controller to assist human operators in collaborative tasks. More in details, an ensemble of Artificial Neural Networks (ANNs) is used to learn a human-robot interaction dynamic model, capturing uncertainties. Such a learned model is kept updated during collaborative tasks execution. In addition, the learned model is used by a Model Predictive Controller (MPC) with Cross-Entropy Method (CEM). The aim of the MPC+CEM is to online optimize the stiffness and damping impedance control parameters minimizing the human effort (i.e, minimizing the human-robot interaction forces). The proposed approach has been validated through an experimental procedure. A lifting task has been considered as the reference validation application (weight of the manipulated part: 10 kg unknown to the robot controller). A KUKA LBR iiwa 14 R820 has been used as a test platform. Qualitative performance (i.e, questionnaire on perceived collaboration) have been evaluated. Achieved results have been compared with previous developed offline model-free optimized controllers and with the robot manual guidance controller. The proposed MBRL variable impedance controller shows improved human-robot collaboration. The proposed controller is capable to actively assist the human in the target task, compensating for the unknown part weight. The human-robot interaction dynamic model has been trained with a few initial experiments (30 initial experiments). In addition, the possibility to keep the learning of the human-robot interaction dynamics active allows accounting for the adaptation of human motor system.

Upper-limb exoskeletons provide high-intensity, repetitive, task-specific, interactive and individualized training, making effective use of neuroplasticity for functional recovery in neurological patients. Most exoskeletons have robot axes aligned with the anatomical axes of the subject and provide direct control of individual joints. Recently, novel mechanical structures and actuation mechanisms have been proposed, but still result in bulky and heavy exoskeletons, limiting their applicability into clinical practice. Technological efforts are needed to promote light and wearable exoskeletons that implement active-assistive controllers, providing “assisted-as-needed” rehabilitation therapy, towards patient’s motivation and self-esteem. An overview of upper-limb exoskeletons, including mechanical design and control algorithms, will be provided. Special focus will be put on the current evidence about the efficacy of wearable robotic technologies on motor recovery and about other therapies that can be combined with exoskeletons to improve their therapeutic effects.

Human-robot cooperation is increasingly demanded in industrial applications. Many tasks require the robot to enhance the capabilities of humans. In this scenario, safety also plays an important role in avoiding any accident involving humans, robots, and the environment. With this aim, the paper proposes a cooperative fuzzy-impedance control with embedded safety rules to assist human operators in heavy industrial applications while manipulating unknown weight parts. The proposed methodology is composed by four main components: (i) an inner Cartesian impedance controller (to achieve the compliant robot behavior), (ii) an outer fuzzy controller (to provide the assistance to the human operator), (iii) embedded safety rules (to limit force/velocity during the human-robot interaction enhancing safety), and (iv) a neural network approach (to optimize the control parameters for the human-robot collaboration on the basis of the target indexes of assistance performance defined for this purpose). The main achieved result refers to the capability of the controller to deal with uncertain payloads while assisting and empowering the human operator, both embedding in the controller safety features at force and velocity levels and minimizing the proposed performance indexes. The effectiveness of the proposed approach is verified with a KUKA iiwa 14 R820 manipulator in an experimental procedure where human subjects evaluate the robot performance in a collaborative lifting task of a 10 kg part.

Socially assistive robots have shown potential benefits in therapy of child and elderly patients with social and cognitive deficits. In particular, for autistic children, humanoid robots could enhance engagement and attention, thanks to their simplified toy-like appearance and the reduced set of possible movements and expressions. The recent focus on autism-related motor impairments has increased the interest on developing new robotic tools aimed at improving not only the social capabilities but also the motor skills of autistic children. To this purpose, we have designed two embodied mirroring setups using the NAO humanoid robot. Two different tracking systems were used and compared: Inertial Measurement Units and the Microsoft Kinect, a marker-less vision based system. Both platforms were able to mirror upper limb basic movements of two healthy subjects, an adult and a child. However, despite the lower accuracy, the Kinect-based setup was chosen as the best candidate for embodied mirroring in autism treatment, thanks to the lower intrusiveness and reduced setup time. A prototype of an interactive mirroring game was developed and successfully tested with the Kinect-based platform, paving the way to the development of a versatile and powerful tool for clinical use with autistic children.

The brain is provided with an enormous computing capability and exploits neural plasticity to store and elaborate complex information. One of the multiple mechanisms that neural circuits express is the Spike-timing-dependent plasticity (STDP), a form of long-term synaptic plasticity exploiting the time relationship between pre- and post-synaptic action potentials (i.e., neuron spikes). It has been found that in certain cases, for instance at the input stage of the cerebellum, between mossy fibers and granular neurons, the plasticity is not only driven by the timing of the spikes, but also by the oscillation frequency of the inputs. This complex behaviour has been implemented in this work, where we developed a novel form of advanced synaptic plasticity model to be used in a well-established neural network simulator (NEST). The subsequent tests proved the proper functioning of the plasticity and its range of applicability, demonstrating the possibility to adopt it in noisy and variable conditions, similar to the biological settings.

Autism spectrum disorder (ASD) is a multifaceted neurodevelopmental disorder characterized by a persistence of social and communication impairment, and restricted and repetitive behaviors. However, motor disorders have also been described, but not objectively assessed. Most studies showed inefficient eye-hand coordination and motor learning in children with ASD; in other experiments, mechanisms of acquisition of internal models in self-generated movements appeared to be normal in autism. In this framework, we have developed a robotic protocol, recording gaze and hand data during upper limb tasks, in which a haptic pen-like handle is moved along specific trajectories displayed on the screen. The protocol includes trials of reaching under a perturbing force field and catching moving targets, with or without visual availability of the whole path. We acquired 16 typically-developing scholar-age children and one child with ASD as a case study. Speed-accuracy tradeoff, motor performance, and gaze-hand spatial coordination have been evaluated. Compared to typically developing peers, in the force field sequence, the child with ASD showed an intact but delayed learning, and more variable gazehand patterns. In the catching trials, he showed less efficient movements, but an intact capability of exploiting the available a-priori plan. The proposed protocol represents a powerful tool, easily tunable, for quantitative (longitudinal) assessment, and for subject-tailored training in ASD.

Switching between contexts affects the mechanisms underlying motion planning, in particular it may entail reranking the variables to be controlled in defining the motor solutions. Three astronauts performed multiple sessions of whole-body pointing, in normogravity before launch, in prolonged weightlessness onboard the International Space Station, and after return. The effect of gravity context on kinematic and dynamic components was evaluated. Hand trajectory was gravity independent; center-of-mass excursion was highly variable within and between subjects. The body-environment effort exchange, expressed as inertial ankle momentum, was systematically lower in weightlessness than in normogravity. After return on Earth, the system underwent a rapid 1-week readaptation. The study indicates that minimizing the control effort is given greater weight when optimizing the motor plan in weightlessness compared to normogravity: the hierarchies of the controlled variables are gravity dependent.

A bioinspired adaptive model, developed by means of a spiking neural network made of thousands of artificial neurons, has been leveraged to control a humanoid NAO robot in real time. The learning properties of the system have been challenged in a classic cerebellum-driven paradigm, a perturbed upper limb reaching protocol. The neurophysiological principles used to develop the model succeeded in driving an adaptive motor control protocol with baseline, acquisition, and extinction phases. The spiking neural network model showed learning behaviours similar to the ones experimentally measured with human subjects in the same task in the acquisition phase, while resorted to other strategies in the extinction phase. The model processed in real-time external inputs, encoded as spikes, and the generated spiking activity of its output neurons was decoded, in order to provide the proper correction on the motor actuators. Three bidirectional long-term plasticity rules have been embedded for different connections and with different time scales. The plasticities shaped the firing activity of the output layer neurons of the network. In the perturbed upper limb reaching protocol, the neurorobot successfully learned how to compensate for the external perturbation generating an appropriate correction. Therefore, the spiking cerebellar model was able to reproduce in the robotic platform how biological systems deal with external sources of error, in both ideal and real (noisy) environments.

Objective: To develop and evaluate a hybrid robotic system for arm recovery after stroke, combining ElectroMyoGraphic (EMG)-triggered functional electrical stimulation (FES) with a passive exoskeleton for upper limb suspension. Methods: The system was used in a structured exercise program resembling activities of daily life. Exercises execution was continuously controlled using angle sensor data and radio-frequency identification technology. The training program consisted of 27 sessions lasting 30 min each. Seven post-acute stroke patients were recruited from two clinical sites. The efficacy of the system was evaluated in terms of action research arm test, motricity index, motor activity log, and box & blocks tests. Furthermore, kinematics-based and EMG-based outcome measures were derived directly from data collected during training sessions. Results: All patients showed an improvement of motor functions at the end of the training program. After training, the exercises were in most cases executed faster, smoother, and with an increased range of motion. Subjects were able to trigger FES, but in some cases, they did not maintain the voluntary effort during task execution. All subjects but one considered the system usable. Conclusion: The preliminary results showed that the system can be used in a clinical environment with positive effects on arm functional recovery. However, only the final results of the currently ongoing clinical trial will unveil the system's full potential. Significance: The presented hybrid robotic system is highly customizable, allows to monitor the daily performance, requires low supervision of the therapist, and might have the potential to enhance arm recovery after stroke.

Sensorimotor signals are integrated and processed by the cerebellar circuit to predict accurate control of actions. In order to investigate how single neuron dynamics and geometrical modular connectivity affect cerebellar processing, we have built an olivocerebellar Spiking Neural Network (SNN) based on a novel simplification algorithm for single point models (Extended Generalized Leaky Integrate and Fire, EGLIF) capturing essential non-linear neuronal dynamics (e.g., pacemaking, bursting, adaptation, oscillation and resonance). EGLIF models specifically tuned for each neuron type were embedded into an olivocerebellar scaffold reproducing realistic spatial organization and physiological convergence and divergence ratios of connections. In order to emulate the circuit involved in an eye blink response to two associated stimuli, we modeled two adjacent olivocerebellar microcomplexes with a common mossy fiber input but different climbing fiber inputs (either on or off). EGLIF-SNN model simulations revealed the emergence of fundamental response properties in Purkinje cells (burst-pause) and deep nuclei cells (pause-burst) similar to those reported in vivo. The expression of these properties depended on the specific activation of climbing fibers in the microcomplexes and did not emerge with scaffold models using simplified point neurons. This result supports the importance of embedding SNNs with realistic neuronal dynamics and appropriate connectivity and anticipates the scale-up of EGLIF-SNN and the embedding of plasticity rules required to investigate cerebellar functioning at multiple scales.

Programming is a creative activity that teaches precision. In Logo, novices write simple programs that draw geometric shapes onto a screen. Logical flaws, however, cause unintended results and pose a major challenge for young programmers who yet need to learn how to search for errors in their code. We discuss the problems novices face when learning to program in Logo. Furthermore, we present a reverse debugger for Logo that enables programmers to step through their code in either direction. Using a stack, previous program states can be retrieved on demand. Our solution balances performance and memory consumption and hence can be used to debug even long and complex programs.

International audience; Human robot collaboration requires new planning strategies to guarantee an efficient and safe coexistence of robots and humans in the workspace. We propose a framework based on a model predictive control approach to trajectory scaling and inverse kinematics. The online modification of the velocity override slows down the task to ensure safety and the redundancy of the system is exploited to maximize the distance from the operator. Experimental results on a 7-degree-of-freedom robotic system prove the effectiveness of the method.

Smart factories must speed up their processes to face new manufacturing challenges and, at the same time, demonstrate an extremely high degree of flexibility to reduce production costs and time. This kind of issues can be addressed by the cooperation between humans and robots in a mixed human-robot working environment. Robots have the compelling advantage of spatial precision and repeatability as well as the capability of applying defined forces. Humans, on the other hand, are especially skilled at complex manipulations and adapting to changing task requirements. In this complicate scenario of co-shared workplace and continuous human-robot interaction, safety strategies are a key requirement to avoid possible injuries to humans or fatal accidents. This chapter proposes a systemic approach to respond to these requirements. The approach merges and manages multiple sensing sources, redundant transmission protocols and software decision mechanisms, aiming to guarantee a continuous and reconfigurable co-share scenario that enables an operative interaction between human workers and robots in a controlled and safe environment. Furthermore, new technological solutions and innovative methodologies are presented for the definition of a safer workplace in human-robot interaction scenarios.

Within the scope of the project EURECA (https://cordis.europa.eu/project/id/738039), the paper describes a human-robot cooperative installation methodology of heavy and bulky components based on marker-based visual servoing, force control, and human-robot cooperation. The main advance in the human-robot cooperation is achieved by a shared-control of the interaction during the in- stallation task, relieving the human operator by the manipulated load and giving to the robot a partially autonomous behaviour in the force-tracking direction. Experimental results are shown in the context of the H2020 CleanSky 2 EURECA project in which a side-wall panel is installed in a 1:1 scale mock-up scenario of an A320 plane fuselage environment.

International audience; The paper presents a fast online predictive method to solve the task-priority differential inverse kinematics of redundant manipula-tors under kinematic constraints. It implements a task-scaling technique to preserve the desired geometrical task, when the trajectory is infeasible for the robot capabilities. Simulation results demonstrate the effectiveness of the methodology.

The aim of this study concerns the evaluation and comparison of different Human-Machine Interfaces for the control of an upper limb motorized exoskeleton for severely impaired patients. Different approaches (i.e. manual, vocal, visual control) are tested in a simulation environment on three subjects affected by muscular dystrophy with the aim of assessing the capability of the system to interact with the user and vice versa. A Graphical User Interface shows the simulated behavior of the exoskeleton to the user which has to perform reaching tasks in the space by moving the exoskeleton end-effector to defined virtual targets that are displayed on the screen. Specific assessment of the interaction of the user with each control interface is achieved, while a quantitative evaluation of the usability of all the three approaches is provided by a System Usability Scale (SUS) questionnaire. All patients were able to interact with all control interfaces without difficulties and to complete reaching tasks in simulation. SUS scores showed overall good usability of the Human-Machine Control Interfaces suggesting that the manual and the vocal control interfaces are preferred by the subjects.

Recent advances in planning and control of robot manipulators make an increasing use of optimization-based techniques, such as model predictive control. In this framework, ensuring the feasibility of the online optimal control problem is a key issue. In the case of manipulators with bounded joint positions, velocities, and accelerations, feasibility can be guaranteed by limiting the set of admissible velocities and positions to a viable set. However, this results in the imposition of nonlinear optimization constraints. In this paper, we analyze the feasibility of the optimal control problem and we propose a method to construct a viable convex polyhedral that ensures feasibility of the optimal control problem by means of a given number of linear constraints. Experimental and numerical results on an industrial manipulator show the validity of the proposed approach.

During natural learning, synaptic plasticity is thought to evolve dynamically and redistribute within and among subcircuits. This process should emerge in plastic neural networks evolving under behavioral feedback and should involve changes distributed across multiple synaptic sites. In eyeblink classical conditioning (EBCC), the cerebellum learns to predict the precise timing between two stimuli, hence EBCC represents an elementary yet meaningful paradigm to investigate the cerebellar network functioning. We have simulated EBCC mechanisms by reconstructing a realistic cerebellar microcircuit model and embedding multiple plasticity rules imitating those revealed experimentally. The model was tuned to fit experimental EBCC human data, estimating the underlying learning time-constants. Learning started rapidly with plastic changes in the cerebellar cortex followed by slower changes in the deep cerebellar nuclei. This process was characterized by differential development of long-term potentiation and depression at individual synapses, with a progressive accumulation of plasticity distributed over the whole network. The experimental data included two EBCC sessions interleaved by a trans-cranial magnetic stimulation (TMS). The experimental and the model response data were not significantly different in each learning phase, and the model goodness-of-fit was [Formula: see text] for all the experimental conditions. The models fitted on TMS data revealed a slowed down re-acquisition (sessions-2) compared to the control condition ([Formula: see text]). The plasticity parameters characterizing each model significantly differ among conditions, and thus mechanistically explain these response changes. Importantly, the model was able to capture the alteration in EBCC consolidation caused by TMS and showed that TMS affected plasticity at cortical synapses thereby altering the fast learning phase. This, secondarily, also affected plasticity in deep cerebellar nuclei altering learning dynamics in the entire sensory-motor loop. This observation reveals dynamic redistribution of changes over the entire network and suggests how TMS affects local circuit computation and memory processing in the cerebellum.

Stroke-related locomotor impairments are often associated with abnormal timing and intensity of recruitment of the affected and non-affected lower limb muscles. Restoring the proper lower limbs muscles activation is a key factor to facilitate recovery of gait capacity and performance, and to reduce maladaptive plasticity. Ekso is a wearable powered exoskeleton robot able to support over-ground gait training. The user controls the exoskeleton by triggering each single step during the gait cycle. The fine-tuning of the exoskeleton control system is crucial-it is set according to the residual functional abilities of the patient, and it needs to ensure lower limbs powered gait to be the most physiological as possible. This work focuses on the definition of an automatic calibration procedure able to detect the best Ekso setting for each patient. EMG activity has been recorded from Tibialis Anterior, Soleus, Rectus Femoris, and Semitendinosus muscles in a group of 7 healthy controls and 13 neurological patients. EMG signals have been processed so to obtain muscles activation patterns. The mean muscular activation pattern derived from the controls cohort has been set as reference. The developed automatic calibration procedure requires the patient to perform overground walking trials supported by the exoskeleton while changing parameters setting. The Gait Metric index is calculated for each trial, where the closer the performance is to the normative muscular activation pattern, in terms of both relative amplitude and timing, the higher the Gait Metric index is. The trial with the best Gait Metric index corresponds to the best parameters set. It has to be noted that the automatic computational calibration procedure is based on the same number of overground walking trials, and the same experimental set-up as in the current manual calibration procedure. The proposed approach allows supporting the rehabilitation team in the setting procedure. It has been demonstrated to be robust, and to be in agreement with the current gold standard (i.e., manual calibration performed by an expert engineer). The use of a graphical user interface is a promising tool for the effective use of an automatic procedure in a clinical context.

Two binding requirements for in vitro studies on long-term neuronal networks dynamics are (i) finely controlled environmental conditions to keep neuronal cultures viable and provide reliable data for more than a few hours and (ii) parallel operation on multiple neuronal cultures to shorten experimental time scales and enhance data reproducibility. In order to fulfill these needs with a Microelectrode Arrays (MEA)-based system, we designed a stand-alone device that permits to uninterruptedly monitor neuronal cultures activity over long periods, overcoming drawbacks of existing MEA platforms. We integrated in a single device: (i) a closed chamber housing four MEAs equipped with access for chemical manipulations, (ii) environmental control systems and embedded sensors to reproduce and remotely monitor the standard in vitro culture environment on the lab bench (i.e. in terms of temperature, air CO2 and relative humidity), and (iii) a modular MEA interface analog front-end for reliable and parallel recordings. The system has been proven to assure environmental conditions stable, physiological and homogeneos across different cultures. Prolonged recordings (up to 10 days) of spontaneous and pharmacologically stimulated neuronal culture activity have not shown signs of rundown thanks to the environmental stability and have not required to withdraw the cells from the chamber for culture medium manipulations. This system represents an effective MEA-based solution to elucidate neuronal network phenomena with slow dynamics, such as long-term plasticity, effects of chronic pharmacological stimulations or late-onset pathological mechanisms.