We present DemoStart, a novel auto-curriculum reinforcement learning method capable of learning complex manipulation behaviors on an arm equipped with a three- fingered robotic hand, from only a sparse reward and a handful of demonstrations in simulation. Learning from simulation drastically reduces the development cycle of behavior generation, and domain randomization techniques are leveraged to achieve successful zero-shot sim-to- real transfer. Transferred policies are learned directly from raw pixels from multiple cameras and robot proprioception. Our approach outperforms policies learned from demonstrations on the real robot and requires 100 times fewer demonstrations, collected in simulation. More details and videos in sites.google.com/view/demostart.

Reinforcement learning solely from an agent’s self-generated data is often believed to be infeasible for learning on real robots, due to the amount of data needed. However, if done right, agents learning from real data can be surprisingly efficient through re-using previously collected sub-optimal data. In this paper we demonstrate how the increased understanding of off-policy learning methods and their embedding in an iterative online/offline scheme ("collect and infer") can drastically improve data-efficiency by using all the collected experience, which empowers learning from real robot experience only. Moreover, the resulting policy improves significantly over the state of the art on a recently proposed real robot manipulation benchmark. Our approach learns end-to-end, directly from pixels, and does not rely on additional human domain knowledge such as a simulator or demonstrations.

In this paper, we describe the design of the Barkour benchmark for measuring robot agility in navigating complex environments. Despite the growing interest in developing agile robot locomotion skills, the field lacks systematic benchmarks to measure the performance of robotic control systems and hardware in agility-focused tasks. This motivated us to propose the Barkour benchmark, an obstacle course designed to quantify agility across various robotic platforms. Inspired by dog agility competitions, the course features diverse obstacles and a time-based scoring mechanism, encouraging researchers to develop controllers that enable robots to move quickly, precisely, and with adaptability. This benchmark is challenging as it demands diverse motion skills and the time-based scoring requires control precision at high speed. Along with the design details presented in the paper, we release our simulated environment setups in MuJoCo-XLA and the CAD model of a custom-designed quadruped robot to facilitate future research to reproduce the Barkour setup (available at sites.google.com/view/barkour). We hope these together will accelerate the pace of robot agility research.

We present a system for applying sim2real approaches to “in the wild” scenes with realistic visuals, and to policies which rely on active perception using RGB cameras. Given a short video of a static scene collected using a generic phone, we learn the scene's contact geometry and a function for novel view synthesis using a Neural Radiance Field (NeRF). We augment the NeRF rendering of the static scene by overlaying the rendering of other dynamic objects (e.g. the robot's own body, a ball). A simulation is then created using the rendering engine in a physics simulator which computes contact dynamics from the static scene geometry (estimated from the NeRF vol-ume density) and the dynamic objects' geometry and physical properties (assumed known). We demonstrate that we can use this simulation to learn vision-based whole body navigation and ball pushing policies for a 20 degree-of-freedom humanoid robot with an actuated head-mounted RGB camera, and we successfully transfer these policies to a real robot.

Overheadwork is classifiedas one of the major risk factors for the onset of shoulder work-related musculoskeletal disorders and muscle fatigue. Upper-limb exoskeletons can be used to assist workers during the execution of industrial overhead tasks to prevent such disorders. Twelve novice participants have been equipped with inertial and force/torque sensors to simultaneously estimate the whole-body kinematics and the joint torques (i.e., internal articular stress) by means of a probabilistic estimator, while performing an overhead task with a pointing tool. An evaluation has been performed to analyze the effect at the whole-body level by considering the conditions of wearing and not-wearing PAEXO, a passive exoskeleton for upper-limb support during overhead work. Results point out that PAEXO provides a reduction of the whole-body joint effort across the experimental task blocks (from 66% to 86%). Moreover, the analysis along with five different body areas shows that 1) the exoskeleton provides support at the human shoulders by reducing the joint effort at the targeted limbs, and 2) that part of the internal wrenches is intuitively transferred from the upper body to the thighs and legs, which is shown with an increment of the torques at the legs joints. The promising outcomes show that the probabilistic estimation algorithm can be used as a validation metric to quantitatively assess PAEXO performances, paving thus the way for the next challenging milestone, such as the optimization of the human joint torques via adaptive exoskeleton control.

Collecting and automatically obtaining reward signals from real robotic visual data for the purposes of training reinforcement learning algorithms can be quite challenging and time-consuming. Methods for utilizing unlabeled data can have a huge potential to further accelerate robotic learning. We consider here the problem of performing manipulation tasks from pixels. In such tasks, choosing an appropriate state representation is crucial for planning and control. This is even more relevant with real images where noise, occlusions and resolution affect the accuracy and reliability of state estimation. In this work, we learn a latent state representation implicitly with deep reinforcement learning in simulation, and then adapt it to the real domain using unlabeled real robot data. We propose to do so by optimizing sequence-based self- supervised objectives. These use the temporal nature of robot experience, and can be common in both the simulated and real domains, without assuming any alignment of underlying states in simulated and unlabeled real images. We further propose a novel such objective, the Contrastive Forward Dynamics loss, which combines dynamics model learning with time-contrastive techniques. The learned state representation that results from our methods can be used to robustly solve a manipulation task in simulation and to successfully transfer the learned skill on a real system. We demonstrate the effectiveness of our approaches by training a vision-based reinforcement learning agent for cube stacking. Agents trained with our method, using only 5 hours of unlabeled real robot data for adaptation, shows a clear improvement over domain randomization, and standard visual domain adaptation techniques for sim-to-real transfer.

It is well known that sensors using strain gauges have a potential dependency on temperature. This creates temperature drift in the measurements of six axis force torque sensors (F/T). The temperature drift can be considerable if an experiment is long or the environmental conditions are different from when the calibration of the sensor was performed. Other in situ methods disregard the effect of temperature on the sensor measurements. Experiments performed using the humanoid robot platform iCub show that the effect of temperature is relevant. The model based in situ calibration of six axis force torque sensors method is extended to perform temperature compensation.

A common approach to the generation of walking patterns for humanoid robots consists in adopting a layered control architecture. This paper proposes an architecture composed of three nested control loops. The outer loop exploits a robot kinematic model to plan the footstep positions. In the mid layer, a predictive controller generates a Center of Mass trajectory according to the well-known table-cart model. Through a whole-body inverse kinematics algorithm, we can define joint references for position controlled walking. The outcomes of these two loops are then interpreted as inputs of a stack-of-task QP-based torque controller, which represents the inner loop of the presented control architecture. This resulting architecture allows the robot to walk also in torque control, guaranteeing higher level of compliance. Real world experiments have been carried on the humanoid robot iCub.

This paper presents a proof of concept to calibrate iCub's skin using vacuum bags. The main idea of the method consists of inserting the skin, made of distributed capacitive sensors, in a vacuum bag and then decreasing the pressure in the bag to create a uniform pressure distribution on the skin surface. The capacitance changes of each sensor were then related to the applied differential pressure using a least square fitting with a fifth order polynomial model. After calibration, integration of the pressure distribution over the skin geometry provides us with the net normal force applied to the skin. Experiments were conducted using the forearm skin of the iCub humanoid robot and the calibration results were validated using standard weights. The validation results indicate acceptable average errors in force prediction.

This work explores the stabilization of desired dynamic motion tasks involving hard impacts at non-negligible speed for humanoid robots. To this end, a so-called reference spreading hybrid control law is designed showing promising results in simulation. The simulations are performed employing a dynamical model of an existing humanoid robot and impacts are assumed to be inelastic. The desired motion task consists of having the robot balancing on one foot while repeatedly making and breaking contact with a wall by means of one hand. The simulation results illustrate that the considered controller is suited to control humanoid robot motions with impacts.

Humanoid robots may require a degree of compliance at joint level for improving efficiency, shock tolerance, and safe interaction with humans. The presence of joint elasticity, however, complexifies the control design of humanoid robots. This paper proposes a control framework to extend momentum based controllers developed for stiff actuation to the case of series elastic actuators. The key point is to consider the motor velocities as an intermediate control input, and then apply high-gain control to stabilise the desired motor velocities achieving momentum control. Simulations carried out on a model of the robot iCub verify the soundness of the proposed approach.

This paper presents a new algorithm for optimal control (OC) of nonlinear dynamical systems. The main feature of this algorithm is that it allows the specification of the control objectives as a hierarchy of tasks, each task representing an action that the robot should perform. Each task is described by a cost function that the algorithm tries to minimize, while not affecting the tasks of higher priority. The concept of strict priority allows for an easier and more robust specification of the control objectives, without hand tuning of task weights. The hierarchy also makes it possible to properly regularize the behavior of each task independently. For the first time, we properly define the problem of regularizing the task cost functions in the presence of a hierarchy and propose an algorithm to compute an approximate solution. Several simulated scenarios with different robots compare our solution with other state-of-the-art methods, validating the interest of the hierarchy in OC and empirically demonstrating the importance of regularization to generate feasible behaviors.

This paper presents a novel approach for incremental semiparametric inverse dynamics learning. In particular, we consider the mixture of two approaches: Parametric modeling based on rigid body dynamics equations and nonparametric modeling based on incremental kernel methods, with no prior information on the mechanical properties of the system. The result is an incremental semiparametric approach, leveraging the advantages of both the parametric and nonparametric models. We validate the proposed technique learning the dynamics of one arm of the iCub humanoid robot.

Envisioned applications for humanoid robots call for the design of balancing and walking controllers. While promising results have been recently achieved, robust and reliable controllers are still a challenge for the control community dealing with humanoid robotics. Momentum-based strategies have proven their effectiveness for controlling humanoids balancing, but the stability analysis of these controllers is still missing. The contribution of this paper is twofold. First, we numerically show that the application of state-of-the-art momentum-based control strategies may lead to unstable zero dynamics. Secondly, we propose simple modifications to the control architecture that avoid instabilities at the zero-dynamics level. Asymptotic stability of the closed loop system is shown by means of a Lyapunov analysis on the linearized system's joint space. The theoretical results are validated with both simulations and experiments on the iCub humanoid robot.

This paper presents a new condition, the fully physical consistency for a set of inertial parameters to determine if they can be generated by a physical rigid body. The proposed condition ensure both the positive definiteness and the triangular inequality of 3D inertia matrices as opposed to existing techniques in which the triangular inequality constraint is ignored. This paper presents also a new parametrization that naturally ensures that the inertial parameters are fully physical consistency. The proposed parametrization is exploited to reformulate the inertial identification problem as a manifold optimization problem, that ensures that the identified parameters can always be generated by a physical body. The proposed optimization problem has been validated with a set of experiments on the iCub humanoid robot.

This paper deals with the generation of motion for complex dynamical systems (such as humanoid robots) to achieve several concurrent objectives. Hierarchy of tasks and optimal control are two frameworks commonly used to this aim. The first one specifies control objectives as a number of quadratic functions to be minimized under strict priorities. The second one minimizes an arbitrary user-defined function of the future state of the system, thus considering its evolution in time. Our recent work on prioritized optimal control merges the advantages of both these methods. This paper reformulates the original prioritized optimal control algorithm with the precise goal of improving its computational speed. We extend the dynamic programming method to work with a hierarchy of tasks. We compared our approach in simulation with both our previous algorithm and classical optimal control. The measured computational improvement represents another step towards the application of prioritized optimal control for online model predictive control of humanoid robots. We believe that this could be the key to unlock the (so far unexploited) dynamic capabilities of these mechanical systems.

Classically robot force control passes through joint torques measurement or estimation. Within this context, classical torque sensing technologies rely on current sensing on motor windings and on torsion sensing on motor shaft. An alternative approach was recently proposed in [1] and combines whole-body distributed 6-axis force/torque (F/T) sensors, gyroscopes, accelerometers and tactile sensors (i.e. artificial skin). A further advantage of this method is that it simultaneously estimates (internal) joint torques and (external) contact forces with no need of joint redesign. As a drawback, the method relies on a model of the robot dynamics, as it consists on reordering the classical recursive Newton-Euler algorithm (RNEA). In this paper we consider the problem of the parametric identification of the robot dynamic model from embedded F/T sensors. We extend recent results on parametric identification [2] by considering an arbitrary reordering of the classical RNEA. The theoretical framework is validated on the iCub humanoid, which is equipped with both 6-axis F/T sensors and joint torque sensors. We estimated the system inertial parameters using only one F/T sensor. We used the obtained parameters to estimate the joint torques (as proposed in [1]) and compared the results with direct joint torque measurements, used in this context only as a ground truth.

This paper proposes techniques to calibrate six-axis force-torque sensors that can be performed in situ, i.e., without removing the sensor from the hosting system. We assume that the force-torque sensor is attached to a rigid body equipped with an accelerometer. Then, the proposed calibration technique uses the measurements of the accelerometer, but requires neither the knowledge of the inertial parameters nor the orientation of the rigid body. The proposed method exploits the geometry induced by the model between the raw measurements of the sensor and the corresponding force-torque. The validation of the approach is performed by calibrating two six-axis force-torque sensors of the iCub humanoid robot.

Towards enhancing the dynamic locomotion and manipulation abilities of bipedal robots in real-world scenarios, a key problem lies in the accurate estimation of the dynamic state of the feet of the robot. In this paper, an approach is presented for estimating the dynamic pose and the internal (body) and external (ground contact) wrenches acting on the individual feet of a bipedal robot fusing haptic (compliant skin), inertial, and force/torque (F/T) measurements. Assuming rigid body dynamics on an individual foot, an Extended Kalman Filter (EKF) is used to combine ankle F/T sensor readings, contact forces computed from a compliant tactile array on the foot sole and accelerometer plus gyroscope measurements, thereby estimating both the state and the external wrenches affecting a foot through a method of state augmentation. Moreover, covariance estimation of the measurement noise was carried out for all sensors, in particular, for the skin, a bayesian-network-based regression method was chosen. The framework was implemented with the iCub humanoid robot under a toppling scenario; the estimated augmented foot state was then used to compute the Foot Rotation Indicator (FRI) trajectory as a validation through prediction of the onset of toppling and instability.

Given an articulated rigid body, we define the problem of estimating its dynamics as the problem of computing all the forces and accelerations acting on the bodies which constitute the articulated system. Similarly, we define the state estimation problem as the problem of computing the system positions and velocities. In the present paper we propose a framework for simultaneous state and dynamics estimation. The estimation is framed in a Bayesian framework and a suitable Bayesian prior is defined to guarantee the physical consistency of the obtained estimation. The Bayesian posterior makes use of all available measurements which include encoders, gyroscopes, accelerometers, force and torque sensors. The proposed theoretical framework is validated both on simulation and on the iCub humanoid. The software that implements the theoretical framework is realised with an open-source license.

Collocated adaptive control of underactuated mechanical systems is still a concern for the control community. The main difficulty comes from the nonlinearity of the collocated inverse dynamics with respect to the base parameters, which forbids the direct application of classical adaptive control schemes. This paper extends and encompasses the Slotine's adaptive control, which was developed for fully actuated mechanical systems, to stabilize the collocated state space of an underactuated mechanical system. The key point is to define the sliding variable as the difference between the system's velocity and an exogenous state whose dynamics is considered as control input. We first revisit the Slotine's result in view of this definition and then show how to extend it to the underactuated case. Stability and convergence of time-varying reference trajectories for the collocated dynamics are shown to be in the sense of Lyapunov. Global well-posedness of the control laws is achieved by means of a new algebraic property of the mass matrix. Simulations, comparisons to existing control strategies, and experimental results on a two-link manipulator verify the soundness of the proposed approach.

Human interaction is based, among other factors, on non verbal and implicit communication. By observing the action of someone else we can automatically infer several non obvious details of what's happening, as the goal of the agent, his mood and even some features of the object he is using, e.g., if it is heavy or light. This action reading skill is developed very early in life and constitutes a fundamental basis for the development of collaboration. A similar capacity to implicitly communicate weight would be desirable also in a humanoid robot, to allow for a natural preparation for hand-over between the robot and the human partner. Here we have investigated how the ability to infer weight from human action observation develops during childhood and whether such ability generalizes to the observation of a humanoid robot. In particular, the robot was not programmed to perform human-like lifting actions, but just to replicate a property of human lifting deemed as determinant for weight reading: exhibiting a velocity proportional to object weight. Our results suggest that although 6-year-olds can already judge weight from the observation of a lifting action, they cannot generalize this skill to simplified robotic actions as the ones proposed here.

This paper presents a new technique to control highly redundant mechanical systems, such as humanoid robots. We take inspiration from two approaches. Prioritized control is a widespread multi-task technique in robotics and animation: tasks have strict priorities and they are satisfied only as long as they do not conflict with any higher-priority task. Optimal control instead formulates an optimization problem whose solution is either a feedback control policy or a feedforward trajectory of control inputs. We introduce strict priorities in multi-task optimal control problems, as an alternative to weighting task errors proportionally to their importance. This ensures the respect of the specified priorities, while avoiding numerical conditioning issues. We compared our approach with both prioritized control and optimal control with tests on a simulated robot with 11 degrees of freedom.

In this paper we tackle the problem of estimating the local compliance of tactile arrays exploiting global measurements from a single force and torque sensor. The proposed procedure exploits a transformation matrix (describing the relative position between the local tactile elements and the global force/torque measurements) to define a linear regression problem on the unknown local stiffness. Experiments have been conducted on the foot of the iCub robot, sensorized with a single force/torque sensor and a tactile array of 250 tactile elements (taxels) on the foot sole. Results show that a simple calibration procedure can be employed to estimate the stiffness parameters of virtual springs over a tactile array and to use these model to predict normal forces exerted on the array based only on the tactile feedback. Leveraging on previous works [1] the proposed procedure does not necessarily need a-priori information on the transformation matrix of the taxels which can be directly estimated from available measurements.

Legged robots are typically in rigid contact with the environment at multiple locations, which add a degree of complexity to their control. We present a method to control the motion and a subset of the contact forces of a floating-base robot. We derive a new formulation of the lexicographic optimization problem typically arising in multi-task motion/force control frameworks. The structure of the constraints of the problem (i.e. the dynamics of the robot) allows us to find a sparse analytical solution. This leads to an equivalent optimization with reduced computational complexity, comparable to inverse-dynamics based approaches. At the same time, our method preserves the flexibility of optimization based control frameworks. Simulations were carried out to achieve different multi-contact behaviors on a 23-degree-of-freedom humanoid robot, validating the presented approach. A comparison with another state-of-the-art control technique with similar computational complexity shows the benefits of our controller, which can eliminate force/torque discontinuities.

In this paper we propose a methodology to control a novel class of actuators that we called passive noise rejection variable stiffness actuators (pnrVSA). Differently from nowadays classical VSA designs, this novel class of actuators mimics the human musculoskeletal ability to increase noise rejection without relying on feedback. To fully highlight the potentialities behind these actuators we consider movement planning under two constraints: (1) absence of feedback, i.e. purely open-loop planning1; (2) uncertain dynamic model. Under these constraints, movement planning can be formalized as an open-loop stochastic optimal control. Due to the lack of classical methods forcing the open-loop nature of the computed solution, we used here a slight modification of available methodologies based on importance sampling of trajectories using forward diffusion processes. Simulations show that the proposed algorithm can be effectively used to plan open-loop movements with pnrVSA. In particular, two different scenarios are considered: the control of a single joint pnrVSA and the control of a two degrees of freedom planar arm equipped with antagonist pnrVSAs at each joint. In both cases, movement has to be planned in presence of uncertain dynamics for unstable tasks. It is shown that open-loop stochastic optimal control can modulate the intrinsic stiffness of the system to cope with both instability and noise.

In this paper we describe the mathematical model of a tendon driven eye. We focus on describing its movements, posing a specific attention on cyclotorsion, that is the rotation around the eye optical axis. This study aims at understanding the cause of a cyclotorsion effect that has been noticed on a real setup. The paper starts from a qualitative analysis of the effect. Then, it proposes two different models for its motion. Finally both models are validated by comparing their predictions with the outcomes on the real robot. Given the complexity of the system and of its motion results are for the moment just qualitative and quantitative comparison will be the goal of our future works.

This paper investigates the role of precise estimation of contact points in force control. This analysis is motivated by scenarios in which robots make contacts, either voluntarily or accidentally, with different parts of their body. Control paradigms that are usually implemented in robots with no tactile system, make the hypothesis that contacts occur at the end-effectors only. In this paper we try to investigate what happens when this assumption is not verified. First we consider a simple feedforward force control law, and then we extend it by introducing a proportional feedback term. For both controllers we find the error in the resulting contact force, that is induced by a hypothetic error in the estimation of the contact point. We show that, depending on the geometry of the contact, incorrect estimation of contact points can induce undesired joint accelerations. We validate the presented analysis with tests on a simulated robot arm. Moreover we consider a complex real world scenario, where most of the assumptions that we make in our analytical derivation do not hold. Through tests on the iCub humanoid robot we see how errors in contact localization affect the performance of a parallel force/position controller. In order to estimate contact points and contact forces on the forearm of the iCub we do not use any model of the environment, but we exploit its 6-axis force/torque sensor and its sensorized skin.

Muscle co-contraction can be modeled as an active modulation of the passive musculo-skeletal compliance. Within this context, recent findings in human motor control have shown that active compliance modulation is fundamental when planning movements in presence of unpredictability and uncertainties. Along this line of research, this paper investigates the link between active impedance control and unpredictability, with special focus on robotic applications. Different types of actuators are considered and confronted to extreme situations such as moving in an unstable force field and controlling a system with significant delays in the feedback loop. We use tools from stochastic optimal control to illustrate the possibility of optimally planning the intrinsic system stiffness when performing movements in such situations. In the extreme case of total feedback absence, different actuators model are considered and their performance in dealing with unpredictability compared. Finally, an application of the proposed theories on planning reaching movements with the iCub humanoid platform is proposed.

Visual motion is a simple yet powerful cue widely used by biological systems to improve their perception and adaptation to the environment. Examples of tasks that greatly benefit from the ability to detect movement are object segmentation, 3D scene reconstruction and control of attention. In computer vision several algorithms for computing visual motion and optic flow exist. However their application in robotics is not straightforward as in these platforms visual motion is often dominated by (self) motion produced by the movement of the robot (egomotion) making it difficult to disambiguate between motion induced by the scene dynamics or by the own actions of the robot. Independent motion detection is an active field in computer vision and robotics, however approaches in this area typically require that some models of both the environment and the robot visual system are available and are hardly suitable for real-time control. In this paper we describe the motionCUT, a derivation of the Lucas-Kanade optical flow algorithm that allows detecting moving objects, irrespectively of the egomotion produced by the robot. Our method is purely visual and does not require information other than the images coming from the cameras. As such it can be easily adapted to any robotic platform. The system was tested on a stereo tracking task on the iCub humanoid robot, demonstrating that the algorithm performs well and can easily execute in real-time.

This paper deals with the problem of estimating the position of tactile elements (i.e. taxels) that are mounted on a robot body part. This problem arises with the adoption of tactile systems with a large number of sensors, and it is particularly critical in those cases in which the system is made of flexible material that is deployed on a curved surface. In this scenario the location of each taxel is partially unknown and difficult to determine manually. Placing the device is in fact an inaccurate procedure that is affected by displacements in both position and orientation. Our approach is based on the idea that it is possible to automatically infer the position of the taxels by measuring the interaction forces exchanged between the sensorized part and the environment. The location of the contact is estimated through force/torque (F/T) measures gathered by a sensor mounted on the kinematic chain of the robot. Our method requires few hypotheses and can be effectively implemented on a real platform, as demonstrated by the experiments with the iCub humanoid robot.

When a robot is required to safely interact with a physical environment, two approaches are typically reported in literature: using a force/torque sensor to regulate the interaction forces at the end effector, or integrating sensors in each robot joint to regulate their torques. In this paper we want to discuss the benefits and the disadvantages of the two approaches, showing a direct comparison between the information which can be obtained from the two categories of sensors. Results obtained on the new iCub arm, which integrates torque sensing capabilities at joint level will be presented and discussed.

An important issue in motor control is understanding the basic principles underlying the accomplishment of natural movements. According to optimal control theory, the problem can be stated in these terms: what cost function do we optimize to coordinate the many more degrees of freedom than necessary to fulfill a specific motor goal? This question has not received a final answer yet, since what is optimized partly depends on the requirements of the task. Many cost functions were proposed in the past, and most of them were found to be in agreement with experimental data. Therefore, the actual principles on which the brain relies to achieve a certain motor behavior are still unclear. Existing results might suggest that movements are not the results of the minimization of single but rather of composite cost functions. In order to better clarify this last point, we consider an innovative experimental paradigm characterized by arm reaching with target redundancy. Within this framework, we make use of an inverse optimal control technique to automatically infer the (combination of) optimality criteria that best fit the experimental data. Results show that the subjects exhibited a consistent behavior during each experimental condition, even though the target point was not prescribed in advance. Inverse and direct optimal control together reveal that the average arm trajectories were best replicated when optimizing the combination of two cost functions, nominally a mix between the absolute work of torques and the integrated squared joint acceleration. Our results thus support the cost combination hypothesis and demonstrate that the recorded movements were closely linked to the combination of two complementary functions related to mechanical energy expenditure and joint-level smoothness.

This paper describes the control of a human-like robotic neck actuated with tendons. The controller regulates the length of the tendons to achieve a desired orientation of the neck and at the same time it maintains the tension of the tendons within certain limits. The solution we propose does not use any model of the system, but it relies on online learning of the different Jacobian mappings required by the controller. Learning, data acquisition and control are simultaneous; thus learning is completely autonomous, and purely online. We show that after enough iterations the controller produces straight trajectories in the task space and is able to maintain the tension of the tendons within safe limits.

During the last decades, interaction (with humans and with the environment) has become an increasingly interesting topic of research within the field of robotics. At the basis of interaction, a fundamental role is played by the ability to actively regulate the interaction forces. In this paper we propose a technique for controlling the interaction forces exploiting a proximal six axes force/torque sensor. The major assumption is the knowledge of the point where external forces are applied. The proposed approach is tested and validated on the four limbs of the iCub, a humanoid robot designed for research in embodied cognition. Remarkably, the proposed approach can be used to implement active compliance in other non passively back-drivable manipulators by simply inserting one or more force/torque sensor anywhere along the kinematic chain.

Online optimal planning of robotic arm movement is addressed. Optimality is inspired by computational models, where a “cost function” is used to describe limb motions according to different criteria. A method is proposed to implement optimal planning in Cartesian space, minimizing some cost function, by means of numerical approximation to a generalized nonlinear model predictive control problem. The Extended RItz Method is applied as a functional approximation technique. Differently from other approaches, the proposed technique can be applied on platforms with strict control temporal constraints and limited processing capability, since the computational burden is completely concentrated in an off-line phase. The trajectory generation on-line is therefore computationally efficient. Task to joint space conversion is implemented on-line by a closed loop inverse kinematics algorithm, taking into account the robot's physical limits. Experimental results, where a 4DOF arm moves according to a particular nonlinear cost, show the effectiveness of the proposed approach, and suggest interesting future developments.

In this paper we describe the design of a Cartesian Controller for a generic robot manipulator. We address some of the challenges that are typically encountered in the field of humanoid robotics. The solution we propose deals with a large number of degrees of freedom, produce smooth, human-like motion and is able to compute the trajectory on-line. In this paper we support the idea that to produce significant advancements in the field of robotics it is important to compare different approaches not only at the theoretical level but also at the implementation level. For this reason we test our software on the iCub platform and compare its performance against other available solutions.

In this paper, we describe the implementation of a precise reaching controller on an upper-torso humanoid robot. The solution we propose does not rely on prior models of the kinematic structure of either the arm or the head. A learning strategy enables the robot to acquire the required sensory-motor transformations. After learning the robot is able to precisely reach for a visually identified object in the 3-dimensional space. In this technique we use the fixation point (represented in the head joints motor space) as a reference frame to code the position of the object and to represent the eye-to-hand Jacobian matrix. This strategy successfully deals with the kinematic redundancy of the structure and constraints the dimensionality of the problem.

Recent experiments on sensory-motor systems of frogs and rats have revealed that those systems have a modular structure. Apparently, control actions form a vector space with a handful of elementary controls as a basis. The reduction of admissible controls to a vector space plays in control theory a similar role to PCA in learning and recognition applications. Inspired by these observations, we first propose a mathematical model of the experimentally observed modular structure. We then show how to choose elementary control actions so as to make the system reach any desired final state. Finally, we determine the minimum number of elementary control actions to perform reaching.