Francesco Nori

Tuomas Haarnoja, Ben Moran, Guy Lever et al. · deepmind

We investigate whether Deep Reinforcement Learning (Deep RL) is able to synthesize sophisticated and safe movement skills for a low-cost, miniature humanoid robot that can be composed into complex behavioral strategies in dynamic environments. We used Deep RL to train a humanoid robot with 20 actuated joints to play a simplified one-versus-one (1v1) soccer game. The resulting agent exhibits robust and dynamic movement skills such as rapid fall recovery, walking, turning, kicking and more; and it transitions between them in a smooth, stable, and efficient manner. The agent's locomotion and tactical behavior adapts to specific game contexts in a way that would be impractical to manually design. The agent also developed a basic strategic understanding of the game, and learned, for instance, to anticipate ball movements and to block opponent shots. Our agent was trained in simulation and transferred to real robots zero-shot. We found that a combination of sufficiently high-frequency control, targeted dynamics randomization, and perturbations during training in simulation enabled good-quality transfer. Although the robots are inherently fragile, basic regularization of the behavior during training led the robots to learn safe and effective movements while still performing in a dynamic and agile way -- well beyond what is intuitively expected from the robot. Indeed, in experiments, they walked 181% faster, turned 302% faster, took 63% less time to get up, and kicked a ball 34% faster than a scripted baseline, while efficiently combining the skills to achieve the longer term objectives.

Konstantinos Bousmalis, Giulia Vezzani, Dushyant Rao et al.

The ability to leverage heterogeneous robotic experience from different robots and tasks to quickly master novel skills and embodiments has the potential to transform robot learning. Inspired by recent advances in foundation models for vision and language, we propose a multi-embodiment, multi-task generalist agent for robotic manipulation. This agent, named RoboCat, is a visual goal-conditioned decision transformer capable of consuming action-labelled visual experience. This data spans a large repertoire of motor control skills from simulated and real robotic arms with varying sets of observations and actions. With RoboCat, we demonstrate the ability to generalise to new tasks and robots, both zero-shot as well as through adaptation using only 100-1000 examples for the target task. We also show how a trained model itself can be used to generate data for subsequent training iterations, thus providing a basic building block for an autonomous improvement loop. We investigate the agent's capabilities, with large-scale evaluations both in simulation and on three different real robot embodiments. We find that as we grow and diversify its training data, RoboCat not only shows signs of cross-task transfer, but also becomes more efficient at adapting to new tasks.

Arunkumar Byravan, Jan Humplik, Leonard Hasenclever et al.

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 volume 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 degrees of freedom humanoid robot with an actuated head-mounted RGB camera, and we successfully transfer these policies to a real robot. Project video is available at https://sites.google.com/view/nerf2real/home

Maria Bauza, Jose Enrique Chen, Valentin Dalibard et al.

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 https://sites.google.com/view/demostart.

Serena Ivaldi, Vincent Padois, Francesco Nori

The number of tools for dynamics simulation has grown in the last years. It is necessary for the robotics community to have elements to ponder which of the available tools is the best for their research. As a complement to an objective and quantitative comparison, difficult to obtain since not all the tools are open-source, an element of evaluation is user feedback. With this goal in mind, we created an online survey about the use of dynamical simulation in robotics. This paper reports the analysis of the participants' answers and a descriptive information fiche for the most relevant tools. We believe this report will be helpful for roboticists to choose the best simulation tool for their researches.

Dhruva Tirumala, Markus Wulfmeier, Ben Moran et al. · deepmind

We apply multi-agent deep reinforcement learning (RL) to train end-to-end robot soccer policies with fully onboard computation and sensing via egocentric RGB vision. This setting reflects many challenges of real-world robotics, including active perception, agile full-body control, and long-horizon planning in a dynamic, partially-observable, multi-agent domain. We rely on large-scale, simulation-based data generation to obtain complex behaviors from egocentric vision which can be successfully transferred to physical robots using low-cost sensors. To achieve adequate visual realism, our simulation combines rigid-body physics with learned, realistic rendering via multiple Neural Radiance Fields (NeRFs). We combine teacher-based multi-agent RL and cross-experiment data reuse to enable the discovery of sophisticated soccer strategies. We analyze active-perception behaviors including object tracking and ball seeking that emerge when simply optimizing perception-agnostic soccer play. The agents display equivalent levels of performance and agility as policies with access to privileged, ground-truth state. To our knowledge, this paper constitutes a first demonstration of end-to-end training for multi-agent robot soccer, mapping raw pixel observations to joint-level actions, that can be deployed in the real world. Videos of the game-play and analyses can be seen on our website https://sites.google.com/view/vision-soccer .

Toshiaki Tsuji, Yasuhiro Kato, Gokhan Solak et al.

This paper comprehensively surveys research trends in imitation learning for contact-rich robotic tasks. Contact-rich tasks, which require complex physical interactions with the environment, represent a central challenge in robotics due to their nonlinear dynamics and sensitivity to small positional deviations. The paper examines demonstration collection methodologies, including teaching methods and sensory modalities crucial for capturing subtle interaction dynamics. We then analyze imitation learning approaches, highlighting their applications to contact-rich manipulation. Recent advances in multimodal learning and foundation models have significantly enhanced performance in complex contact tasks across industrial, household, and healthcare domains. Through systematic organization of current research and identification of challenges, this survey provides a foundation for future advancements in contact-rich robotic manipulation.

Ken Caluwaerts, Atil Iscen, J. Chase Kew et al.

Animals have evolved various agile locomotion strategies, such as sprinting, leaping, and jumping. There is a growing interest in developing legged robots that move like their biological counterparts and show various agile skills to navigate complex environments quickly. Despite the interest, the field lacks systematic benchmarks to measure the performance of control policies and hardware in agility. We introduce the Barkour benchmark, an obstacle course to quantify agility for legged robots. Inspired by dog agility competitions, it consists of diverse obstacles and a time based scoring mechanism. This encourages researchers to develop controllers that not only move fast, but do so in a controllable and versatile way. To set strong baselines, we present two methods for tackling the benchmark. In the first approach, we train specialist locomotion skills using on-policy reinforcement learning methods and combine them with a high-level navigation controller. In the second approach, we distill the specialist skills into a Transformer-based generalist locomotion policy, named Locomotion-Transformer, that can handle various terrains and adjust the robot's gait based on the perceived environment and robot states. Using a custom-built quadruped robot, we demonstrate that our method can complete the course at half the speed of a dog. We hope that our work represents a step towards creating controllers that enable robots to reach animal-level agility.

Alex X. Lee, Coline Devin, Yuxiang Zhou et al.

We study the problem of robotic stacking with objects of complex geometry. We propose a challenging and diverse set of such objects that was carefully designed to require strategies beyond a simple "pick-and-place" solution. Our method is a reinforcement learning (RL) approach combined with vision-based interactive policy distillation and simulation-to-reality transfer. Our learned policies can efficiently handle multiple object combinations in the real world and exhibit a large variety of stacking skills. In a large experimental study, we investigate what choices matter for learning such general vision-based agents in simulation, and what affects optimal transfer to the real robot. We then leverage data collected by such policies and improve upon them with offline RL. A video and a blog post of our work are provided as supplementary material.

Lorenzo Natale, Chiara Bartolozzi, Francesco Nori et al.

In this chapter we describe the history and evolution of the iCub humanoid platform. We start by describing the first version as it was designed during the RobotCub EU project and illustrate how it evolved to become the platform that is adopted by more than 30 laboratories world wide. We complete the chapter by illustrating some of the research activities that are currently carried out on the iCub robot, i.e. visual perception, event driven sensing and dynamic control. We conclude the Chapter with a discussion of the lessons we learned and a preview of the upcoming next release of the robot, iCub 3.0.

Joan Kangro, Anand Vazhapilli Sureshbabu, Silvio Traversaro et al.

In modern robotic applications, tactile sensor arrays (i.e., artificial skins) are an emergent solution to determine the locations of contacts between a robot and an external agent. Localizing the point of contact is useful but determining the force applied on the skin provides many additional possibilities. This additional feature usually requires time-consuming calibration procedures to relate the sensor readings to the applied forces. This letter presents a novel device that enables the calibration of tactile sensor arrays in a fast and simple way. The key idea is to design a plenum chamber where the skin is inserted, and then the calibration of the tactile sensors is achieved by relating the air pressure and the sensor readings. This general concept is tested experimentally to calibrate the skin of the iCub robot. The validation of the calibration device is achieved by placing the masses of known weight on the artificial skin and comparing the applied force against the one estimated by the sensors.

Giulia Vezzani, Michael Neunert, Markus Wulfmeier et al.

Robot manipulation requires a complex set of skills that need to be carefully combined and coordinated to solve a task. Yet, most ReinforcementLearning (RL) approaches in robotics study tasks which actually consist only of a single manipulation skill, such as grasping an object or inserting a pre-grasped object. As a result the skill ('how' to solve the task) but not the actual goal of a complete manipulation ('what' to solve) is specified. In contrast, we study a complex manipulation goal that requires an agent to learn and combine diverse manipulation skills. We propose a challenging, highly under-actuated peg-in-hole task with a free, rotational asymmetrical peg, requiring a broad range of manipulation skills. While correct peg (re-)orientation is a requirement for successful insertion, there is no reward associated with it. Hence an agent needs to understand this pre-condition and learn the skill to fulfil it. The final insertion reward is sparse, allowing freedom in the solution and leading to complex emerging behaviour not envisioned during the task design. We tackle the problem in a multi-task RL framework using Scheduled Auxiliary Control (SAC-X) combined with Regularized Hierarchical Policy Optimization (RHPO) which successfully solves the task in simulation and from scratch on a single robot where data is severely limited.

Rae Jeong, Jost Tobias Springenberg, Jackie Kay et al.

Learning dexterous manipulation in high-dimensional state-action spaces is an important open challenge with exploration presenting a major bottleneck. Although in many cases the learning process could be guided by demonstrations or other suboptimal experts, current RL algorithms for continuous action spaces often fail to effectively utilize combinations of highly off-policy expert data and on-policy exploration data. As a solution, we introduce Relative Entropy Q-Learning (REQ), a simple policy iteration algorithm that combines ideas from successful offline and conventional RL algorithms. It represents the optimal policy via importance sampling from a learned prior and is well-suited to take advantage of mixed data distributions. We demonstrate experimentally that REQ outperforms several strong baselines on robotic manipulation tasks for which suboptimal experts are available. We show how suboptimal experts can be constructed effectively by composing simple waypoint tracking controllers, and we also show how learned primitives can be combined with waypoint controllers to obtain reference behaviors to bootstrap a complex manipulation task on a simulated bimanual robot with human-like hands. Finally, we show that REQ is also effective for general off-policy RL, offline RL, and RL from demonstrations. Videos and further materials are available at sites.google.com/view/rlfse.

Rae Jeong, Jackie Kay, Francesco Romano et al.

Learning robotic control policies in the real world gives rise to challenges in data efficiency, safety, and controlling the initial condition of the system. On the other hand, simulations are a useful alternative as they provide an abundant source of data without the restrictions of the real world. Unfortunately, simulations often fail to accurately model complex real-world phenomena. Traditional system identification techniques are limited in expressiveness by the analytical model parameters, and usually are not sufficient to capture such phenomena. In this paper we propose a general framework for improving the analytical model by optimizing state dependent generalized forces. State dependent generalized forces are expressive enough to model constraints in the equations of motion, while maintaining a clear physical meaning and intuition. We use reinforcement learning to efficiently optimize the mapping from states to generalized forces over a discounted infinite horizon. We show that using only minutes of real world data improves the sim-to-real control policy transfer. We demonstrate the feasibility of our approach by validating it on a nonprehensile manipulation task on the Sawyer robot.

Rae Jeong, Yusuf Aytar, David Khosid et al.

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 exploit 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 propose 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.

Devin Schwab, Tobias Springenberg, Murilo F. Martins et al.

We present a method for fast training of vision based control policies on real robots. The key idea behind our method is to perform multi-task Reinforcement Learning with auxiliary tasks that differ not only in the reward to be optimized but also in the state-space in which they operate. In particular, we allow auxiliary task policies to utilize task features that are available only at training-time. This allows for fast learning of auxiliary policies, which subsequently generate good data for training the main, vision-based control policies. This method can be seen as an extension of the Scheduled Auxiliary Control (SAC-X) framework. We demonstrate the efficacy of our method by using both a simulated and real-world Ball-in-a-Cup game controlled by a robot arm. In simulation, our approach leads to significant learning speed-ups when compared to standard SAC-X. On the real robot we show that the task can be learned from-scratch, i.e., with no transfer from simulation and no imitation learning. Videos of our learned policies running on the real robot can be found at https://sites.google.com/view/rss-2019-sawyer-bic/.

Francisco Javier Andrade Chavez, Gabriele Nava, Silvio Traversaro et al.

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 \textit{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 \textit{in situ} calibration of six axis force torque sensors method is extended to perform temperature compensation.

Yeshasvi Tirupachuri, Gabriele Nava, Claudia Latella et al.

The topic of physical human-robot interaction received a lot of attention from the robotics community because of many promising application domains. However, studying physical interaction between a robot and an external agent, like a human or another robot, without considering the dynamics of both the systems may lead to many short-comings in fully exploiting the interaction. In this paper, we present a coupled-dynamics formalism followed by a sound approach in exploiting helpful interaction with a humanoid robot. In particular, we propose the first attempt to define and exploit the human help for the robot to accomplish a specific task. As a result, we present a task-based partner-aware robot control techniques. The theoretical results are validated by conducting experiments with two iCub humanoid robots involved in physical interaction.

Yeshasvi Tirupachuri, Silvio Traversaro, Francesco Nori et al.

Articulated objects like doors, drawers, valves, and tools are pervasive in our everyday unstructured dynamic environments. Articulation models describe the joint nature between the different parts of an articulated object. As most of these objects are passive, a robot has to interact with them to infer all the articulation models to understand the object topology. We present a general algorithm to estimate the inherent articulation models by exploiting the momentum of the articulated system along with the interaction wrench while manipulating the object. We validate our approach with experiments in a simulation environment.

Stefano Dafarra, Gabriele Nava, Marie Charbonneau et al.

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.

Francisco Javier Andrade Chavez, Joan Kangro, Silvio Traversaro et al.

In this paper, we present algorithms to estimate external contact forces and joint torques using only skin, i.e. distributed tactile sensors. To deal with gaps between the tactile sensors (taxels), we use interpolation techniques. The application of these interpolation techniques allows us to estimate contact forces and joint torques without the need for expensive force-torque sensors. Validation was performed using the iCub humanoid robot.

Gabriele Nava, Daniele Pucci, Nuno Guedelha et al.

Forthcoming applications concerning humanoid robots may involve physical interaction between the robot and a dynamic environment. In such scenario, classical balancing and walking controllers that neglect the environment dynamics may not be sufficient for achieving a stable robot behavior. This paper presents a modeling and control framework for balancing humanoid robots in contact with a dynamic environment. We first model the robot and environment dynamics, together with the contact constraints. Then, a control strategy for stabilizing the full system is proposed. Theoretical results are verified in simulation with robot iCub balancing on a seesaw.

Marie Charbonneau, Gabriele Nava, Francesco Nori et al.

A whole-body torque control framework adapted for balancing and walking tasks is presented in this paper. In the proposed approach, centroidal momentum terms are excluded in favor of a hierarchy of high-priority position and orientation tasks and a low-priority postural task. More specifically, the controller stabilizes the position of the center of mass, the orientation of the pelvis frame, as well as the position and orientation of the feet frames. The low-priority postural task provides reference positions for each joint of the robot. Joint torques and contact forces to stabilize tasks are obtained through quadratic programming optimization. Besides the exclusion of centroidal momentum terms, part of the novelty of the approach lies in the definition of control laws in SE(3) which do not require the use of Euler parameterization. Validation of the framework was achieved in a scenario where the robot kept balance while walking in place. Experiments have been conducted with the iCub robot, in simulation and in real-world experiments.

Stefano Dafarra, Francesco Romano, Francesco Nori

Balancing and reacting to strong and unexpected pushes is a critical requirement for humanoid robots. We recently designed a capture point based approach which interfaces with a momentum-based torque controller and we implemented and validated it on the iCub humanoid robot. In this work we implement a Receding Horizon control, also known as Model Predictive Control, to add the possibility to predict the future evolution of the robot, especially the constraints switching given by the hybrid nature of the system. We prove that the proposed MPC extension makes the step-recovery controller more robust and reliable when executing the recovery strategy. Experiments in simulation show the results of the proposed approach.

Stefano Dafarra, Francesco Romano, Gabriele Nava et al.

When balancing, a humanoid robot can be easily subjected to unexpected disturbances like external pushes. In these circumstances, reactive movements as steps become a necessary requirement in order to avoid potentially harmful falling states. In this paper we conceive a Model Predictive Controller which determines a desired set of contact wrenches by predicting the future evolution of the robot, while taking into account constraints switching in case of steps. The control inputs computed by this strategy, namely the desired contact wrenches, are directly obtained on the robot through a modification of the momentum-based whole-body torque controller currently implemented on iCub. The proposed approach is validated through simulations in a stepping scenario, revealing high robustness and reliability when executing a recovery strategy.

Stefano Dafarra, Francesco Romano, Francesco Nori

One of the challenges for the robotics community is to deploy robots which can reliably operate in real world scenarios together with humans. A crucial requirement for legged robots is the capability to properly balance on their feet, rejecting external disturbances. iCub is a state-of-the-art humanoid robot which has only recently started to balance on its feet. While the current balancing controller has proved successful in various scenarios, it still misses the capability to properly react to strong pushes by taking steps. This paper goes in this direction. It proposes and implements a control strategy based on the Capture Point concept [1]. Instead of relying on position control, like most of Capture Point related approaches, the proposed strategy generates references for the momentum-based torque controller already implemented on the iCub, thus extending its capabilities to react to external disturbances, while retaining the advantages of torque control when interacting with the environment. Experiments in the Gazebo simulator and on the iCub humanoid robot validate the proposed strategy.

Francesco Nori

We propose an algorithm to compute the dynamics of articulated rigid-bodies with different sensor distributions. Prior to the on-line computations, the proposed algorithm performs an off-line optimisation step to simplify the computational complexity of the underlying solution. This optimisation step consists in formulating the dynamic computations as a system of linear equations. The computational complexity of computing the associated solution is reduced by performing a permuted LU-factorisation with off-line optimised permutations. We apply our algorithm to solve classical dynamic problems: inverse and forward dynamics. The computational complexity of the proposed solution is compared to `gold standard' algorithms: recursive Newton-Euler and articulated body algorithm. It is shown that our algorithm reduces the number of floating point operations with respect to previous approaches. We also evaluate the numerical complexity of our algorithm by performing tests on dynamic computations for which no gold standard is available.

Gabriele Nava, Daniele Pucci, Francesco Nori

Humanoid robots may require a degree of compliance at the joint level for improving efficiency, shock tolerance, and safe interaction with humans. The presence of joint elasticity, however, complexifies the design of balancing and walking controllers. This paper proposes a control framework for extending momentum based controllers developed for stiff actuators 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.

Daniele Pucci, Silvio Traversaro, Francesco Nori

The paper takes the first step towards the de- velopment of a control framework for underactuated flying humanoid robots. These robots may thus have the capacities of flight, contact locomotion, and manipulation, and benefit from technologies and methods developed for Whole-Body Control and Aerial Manipulation. As in the case of quadrotors, we as- sume that the humanoid robot is powered by four thrust forces. For convenience, these forces are placed at the robot hands and feet. The control objective is defined as the asymptotic stabilization of the robot centroidal momentum. This objective allows us to track a desired trajectory for the robot center of mass and keep small errors between a reference orientation and the robot base frame. Stability and convergence of the robot momentum are shown to be in the sense of Lyapunov. Simulations carried out on a model of the humanoid robot iCub verify the soundness of the proposed approach.

Alessandro Saccon, Silvio Traversaro, Francesco Nori et al.

In the literature on robotics and multibody dynamics, the concept of average angular velocity has received considerable attention in recent years. We address the question of whether the average angular velocity defines an orientation framethat depends only on the current robot configuration and provide a simple algebraic condition to check whether this holds. In the language of geometric mechanics, this condition corresponds to requiring the flatness of the mechanical connection associated to the robotic system. Here, however, we provide both a reinterpretation and a proof of this result accessible to readers with a background in rigid body kinematics and multibody dynamics but not necessarily acquainted with differential geometry, still providing precise links to the geometric mechanics literature. This should help spreading the algebraic condition beyond the scope of geometric mechanics,contributing to a proper utilization and understanding of the concept of average angular velocity.

Francesco Romano, Silvio Traversaro, Daniele Pucci et al.

In this paper, we propose a software abstraction layer to simplify the design and synthesis of whole-body controllers without making any preliminary assumptions on the control law to be implemented. The main advantage of the proposed library is the decoupling of the control software from implementation details, which are related to the robotic platform. Furthermore, the resulting code is more clean and concise than ad-hoc code, as it focuses only on the implementation of the control law. In addition, we present a reference implementation of the abstraction layer together with a Simulink interface to provide support to Model-Driven based development. We also show the implementation of a simple proportional-derivative plus gravity compensation control together with a more complex momentum-based bipedal balance controller.

Silvio Traversaro, Stanislas Brossette, Adrien Escande et al.

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.

Francisco Javier Andrade Chavez, Silvio Traversaro, Daniele Pucci et al.

This paper proposes and validates an in situ calibration method to calibrate six axis force torque (F/T) sensors once they are mounted on the system. This procedure takes advantage of the knowledge of the model of the robot to generate the expected wrenches of the sensors during some arbitrary motions. It then uses this information to train and validate new calibration matrices, taking into account the calibration matrix obtained with a classical Workbench calibration. The proposed calibration algorithm is validated on the F/T sensors mounted on the iCub humanoid robot legs.

Daniele Pucci, Gabriele Nava, Francesco Nori

This paper proposes a technique for automatic gain tuning of a momentum based balancing controller for humanoid robots. The controller ensures the stabilization of the centroidal dynamics and the associated zero dynamics. Then, the closed-loop, constrained joint space dynamics is linearized and the controller's gains are chosen so as to obtain desired properties of the linearized system. Symmetry and positive definiteness constraints of gain matrices are enforced by proposing a tracker for symmetric positive definite matrices. Simulation results are carried out on the humanoid robot iCub.

Francesco Romano, Daniele Pucci, Silvio Traversaro et al.

Legged locomotion has received increasing attention from the robotics community. In this respect, contact stability plays a critical role in ensuring that robots maintain balance, and it is a key element for balancing and walking controllers. The Center of Pressure is a contact stability criterion that defines a point that must be kept strictly inside the support polygon in order to ensure postural stability. In this paper, we introduce the concept of the sensitivity of the static center of pressure: roughly speaking, the rate of change of the center of pressure with respect to the system equilibrium configurations. This new concept can be used as an additional criterion to assess the robustness of the contact stability. We show how the sensitivity of the center of pressure can also be used as a metric to assess balancing controllers by considering two state-of-the-art control strategies. The analytical analysis is performed on a simplified model, and validated during balancing tasks on the iCub humanoid robot.

Joan Kangro, Silvio Traversaro, Daniele Pucci et al.

The paper presents a proof of concept to calibrate iCub's skin using vacuum bags. The method's main idea consists in inserting the skin in a vacuum bag, and then decreasing the pressure in the bag to create a uniform pressure distribution on the skin surface. Acquisition and data processing of the bag pressure and sensors' measured capacitance allow us to characterize the relationship between the pressure and the measured capacitance of each sensor. After calibration, integration of the pressure distribution over the skin geometry provides us with the net normal force applied to the skin. Experiments are conducted using the forearm skin of the iCub humanoid robot, and validation results indicate acceptable average errors in force prediction.

Marie Charbonneau, Francesco Nori, Daniele Pucci

This paper proposes control laws ensuring the stabilization of a time-varying desired joint trajectory, as well as joint limit avoidance, in the case of fully-actuated manipulators. The key idea is to perform a parametrization of the feasible joint space in terms of exogenous states. It follows that the control of these states allows for joint limit avoidance. One of the main outcomes of this paper is that position terms in control laws are replaced by parametrized terms, where joint limits must be avoided. Stability and convergence of time-varying reference trajectories obtained with the proposed method are demonstrated to be in the sense of Lyapunov. The introduced control laws are verified by carrying out experiments on two degrees-of-freedom of the humanoid robot iCub.

Yue Hu, Jorhabib Eljaik, Kevin Stein et al.

The humanoid robot iCub is a research platform of the Fondazione Istituto Italiano di Tecnologia (IIT), spread among different institutes around the world. In the most recent version of iCub, the robot is equipped with stronger legs and bigger feet, allowing it to perform balancing and walking motions that were not possible with the first generations. Despite the new legs hardware, walking has been rarely performed on the iCub robot. In this work the objective is to implement walking motions on the robot, from which we want to analyze its walking capabilities. We developed software modules based on extensions of classic techniques such as the ZMP based pattern generator and position control to identify which are the characteristics as well as limitations of the robot against different walking tasks in order to give the users a reference of the performance of the robot. Most of the experiments have been performed with HeiCub, a reduced version of iCub without arms and head.

Georg Wiedebach, Sylvain Bertrand, Tingfan Wu et al.

We present a method for humanoid robot walking on partial footholds such as small stepping stones and rocks with sharp surfaces. Our algorithm does not rely on prior knowledge of the foothold, but information about an expected foothold can be used to improve the stepping performance. After a step is taken, the robot explores the new contact surface by attempting to shift the center of pressure around the foot. The available foothold is inferred by the way in which the foot rotates about contact edges and/or by the achieved center of pressure locations on the foot during exploration. This estimated contact area is then used by a whole body momentum-based control algorithm. To walk and balance on partial footholds, we combine fast, dynamic stepping with the use of upper body angular momentum to regain balance. We applied this method to the Atlas humanoid designed by Boston Dynamics to walk over small contact surfaces, such as line and point contacts. We present experimental results and discuss performance limitations.

Gabriele Nava, Francesco Romano, Francesco Nori et al.

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.

Raffaello Camoriano, Silvio Traversaro, Lorenzo Rosasco et al.

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. This yields to 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.

Andrea Del Prete, Nicolas Mansard, Francesco Nori et al.

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 multitask 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-offreedom 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.

Andrea Del Prete, Francesco Romano, Lorenzo Natale et al.

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.

Silvio Traversaro, Andrea Del Prete, Riccardo Muradore et al.

Identification of inertial parameters is fundamental for the implementation of torque-based control in humanoids. At the same time, good models of friction and actuator dynamics are critical for the low-level control of joint torques. We propose a novel method to identify inertial, friction and motor parameters in a single procedure. The identification exploits the measurements of the PWM of the DC motors and a 6-axis force/torque sensor mounted inside the kinematic chain. The partial least-square (PLS) method is used to perform the regression. We identified the inertial, friction and motor parameters of the right arm of the iCub humanoid robot. We verified that the identified model can accurately predict the force/torque sensor measurements and the motor voltages. Moreover, we compared the identified parameters against the CAD parameters, in the prediction of the force/torque sensor measurements. Finally, we showed that the estimated model can effectively detect external contacts, comparing it against a tactile-based contact detection. The presented approach offers some advantages with respect to other state-of-the-art methods, because of its completeness (i.e. it identifies inertial, friction and motor parameters) and simplicity (only one data collection, with no particular requirements).

Andrea Del Prete, Francesco Nori, Giorgio Metta et al.

We present a new framework for prioritized multi-task motion-force control of fully-actuated robots. This work is established on a careful review and comparison of the state of the art. Some control frameworks are not optimal, that is they do not find the optimal solution for the secondary tasks. Other frameworks are optimal, but they tackle the control problem at kinematic level, hence they neglect the robot dynamics and they do not allow for force control. Still other frameworks are optimal and consider force control, but they are computationally less efficient than ours. Our final claim is that, for fully-actuated robots, computing the operational-space inverse dynamics is equivalent to computing the inverse kinematics (at acceleration level) and then the joint-space inverse dynamics. Thanks to this fact, our control framework can efficiently compute the optimal solution by decoupling kinematics and dynamics of the robot. We take into account: motion and force control, soft and rigid contacts, free and constrained robots. Tests in simulation validate our control framework, comparing it with other state-of-the-art equivalent frameworks and showing remarkable improvements in optimality and efficiency.

Silvio Traversaro, Daniele Pucci, Francesco Nori

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.