Search

Differential dynamic programming (DDP) is a direct single shooting method for trajectory optimization. Its efficiency derives from the exploitation of temporal structure (inherent to optimal control problems) and explicit roll-out/integration of the system dynamics. However, it suffers from numerical instability and, when compared to direct multiple shooting methods, it has limited initialization options (allows initialization of controls, but not of states) and lacks proper handling of control constraints. In this work, we tackle these issues with a feasibility-driven approach that regulates the dynamic feasibility during the numerical optimization and ensures control limits. Our feasibility search emulates the numerical resolution of a direct multiple shooting problem with only dynamics constraints. We show that our approach (named BOX-FDDP) has better numerical convergence than BOX-DDP+ (a single shooting method), and that its convergence rate and runtime performance are competitive with state-of-the-art direct transcription formulations solved using the interior point and active set algorithms available in KNITRO. We further show that BOX-FDDP decreases the dynamic feasibility error monotonically--as in state-of-the-art nonlinear programming algorithms. We demonstrate the benefits of our approach by generating complex and athletic motions for quadruped and humanoid robots. Finally, we highlight that BOX-FDDP is suitable for model predictive control in legged robots., Comment: 17 pages, 9 figures, draft version

The navigation of robots in dynamic urban environments, requires elaborated anticipative strategies for the robot to avoid collisions with dynamic objects, like bicycles or pedestrians, and to be human aware. We have developed and analyzed three anticipative strategies in motion planning taking into account the future motion of the mobile objects that can move up to 18 km/h. First, we have used our hybrid policy resulting from a Deep Deterministic Policy Gradient (DDPG) training and the Social Force Model (SFM), and we have tested it in simulation in four complex map scenarios with many pedestrians. Second, we have used these anticipative strategies in real-life experiments using the hybrid motion planning method and the ROS Navigation Stack with Dynamic Windows Approach (NS-DWA). The results in simulations and real-life experiments show very good results in open environments and also in mixed scenarios with narrow spaces., This is a preprint. This paper has been accepted to the fifth Iberian Robotics Conference (Robot 2022)

The Institut de Robòtica i Informàtica Industrial is a Joint University Research Institute participated by the Spanish National Research Council and the Universitat Politècnica de Catalunya. Founded in 1995, its scientists have addressed over the years many research topics spanning from robot kinematics, to computer graphics, automatic control, energy systems, and human-robot interaction, among others. This book, prepared for its 25th anniversary, covers its evolution over the years, and serves as a mean of appreciation to the many students, administrative personnel, research engineers, or scientists that have formed part of it.

Advances in computer vision and robotic manipulation are enabling assisted dressing.

Being able to teach complex capabilities, such as folding garments, to a bi-manual robot is a very challenging task, which is often tackled using learning from demonstration datasets. The few garment folding datasets available nowadays to the robotics research community are either gathered from human demonstrations or generated through simulation. The former have the huge problem of perceiving human action and transferring it to the dynamic control of the robot, while the latter requires coding human motion into the simulator in open loop, resulting in far-from-realistic movements. In this article, we present a reduced but very accurate dataset of human cloth folding demonstrations. The dataset is collected through a novel virtual reality (VR) framework we propose, based on Unity¿s 3D platform and the use of a HTC Vive Pro system. The framework is capable of simulating very realistic garments while allowing users to interact with them, in real time, through handheld controllers. By doing so, and thanks to the immersive experience, our framework gets rid of the gap between the human and robot perception-action loop, while simplifying data capture and resulting in more realistic samples., This work was developed in the context of the project CLOTHILDE (”CLOTH manIpulation Learning from DEmonstrations”) which has received funding from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation program (grant agreement No. 741930) and is also supported by the BURG project PCI2019-103447 funded by MCIN/ AEI /10.13039/501100011033 and by the ”European Union”.

En este capítulo afirmamos que la ciencia ficción anticipatoria tiene un papel importante que desempeñar en el actual contexto de desarrollo tecnológico acelerado, ya que puede ayudar a prever posibles futuros alternativos y promover debates éticos que cristalizen en iniciativas educativas y de regulación normativa en beneficio de la humanidad.

Forty years ago the notion of configuration space (C-space) revolutionised robot motion planning for rigid and articulated objects. Despite great progress, handling deformable materials has remained elusive because of their infinite-dimensional shape-state space. Finding low-complexity representations has become a pressing research goal. This work tries to make a tiny step in this direction by proposing a state representation for textiles relying on the C-space of some distinctive points. A stratification of the configuration space for n points in the cloth is derived from that of the flag manifold, and topological techniques to determine adjacencies in manipulation-centred state graphs are developed. Their algorithmic implementation permits obtaining cloth state–space representations of different granularities and tailored to particular purposes. An example of their usage to distinguish between cloth states having different manipulation affordances is provided. Suggestions on how the proposed state graphs can serve as a common ground to link the perception, planning and manipulation of textiles are also made., This work is supported by the European Research Council (ERC) within the European Union Horizon 2020 Programme under grant agreement ERC-2016-ADG-741930 (CLOTHILDE: CLOTH manIpulation Learning from DEmonstrations) and by the Spanish State Research Agency through the María de Maeztu Seal of Excellence to IRI (MDM-2016-0656). We thank Dr Jordi Sanchez Riera for providing us with cloth simulations.

This paper presents a new framework for how autonomous social robots approach and accompany people in urban environments. The method discussed allows the robot to accompany a person and approach to other one, by adapting its own navigation in anticipation of future interactions with other people or contact with static obstacles. The contributions of the paper are manifold: firstly, we extended the Social Force model and the Anticipative Kinodynamic Planner (Ferrer and Sanfeliu, in: IEEE/RSJ international conference on intelligent robots and systems. IEEE, 2014) to the case of an adaptive side-by-side navigation; secondly, we enhance side-by-side navigation with an approaching task and a final positioning that allows the robot to interact with both people; and finally, we use findings from experiments of real-life observations of people walking in pairs to define the parameters of the human–robot interaction in our case of adaptive side-by-side. The method was validated by a large set of simulations; we also conducted real-life experiments with our robot, Tibi, to validate the framework described for the interaction process. In addition, we carried out various surveys and user studies to indicate the social acceptability of the robots performance of the accompanying, approaching and positioning tasks., Work supported by the Spanish Ministry of Science and Innovation under project ColRobTransp (DPI2016-78957-R. AEI/FEDER EU), by the Spanish State Research Agency through the Maria de Maeztu Seal of Excellence to IRI (MDM-2016-0656) and EU TERRINET project (H2020-INFRAIA-2017-1-two-stage-730994). Ely Repiso is also supported by Spanish Ministry of Science and Innovation under a FPI-grant (BES-2014-067713).

This technical report complements the results recently presented in [1] showing that they can be extended to define an efficient and robust method to determine the rotation matrix nearest to an arbitrary 3 × 3 matrix. This problem arises in different areas of robotics that range from the simple case in which we have to restore the orthogonality of a noisy rotation matrix to point-cloud registration or hand-eye calibration. We show that the new method compares favorably with the classical approaches to address this problem and also with more recent methods

Using dual quaternions, the closure equations of a kinematic loop can be expressed as a system of multiaffine equations. In this paper, this property is leveraged to introduce a branch-and-prune method specially tailored for solving such systems of equations. The new method is objectively simpler (in the sense that it is easier to understand and to implement) than previous approaches relying on general techniques such as interval Newton methods or methods based on Bernstein polynomials or linear relaxations. Moreover, it relies on two basic operations — linear interpolation and projection onto coordinate planes— that can be efficiently computed. The generality of the proposed method is evaluated on position analysis problems with 0-, 1-, and 2-dimensional solution sets, including the inverse kinematics of serial robots and the forward kinematics of parallel ones. The results obtained on these problems show that the efficiency of the method compares favorably to state-of-the-art alternatives., This work was partially supported by the Spanish Ministry of Economy and Competitiveness through the project DPI2017-88282-P.

This paper proposes to enrich robot motion data with trajectory curvature information. To do so, we use an approximate implementation of a topological feature named writhe, which measures the curling of a closed curve around itself, and its analog feature for two closed curves, namely the linking number. Despite these features have been established for closed curves, their definition allows for a discrete calculation that is well-defined for non-closed curves and can thus provide information about how much a robot trajectory is curling around a line in space. Such lines can be predefined by a user, observed by vision or, in our case, inferred as virtual lines in space around which the robot motion is curling. We use these topological features to augment the data of a trajectory encapsulated as a Movement Primitive (MP). We propose a method to determine how many virtual segments best characterize a trajectory and then find such segments. This results in a generative model that permits modulating curvature to generate new samples, while still staying within the dataset distribution and being able to adapt to contextual variables., Thiswork has been carried out within the projectCLOTHILDE (“CLOTH manIpulation Learning from DEmonstrations”) funded by the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme (Advanced Grant agreementNo 741930).Research at IRI is also supported by the Spanish State Research Agency through theMaría de Maeztu Seal of Excellence to IRI MDM-2016-0656.

Autonomous and Intelligent Systems (A/IS, to adhere to the terminology of the IEEE Ethically Aligned Design report) can gather their knowledge by different means and from different sources. In principle, learning algorithms are neutral; rather, it is the data they are fed during the learning period that can introduce biases or a specific ethical orientation. Human control over the learning process is more straightforward in learning from demonstration, where data sources are restricted to the choices of the demonstrator (or teacher), but even in unsupervised versions of reinforcement learning, biases are present via the definition of the reward function. In this paper we provide an overview of learning paradigms of artificial systems: supervised and unsupervised methods, with the most striking examples in each category, without too much technical detail. Furthermore, we describe the types of data sources that are presently available and in use by the robotics community. We also focus on observable bias in image datasets and originated by human annotation. We point at quite recent research on bias in social robot navigation and end with a brief reflection about ambient influences on future learning robots., Los sistemas autónomos e inteligentes (A/IS por sus siglas en inglés, en concordancia con el informe del IEEE sobre diseño alineado con la ética) pueden obtener sus conocimientos a través de diferentes procedimientos y de fuentes diversas. Los algoritmos de aprendizaje son neutros en principio, son más bien los datos con los que se alimentan durante el período de aprendizaje que pueden introducir sesgos o una orientación ética específica. El control humano sobre el proceso de aprendizaje es más directo en aprendizaje por demostración, donde las fuentes de datos están restringidas a las elecciones del demostrador (o profesor), pero incluso en las versiones no supervisadas del aprendizaje por refuerzo los sesgos están presentes a través de la definición de la función de recompensa. En este artículo proporcionamos una visión general de los paradigmas de aprendizaje de los sistemas artificiales: métodos supervisados y no supervisados, con los ejemplos más destacados de cada categoría, sin profundizar demasiado en el detalle técnico. Además describimos los tipos de fuentes de datos disponibles actualmente y su uso por la comunidad robótica. También enfatizamos el sesgo que se observa en bases de datos de imágenes y originados por anotación humana. Destacamos una investigación muy reciente sobre sesgo en navegación de robots sociales y finalizamos con una breve reflexión sobre influencia del ambiente sobre futuros robots que aprenden.

During the last years, major advances on artificial intelligence have successfully allowed robots to perceive their environment, which not only includes static but also dynamic objects such as humans. Indeed, robotic perception is a fundamental feature to achieve safe robots' autonomy in human-robot collaboration. However, in order to have true collaboration, both robots and humans should perceive each other’s intentions and interpret which actions they are performing. In this work, we developed a visual representation tool that illustrates the robot's perception of the space that is shared with a person. Specifically, we adapted an existent system to estimate the human pose, and we created a visualisation tool to represent the robot's perception about the human-robot closeness. We also performed a first evaluation of the system working in realistic conditions using the Tiago robot and a person as a test subject. This work is a first step towards allowing humans to have a better understanding about robots' perception in collaborative scenarios.

Trabajo presentado en la IEEE/RSJ International Conference on Intelligent Robots and System, celebrada en Las Vegas (Estados Unidos), del 25 al 29 de octubre de 2020, Recently, Differential Dynamic Programming (DDP) and other similar algorithms have become the solvers of choice when performing non-linear Model Predictive Control (nMPC) with modern robotic devices. The reason is that they have a lower computational cost per iteration when compared with off-the-shelf Non-Linear Programming (NLP) solvers, which enables its online operation. However, they cannot handle constraints, and are known to have poor convergence capabilities. In this paper, we propose a method to solve the optimal control problem with control bounds through a squashing function (i.e., a sigmoid, which is bounded by construction). It has been shown that a naive use of squashing functions damage the convergence rate. To tackle this, we first propose to add a quadratic barrier that avoids the difficulty of the plateau produced by the sigmoid. Second, we add an outer loop that adapts both the sigmoid and the barrier; it makes the optimal control problem with the squashing function converge to the original control-bounded problem. To validate our method, we present simulation results for different types of platforms including a multi-rotor, a biped, a quadruped and a humanoid robot.

Trabajo presentado en el International Conference of IFToMM ITALY, celebrada en Nápoles (Italia) del 9 al 11 de septiembre de 2020, Parallel robots with a configurable platform are a class of parallel robots in which the end-effector is a closed-loop kinematic chain. In n-RRR planar robots the end-effector is a n-bar chain controlled by n actuated RRR chains connected to the base. We solve the direct kinematic problem for 4, 5 and 6-RRR mechanisms by using bilateration, a method that easily lends itself to generalization. Finally, we present the results from experimental tests that have been performed on a 5-RRR prototype., F. Thomas and J. M. Porta were partially supported by the Spanish Ministry of Economy and Competitiveness through projects DPI2017-88282-P and MDM-2016-0656. The support of Dr. Patrick Grosch in building the prototype is gratefully acknowledged.

Trabajo presentado en la IEEE International Conference on Robotics and Automation (ICRA), celebrada de forma virtual en París (Francia) del 31 de mayo al 31 de agosto de 2020, Today's complex robotic designs comprise in some cases a large number of degrees of freedom, enabling for multi-objective task resolution (e.g., humanoid robots or aerial manipulators). This paper tackles the local stability problem of a hierarchical closed-loop inverse kinematics algorithm for such highly redundant robots. We present a method to guarantee this system stability by performing an online tuning of the closed-loop control gains. We define a semi-definite programming problem (SDP) with these gains as decision variables and a discrete-time Lyapunov stability condition as a linear matrix inequality, constraining the SDP optimization problem and guaranteeing the local stability of the prioritized tasks. To the best of authors' knowledge, this work represents the first mathematical development of an SDP formulation that introduces these stability conditions for a multi-objective closed-loop inverse kinematic problem for highly redundant robots. The validity of the proposed approach is demonstrated through simulation case studies, including didactic examples and a Matlab toolbox for the benefit of the community.

In this chapter, the problem of fault-tolerant control of a service robot is addressed. The proposed approach is based on using a fault estimation scheme based on an Robust Unknown-Input Observer (RUIO) that allows to estimate the fault as well as the robot state. This fault estimation scheme is integrated with the control algorithm that is based on a observer-based state feedback control. After the fault occurrence, from the fault estimation, a feedforward control action is added to the feedback control action to compensate the fault effect. To cope with the robot non-linearity, its non-linear model is transformed into a Takagi-Sugeno model. Then, the statefeedback and RUIO are designed using an LMI-based approach considering a gainscheduling scheme. To illustrate the proposed fault-tolerant scheme a mobile service robot TIAGo, developed by PAL robotics, is used., This work is supported by the Spanish State Research Agency through the María de Maeztu Seal of Excellence to IRI MDM-2016-0656

Trabajo presentado en la 29th IEEE International Conference on Robot and Human Interactive Communication (RO-MAN), celebrada en Nápoles (Italia) del 31 de agosto al 4 de septiembre de 2020, Robots are expected to become part of everyday life. However, while there have been important breakthroughs during the recent decades in terms of technological advances, the ability of robots to interact with humans intuitively and effectively is still an open challenge. In this paper, we aim to evaluate how humans interpret and leverage backchannel cues exhibited by a robot which interacts with them in an entertainment context. To do so, a conceptual model was designed to investigate the legibility and the effectiveness of a designed social cue, called SOCial ImmediAcy BackchanneL cuE (SOCIABLE), on participant’s performance. In addition, user’s attitude and cognitive capability were integrated into the model as an estimator of participants’ motivation and ability to process the cue. In working toward such a goal, we conducted a two-day long user study (N=114) at an international event with untrained participants who were not aware of the social cue the robot was able to provide. The results showed that participants were able to perceive the social signal generated from SOCIABLE and thus, they benefited from it. Our findings provide some important insights for the design of effective and instantaneous backchannel cues and the methodology for evaluating them in social robots.

The presented method implements a robot accompaniment in an adaptive side-by-side formation of a single person or a group of people. The method enhances our previous robot adaptive side-by-side behavior allowing the robot to accompany a group of people, not only one person. Adaptive means that the robot is capable to adjust its position and velocity to the behavior of the group being accompanied, without bothering other pedestrians in the environment, as well as facilitating the group navigation to avoid static and dynamic obstacles. Furthermore, the robot can deal with the random factor of human behavior in several situations. Firstly, if other people interfere in the path of the companions, the robot leaves space for the person of the group that has to avoid those other people, by approaching the other companion. Also, without invading any personal space. Secondly, if the people of the group changes their physical position inside the group formation, then the robot adapts to them dynamically by changing from the lateral position to the central position of the formation or otherwise. Thirdly, the robot adapts to changes in the velocity of people in the group and other people that interfere in the path of the group, in magnitude and orientation. Fourthly, the robot can deal with occlusions of one accompanied person by the other. The method has been validated using synthetic experiments and real-life experiments with our robot. Finally, we developed an user study comparing the method with a Wizard of Oz., Work supported by the Spanish Ministry of Science and Innovation under project ColRobTransp (DPI2016-78957-RAEI/FEDER EU), TERRINet (H2020-INFRAIA-2017-1-two-stage-730994), by the Spanish State Research Agency through the Mar´ıa de Maeztu Seal of Excellence to IRI (MDM-2016-0656). Ely Repiso is also suported by Spanish Ministry of Science and Innovation under a FPI-grant, BES-2014-067713

Digital technologies have become part of our everyday lives and are increasingly acting as intermediaries in our workplaces and personal relationships or even substituting them. This poses a series of ethical questions that were not relevant for other types of machines and about which we have no previous experience, nor can we reliably predict how they will ultimately influence the evolution of humankind. This has led to the confluence of artificial intelligence (AI) with the humanities in an ethical debate that is starting to bear fruit, not only with the establishment of regulations and standards, but also with educational initiatives in university teaching, professional improvement, and the conformation of public opinion. Interestingly, science fiction (SF) often plays a prominent speculative role in highlighting the pros and cons of potential scenarios, thus favoring an engaging debate on AI and ethics. This work has been partly supported by the E uropean R esearch C ouncil (E R C ) under theE uropean U nion’s H orizon 2020 research and innovation programme through the projectC L O T H I L D E – C L O T H manI pulation L earning from D E monstrations (A dvanced G rantagreement N o 741930), and the S panish R esearch A gency through the M aría de M aeztu S eal ofE xcellence to I R I (M D M -2016-0656

Digital technologies have become part of our everyday lives and are increasingly acting as intermediaries in our workplaces and personal relationships or even substituting them. This poses a series of ethical questions that were not relevant for other types of machines and about which we have no previous experience, nor can we reliably predict how they will ultimately influence the evolution of humankind. This has led to the confluence of artificial intelligence (AI) with the humanities in an ethical debate that is starting to bear fruit, not only with the establishment of regulations and standards, but also with educational initiatives in university teaching, professional improvement, and the conformation of public opinion. Interestingly, science fiction (SF) often plays a prominent speculative role in highlighting the pros and cons of potential scenarios, thus favoring an engaging debate on AI and ethics. This work has been partly supported by the E uropean R esearch C ouncil (E R C ) under theE uropean U nion’s H orizon 2020 research and innovation programme through the projectC L O T H I L D E – C L O T H manI pulation L earning from D E monstrations (A dvanced G rantagreement N o 741930), and the S panish R esearch A gency through the M aría de M aeztu S eal ofE xcellence to I R I (M D M -2016-0656

Trabajo presentado en el International Symposium on Advances in Robot Kinematics, celebrado en Ljubljana (Eslovenia) del 6 al 10 de diciembre de 2020, Distance geometry is usually defined as the characterization and study of point sets in Rk , the k-dimensional Euclidean space, based on the pairwise distances between their points. In this paper, we use Clifford’s identity to extend this kind of characterization to sets of n hyperspheres embedded in Sn−3 or Rn−3 where the role of the Euclidean distance between two points is replaced by the so-called power between two hyperspheres. By properly choosing the value of n and the radii of these hyperspheres, Clifford’s identity reduces to conditions in terms of generalized Cayley-Menger determinants which has been previously obtained on the basis of a case-by-case analysis., This work was partially supported by the Spanish Ministry of Economy and Competitiveness through the projects DPI2017-88282-P and MDM-2016-0656.

Recent research and developments in cloud robotics (CR) require appropriate knowledge representation to ensure interoperable data, information, and knowledge sharing within cloud infrastructures. As an important branch of the Internet of Things (IoT), these demands to advance it forward motivates academic and industrial sectors to invest on it. The IEEE ’Ontologies for Robotics and Automation’ Working Group (ORA WG) has been developing standard ontologies for different robotic domains, including industrial and autonomous robots. The use of such robotic standards has the potential to benefit the Cloud Robotic Community (CRC) as well, supporting the provision of ubiquitous intelligent services by the CR-based systems. This paper explores this potential by developing an ontological approach for effective information sharing in cloud robotics scenarios. It presents an extension to the existing ontological standards to cater for the CR domain. The use of the new ontological elements is illustrated through its use in a couple of CR case studies. To the best of our knowledge, this is the first work ever that implements an ontology comprising concepts and axioms applicable to the CR domain., This work was partly supported by Innovate UK; CNPq and CAPES in Brazil; FRGS/1/2015/TK08/MUSM/02/1 in Malaysia.

The influence of the humanities on the study of a technological subject like robotics needs to rapidly grow, for the simple reason that robotics is becoming a part of humanity: assisting, interacting, and enabling people in an increasing number of ways in daily life. The robotics research community is well aware of the need for such a crossover with the humanities and many joint ventures are being undertaken, such as forums on “Robotics meets the Humanities” at main robotics conferences, the launching of research projects, and the publication of special issues in scientific journals. This cross-cutting has even led to a new discipline: Roboethics, a subfield of applied ethics studying both the positive and negative implications of robotics for individuals and society, with a view to inspire the moral design, development and use of so-called intelligent/autonomous robots, and help prevent their misuse against humankind. The discipline involves two main areas: legal regulation and ethical education. Regarding the former, institutions such as the European Parliament, the South Korean Robot Ethics Charter, the IEEE Standards Association, and the British Standards Institution are developing regulations for robot designers, programmers, and users. There are many options to integrate ethics education (or Humanities) in technological university degrees, ranging from including a professional ethics course in the syllabus, to allowing students to take certain credits or a minor in a Humanities Department, to even offering a combined degree, like the Computer Science and Philosophy degree at the University of Oxford. Prestigious associations such as IEEE and ACM include 18 knowledge areas in their Computer Science curricula, one of which is “Social Issues and Professional Practice”, so that “students develop an understanding of the relevant social, ethical, legal and professional issues”. To this end, some courses in this area recur to science fiction to exemplify conflictive situations, since narrative is a good way to engage students in safe discussion Carme Torras and reasoning about difficult and emotionally charged issues without making it personal. Some experiences along this line will be described, This work has been partly supported by the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme through the project CLOTHILDE - CLOTH manIpulation Learning from DEmonstrations (Advanced Grant agreement No 741930), and the Spanish Research Agency through the María de Maeztu Seal of Excellence to IRI (MDM-2016-0656).

© 2018 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works. This study presents a solution for the integrated longitudinal and lateral control problem of urban autonomousvehicles. It is based on a gain-scheduling linear parameter-varying (LPV) control approach combined with the use of anUnknown Input Observer (UIO) for estimating the vehicle states and friction force. Two gain-scheduling LPV controllers are usedin cascade configuration that use the kinematic and dynamic vehicle models and the friction and observed states provided bythe Unknown Input Observer (UIO). The LPV–UIO is designed in an optimal manner by solving a set of linear matrix inequalities(LMIs). On the other hand, the design of the kinematic and dynamic controllers lead to solve separately two LPV–LinearQuadratic Regulator problems formulated also in LMI form. The UIO allows to improve the control response in disturbanceaffected scenarios by estimating and compensating the friction force. The proposed scheme has been integrated with atrajectory generation module and tested in a simulated scenario. A comparative study is also presented considering the casesthat the friction force estimation is used or not to show its usefulness

© 20xx IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works. Current solutions to the simultaneous localization and mapping (SLAM) problem approach it as the optimization of a graph of geometric constraints. Scalability is achieved by reducing the size of the graph, usually in two phases. First, some selected nodes in the graph are marginalized and then, the dense and non-relinearizable result is sparsified. The sparsified network has a new set of relinearizable factors and is an approximation to the original dense one. Sparsification is typically approached as a Kullback-Liebler divergence (KLD) minimization between the dense marginalization result and the new set of factors. For a simple topology of the new factors, such as a tree, there is a closed form optimal solution. However, more populated topologies can achieve a much better approximation because more information can be encoded, although in that case iterative optimization is needed to solve the KLD minimization. Iterative optimization methods proposed by the state-of-art sparsification require parameter tuning which strongly affect their convergence. In this paper, we propose factor descent and non-cyclic factor descent, two simple algorithms for SLAM sparsification that match the state-of-art methods without any parameters to be tuned. The proposed methods are compared against the state of the art with regards to accuracy and CPU time, in both synthetic and real world datasets.

We address in this paper the problem of loop closure detection for laser-based simultaneous localization and mapping (SLAM) of very large areas. Consistent with the state of the art, the map is encoded as a graph of poses, and to cope with very large mapping capabilities, loop closures are asserted by comparing the features extracted from a query laser scan against a previously acquired corpus of scan features using a bag-ofwords (BoW) scheme. Two contributions are here presented. First, to benefit from the graph topology, feature frequency scores in the BoW are computed not only for each individual scan but also from neighboring scans in the SLAM graph. This has the effect of enforcing neighbor relational information during document matching. Secondly, a weak geometric check that takes into account feature ordering and occlusions is introduced that substantially improves loop closure detection performance. The two contributions are evaluated both separately and jointly on four common SLAM datasets, and are shown to improve the state-of-the-art performance both in terms of precision and recall in most of the cases. Moreover, our current implementation is designed to work at nearly frame rate, allowing loop closure query resolution at nearly 22 Hz for the best case scenario and 2 Hz for the worst case scenario., This work was supported by the Spanish Ministry of Economy and Competitiveness project ROBINSTRUCT TIN2014-58178-R, the EU H2020 LOGMATIC project 687534, and the Industrial Doctoral program of the Catalan Agency for Management of University and Research Grants.

Trabajo presentado en el 4th Iberian Robotics Conference, celebrado en Oporto (Portugal) en noviembre de 2019, This paper proposes a Social Reward Sources (SRS) design for a Human-Robot Collaborative Navigation (HRCN) task: human-robot collaborative search. It is a flexible approach capable of handling the collaborative task, human-robot interaction and environment restrictions, all integrated on a common environment. We modelled task rewards based on unexplored area observability and isolation and evaluated the model through different levels of human-robot communication.The models are validated through quantitative evaluation against both agents¿ individual performance and qualitative surveying of participants¿ perception. After that, the three proposed communication levels are com-pared against each other using the previous metrics., Work supported under projects ColRobTransp (DPI2016-78957-RAEI/FEDER EU), TERRINet (H2020-INFRAIA-2017-1-two-stage-730994) and by the Spanish State Research Agency through the Maria de Maeztu Seal of Excellence to IRI (MDM-2016-0656).

Trabajo presentado en la European Conference on Mobile Robots (ECMR), celebrada en Praga (República Checa), del 4 al 6 de septiembre de 2019, This paper describes a self-calibration procedure that jointly estimates the extrinsic parameters of an exteroceptive sensor able to observe ego-motion, and the intrinsic parameters of an odometry motion model, consisting of wheel radii and wheel separation. We use iterative nonlinear onmanifold optimization with a graphical representation of the state, and resort to an adaptation of the pre-integration theory, initially developed for the IMU motion sensor, to be applied to the differential drive motion model. For this, we describe the construction of a pre-integrated factor for the differential drive motion model, which includes the motion increment, its covariance, and a first-order approximation of its dependence with the calibration parameters. As the calibration parameters change at each solver iteration, this allows a posteriori factor correction without the need of re-integrating the motion data. We validate our proposal in simulations and on a real robot and show the convergence of the calibration towards the true values of the parameters. It is then tested online in simulation and is shown to accommodate to variations in the calibration parameters when the vehicle is subject to physical changes such as loading and unloading a freight., This work has been supported by the Spanish Ministry of Science, Innovation, and Universities project EB-SLAM (DPI2017-89564-P), by the EU H2020 project LOGIMATIC (H2020-Galileo-2015-1-687534) and by the Spanish State Research Agency through the Maria de Maeztu Seal of Excellence to IRI MDM-2016-0656. J. Deray acknowledges support from the Industrial Doctorate Program of the Catalan Agency for Management of University and Research Grants.

The final publication is available at link.springer.com A novel application is presented in this paper. Wastewater treatment plants need to sample water to follow the cleaned process continuously. This task is very tedious for laboratory workers and, mainly, can be done in reduced areas of the large basins that form the whole plant, specifically only in the edges of those basins. The new proposal is to enlarge the sampling area and reduce the load for workers. A new unmanned aerial vehicle has been designed together with a novel tool to take water samples. Moreover, a new architecture for the mission plan is presented in form of multi-agents. Experiments have been done in order to test the feasibility of the proposal. This research has been funded by AEROARMS EU Project H2020-ICT-2014-1-644271

Trabajo presentado en la IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), celebrada en Madrid, del 1 al 5 de octubre de 2018, Interaction of socially assistive robots with users is based on social cues coming from different interaction modalities, such as speech or gestures. However, using all modalities at all times may be inefficient as it can overload the user with redundant information and increase the task completion time. Additionally, users may favor certain modalities over the other as a result of their disability or personal preference. In this paper, we propose an Adaptive Modality Selection (AMS) algorithm that chooses modalities depending on the state of the user and the environment, as well as user preferences. The variables that describe the environment and the user state are defined as resources, and we posit that modalities are successful if certain resources possess specific values during their use. Besides the resources, the proposed algorithm takes into account user preferences which it learns while interacting with users. We tested our algorithm in simulations, and we implemented it on a robotic system that provides cognitive training, specifically Sequential memory exercises. Experimental results show that it is possible to use only a subset of available modalities without compromising the interaction. Moreover, we see a trend for users to perform better when interacting with a system with implemented AMS algorithm., This work was supported by the SOCRATES project funded from the European Union’s Horizon 2020 research and innovation programme under the Marie Skłodowska-Curie grant agreement No 721619 and by the Spanish State Research Agency through the Mar´ıa de Maeztu Seal of Excellence to IRI (MDM-2016-0656).

© 2019 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works. Robot-assisted dressing is performed in close physical interaction with users who may have a wide range of physical characteristics and abilities. Design of user adaptive and personalized robots in this context is still indicating limited, or no consideration, of specific user-related issues. This paper describes the development of a multi-modal robotic system for a specific dressing scenario -putting on a shoe, where users’ personalized inputs contribute to a much improved task success rate. We have developed: 1) user tracking, gesture recognition and posture recognition algorithms relying on images provided by a depth camera; 2) a shoe recognition algorithm from RGB and depth images; 3) speech recognition and text-to-speech algorithms implemented to allow verbal interaction between the robot and user. The interaction is further enhanced by calibrated recognition of the users’ pointing gestures and adjusted robot’s shoe delivery position. A series of shoe fitting experiments have been performed on two groups of users, with and without previous robot personalization, to assess how it affects the interaction performance. Our results show that the shoe fitting task with the personalized robot is completed in shorter time, with a smaller number of user commands and reduced workload.

Exploration of a complex underwater environment without an a priori map is beyond the state of the art for autonomous underwater vehicles (AUVs). Despite several efforts regarding simultaneous localization and mapping (SLAM) and view planning, there is no exploration framework, tailored to underwater vehicles, that faces exploration combining mapping, active localization, and view planning in a unified way. We propose an exploration framework, based on an active SLAM strategy, that combines three main elements: a view planner, an iterative closest point algorithm (ICP)-based pose-graph SLAM algorithm, and an action selection mechanism that makes use of the joint map and state entropy reduction. To demonstrate the benefits of the active SLAM strategy, several tests were conducted with the Girona 500 AUV, both in simulation and in the real world. The article shows how the proposed framework makes it possible to plan exploratory trajectories that keep the vehicle's uncertainty bounded; thus, creating more consistent maps., This research was funded by the Spanish State Research Agency through grants DPI2015-73978-JIN and DPI2017-89564-P, and the María de Maeztu Seal of Excellence to IRI (MDM-2016-0656).

[EN]Assistive robotics is a fast growing field aimed at helping healthcarers in hospitals, rehabilitation centers and nursery homes, as well as empowering people with reduced mobility at home, so that they can autonomously fulfill their daily living activities. The need to function in dynamic human-centered environments poses new research challenges: robotic assistants need to have friendly interfaces, be highly adaptable and customizable, very compliant and intrinsically safe to people, as well as able to handle deformable materials. Besides technical challenges, assistive robotics raises also ethical defies, which have led to the emergence of a new discipline: Roboethics. Several institutions are developing regulations and standards, and many ethics education initiatives include contents on human-robot interaction and human dignity in assistive situations. In this paper, the state of the art in assistive robotics is briefly reviewed, and educational materials from a university course on Ethics in Social Robotics and AI focusing on the assistive context are presented., [ES] La robótica asistencial es un campo en rápido crecimiento des-tinado a ayudar a los cuidadores en hospitales, centros de rehabilitación y residencias, así como a capacitar a las personas con movilidad reducida en el hogar, para que puedan realizar de forma autónoma sus activida-des cotidianas. La necesidad de desempeñarse en entornos dinámicos centrados en el ser humano plantea nuevos retos de investigación: los asistentes robóticos deben tener interfaces amigables, ser altamente adaptables y personalizables, muy compatibles y seguros para las perso-nas, así como ser capaces de manejar materiales deformables.Además de los desafíos técnicos, la robótica asistencial plantea también desafíos éticos, que han llevado a la aparición de una nueva disciplina: la roboética. Numerosas instituciones están elaborando reglamentos y normas, y muchas iniciativas de educación en ética incluyen contenidos sobre la interacción humano-robot y la dignidad humana en situaciones de asistencia.En este trabajo se revisa brevemente el estado del arte de la robótica asistencial y se presentan materiales educativos de un curso universitario sobre Ética de la Robótica Social e IA en el contexto asistencial., This work is partly supported by the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme through the project CLOTHIL-DE - CLOTH manIpulation Learning from DEmonstrations (Advanced Grant agreement No 741930), and the Spanish Research Agency through the María de Maeztu Seal of Excellence to IRI (MDM-2016-0656)

An analytic method to compute the solution intervals for the input variables of spatial RCRCR linkages and their inversions is presented. The input-output equation is formulated as the intersection of a single ellipse with a parameterized family of ellipses, both related with the possible values that certain dual angles determined by the configuration of the mechanism can take. Bounds for the angles of the input pairs of the RCRCR and RRCRC inversions are found by imposing the tangency of two ellipses, what reduces to analyzing the discriminant of a fourth degree polynomial. The bounds for the input pair of the RCRRC inversion is found as the intersection of a single ellipse with the envelope of the parameterized family of ellipses. The method provides the bounds of each of the assembly modes of the mechanism as well as the local extrema that may exist for the input variable., This work is supported by the Spanish Ministry of Economy and Competitiveness(MINECO) through the I+D projects DPI2014-57220-C2-2-P and DPI2017-88282-P.

Healthcare robots will be the next big advance in humans’ domestic welfare, with robots able to assist elderly people and users with disabilities. However, each user has his/her own preferences, needs and abilities. Therefore, robotic assistants will need to adapt to them, behaving accordingly. Towards this goal, we propose a method to perform behavior adaptation to the user preferences, using symbolic task planning. A user model is built from the user’s answers to simple questions with a fuzzy inference system, and it is then integrated into the planning domain. We describe an adaptation method based on both the user satisfaction and the execution outcome, depending on which penalizations are applied to the planner’s rules. We demonstrate the application of the adaptation method in a simple shoe-fitting scenario, with experiments performed in a simulated user environment. The results show quick behavior adaptation, even when the user behavior changes, as well as robustness to wrong inference of the initial user model. Finally, some insights in a non-simulated world shoe-fitting setup are also provided., This work has been partially supported by the CHIST-ERA Project I-DRESS PCIN-2015-147, by the Spanish Ministry of Science and Innovation HuMoUR TIN2017-90086-R and the AEI through the María de Maeztu Seal of Excellence to IRI (MDM-2016-0656). Gerard Canal is also supported by the Spanish Ministry of Education, Culture and Sport via a doctoral grant FPU15/00504.

The influence of the humanities on the study of a technological subject like robotics needs to rapidly grow, for the simple reason that robotics is becoming a part of humanity: assisting, interacting, and enabling people in an increasing number of ways in daily life. The robotics research community is well aware of the need for such a crossover with the humanities and many joint ventures are being undertaken, such as forums on “Robotics meets the Humanities” at main robotics conferences, the launching of research projects, and the publication of special issues in scientific journals. This cross-cutting has even led to a new discipline: Roboethics, a subfield of applied ethics studying both the positive and negative implications of robotics for individuals and society, with a view to inspire the moral design, development and use of so-called intelligent/autonomous robots, and help prevent their misuse against humankind. The discipline involves two main areas: legal regulation and ethical education. Regarding the former, institutions such as the European Parliament, the South Korean Robot Ethics Charter, the IEEE Standards Association, and the British Standards Institution are developing regulations for robot designers, programmers, and users. There are many options to integrate ethics education (or Humanities) in technological university degrees, ranging from including a professional ethics course in the syllabus, to allowing students to take certain credits or a minor in a Humanities Department, to even offering a combined degree, like the Computer Science and Philosophy degree at the University of Oxford. Prestigious associations such as IEEE and ACM include 18 knowledge areas in their Computer Science curricula, one of which is “Social Issues and Professional Practice”, so that “students develop an understanding of the relevant social, ethical, legal and professional issues”. To this end, some courses in this area recur to science fiction to exemplify conflictive situations, since narrative is a good way to engage students in safe discussion and reasoning about difficult and emotionally charged issues without making it personal. Some experiences along this line will be described.

Inferring human operators&rsquo, actions in shared collaborative tasks plays a crucial role in enhancing the cognitive capabilities of industrial robots. In all these incipient collaborative robotic applications, humans and robots not only should share space, but also forces and the execution of a task. In this article, we present a robotic system that is able to identify different human&rsquo, s intentions and to adapt its behavior consequently, only employing force data. In order to accomplish this aim, three major contributions are presented: (a) a force based operator&rsquo, s intention recognition system based on data from only two users, (b) a force based dataset of physical human&ndash, robot interaction, and (c) validation of the whole system with 15 people in a scenario inspired by a realistic industrial application. This work is an important step towards a more natural and user-friendly manner of physical human&ndash, robot interaction in scenarios where humans and robots collaborate in the accomplishment of a task.

Information technologies have become part of our everyday lives and are increasingly acting as intermediaries in our workplaces and personal relationships or even substituting them. This growing interaction with machines poses several questions about which we have no previous experience, nor can we reliably predict how they will influence the evolution of society. This has led to the convergence of technoscience and humanities in an ethical debate that is starting to bear fruit, not only with the setting of regulations and standards, but also with educational initiatives in university teaching, professional improvement, and the conformation of public opinion. Interestingly, science fiction often plays a prominent speculative role in highlighting the pros and cons of potential scenarios.

Trabajo presentado en el IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), celebrado en Macao (China), del 3 al 8 de noviembre de 2019, This paper presents a new method to allow robots to accompany a person or a group of people imitating pedestrians behavior. Two-people groups usually walk in a side-by-side formation and three-people groups walk in a V-formation so that they can see each other. For this reason, the proposed method combines a Side-by-side and V-formation pedestrian model with the Anticipative Kinodynamic Planner (AKP). Combining these methods, the robot is able to do an anticipatory accompaniment of groups of humans, as well as to avoid static and dynamic obstacles in advance, while keeping the prescribed formations. The proposed framework allows also a dynamical re-positioning of the robot, if the physical position of the partners change in the group formation. Furthermore, people have a randomness factor that the robot has to manage, for that reason, the system was adapted to deal with changes in people's velocity, orientation and occlusions. Finally, the method has been validated using synthetic experiments and real-life experiments with our Tibi robot. In addition, a user study has been realized to reveal the social acceptability of the method.

The final publication is available at link.springer.com This chapter is devoted to explaining a method for precise localization using augmented reality markers. This method can achieve precision of less of 5 mm in position at a distance of 0.7 m, using a visual mark of 17 mm × 17 mm, and it can be used by controller when the aerial robot is doing a manipulation task. The localization method is based on optimizing the alignment of deformable contours from textureless images working from the raw vertexes of the observed contour. The algorithm optimizes the alignment of the XOR area computed by means of computer graphics clipping techniques. The method can run at 25 frames per second.

Within the next decades, robots will need to be able to execute a large variety of tasks autonomously in a large variety of environments. To relax the resulting programming effort, a knowledge-enabled approach to robot programming can be adopted to organize information in re-usable knowledge pieces. However, for the ease of reuse, there needs to be an agreement on the meaning of terms. A common approach is to represent these terms using ontology languages that conceptualize the respective domain. In this work, we will review projects that use ontologies to support robot autonomy. We will systematically search for projects that fulfill a set of inclusion criteria and compare them with each other with respect to the scope of their ontology, what types of cognitive capabilities are supported by the use of ontologies, and which is their application domain., This work was partially funded by the following agencies: Deutsche Forschungsgemeinschaft (DFG) through the Collaborative Research Center EASE (1320); Regional Catalan Agency ACCIO through the RIS3CAT 2016 project SIMBIOTS (COMRDI16-1-0017); Spanish State ´ Research Agency through the Mar´ıa de Maeztu Seal of Excellence to IRI (MDM-2016-0656); Spanish Government through the project DPI2016-80077-R and grant 2017; FCT, through IDMEC, under LAETA, project UID/EMS/50022/2019; Project 0043- EUROAGE-4-E (POCTEP Programa Interreg V-A Spain-Portugal). J. Olszewska was partially supported by Innovate UK.

© 2019 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works. This work presents a solution to automatize sampling tasks in a wastewater treatment plant with open air basins. At the behest of human operators, a set of UAVs managed as a network of autonomous agents will perform sample missions by taking direct measurements (through a multiparametric probe) or capturing samples and carrying them to the laboratory with specific developed hardware. These capabilities allow the proposed solution to act as a virtual sensor network with sampling points deployed and connected at any point reachable by UAVs. The hardware prototypes are fully described, with focus on the integration of systems, and the software architecture used is analysed and fully justified. Special focus was put on the localization problem, and several solutions were evaluated. Experimental results of the prototype UAV and sampling probe built are provided to validate the hardware designs, with focus on the localizations tasks. This research has been funded by Spanish Ministery of Innovation and Science CICYT DPI2016-78957-R

Trabajo presentado en el International Symposium on Anticipation and Anticipatory Systems: Humans Meet AI (ANTICIPATION-2019), celebrado en Örebro (Suecia), del 10 al 13 de junio de 2019, The influence that hyperconnectivity and our increasing interaction with machines will have on the evolution of society and on people¿s daily lives is difficult to predict. Thus, when trying to anticipate the potential benefits and risks of information technologies, not just lay people resort to science fiction (SF), but some scholars and even companies do so. Several universities, particularly in the US, include in their Computer Science and Engineering degrees a course on ethics and human values relative to technology. In this context of university education, my novel The Vestigial Heart (MIT Press, 2018) has been published together with online materials to teach a course on Ethics in Social Robotics and AI. Anticipatory literature has always taken science seriously and has tried to project its accomplishments into the future. It seems that science is also starting to take this literature seriously and find inspiration therein. This confluence could be extremely productive and is very good news, opening up interesting perspectives for the coming years.

Assistive robotics is a fast growing field aimed at helping healthcarers in hospitals, rehabilitation centers and nursery homes, as well as empowering people with reduced mobility at home, so that they can autonomously fulfill their daily living activities. The need to function in dynamic human-centered environments poses new research challenges: robotic assistants need to have friendly interfaces, be highly adaptable and customizable, very compliant and intrinsically safe to people, as well as able to handle deformable materials. Besides technical challenges, assistive robotics raises also ethical defies, which have led to the emergence of a new discipline: Roboethics. Several institutions are developing regulations and standards, and many ethics education initiatives include contents on human-robot interaction and human dignity in assistive situations. In this paper, the state of the art in assistive robotics is briefly reviewed, and educational materials from a university course on Ethics in Social Robotics and AI focusing on the assistive context are presented.

The final publication is available at link.springer.com This research presents a methodology for the detection of the crawler used in the project AEROARMS. The approach consisted on using a two-step progressive strategy, going from rough detection and tracking, for approximation maneuvers, to an accurate positioning step based on fiducial markers. Two different methods are explained for the first step, one using efficient image segmentation approach; and the second one using Deep Learning techniques to detect the center of the crawler. The fiducial markers are used for precise localization of the crawler in a similar way as explained in earlier chapters. The methods can run in real-time.

The influence that hyperconnectivity and our increasing interaction with machines will have on the evolution of society and on people’s daily lives is difficult to predict. Thus, when trying to anticipate the potential benefits and risks of information technologies, not just lay people resort to science fiction (SF), but some scholars and even companies do so. Several universities, particularly in the US, include in their Computer Science and Engineering degrees a course on ethics and human values relative to technology. In this context of university education, my novel The Vestigial Heart (MIT Press, 2018) has been published together with online materials to teach a course on Ethics in Social Robotics and AI. Anticipatory literature has always taken science seriously and has tried to project its accomplishments into the future. It seems that science is also starting to take this literature seriously and find inspiration therein. This confluence could be extremely productive and is very good news, opening up interesting perspectives for the coming years.

The current fourth industrial revolution, or `Industry 4.0¿ (I4.0), is driven by digital data, connectivity, and cyber systems, and it has the potential to create impressive/new business opportunities. With the arrival of I4.0, the scenario of various intelligent systems interacting reliably and securely with each other becomes a reality which technical systems need to address. One major aspect of I4.0 is to adopt a coherent approach for the semantic communication in between multiple intelligent systems, which include human and artificial (software or hardware) agents. For this purpose, ontologies can provide the solution by formalizing the smart manufacturing knowledge in an interoperable way. Hence, this paper presents the few existing ontologies for I4.0, along with the current state of the standardization effort in the factory 4.0 domain and examples of real-world scenarios for I4.0., S. Veera Ragavan was supported by Monash University and MOHE grant FRGS/1/2015/TK08/ MUSM/02/1, entitled ‘Towards a model synthesis framework for conceptual modeling of cyber-physical systems’. Paulo J. S. Goncalves was partially supported by FCT, through IDMEC, uner LAETA, project UID/EMS/50022/2109 and project 0043-EUROAGE-4-E (POCTEP Programa Interreg V-A Spain-Portugal). J.I. Olszewska was partially supported by Innovate UK.

Trabajo presentado en la 4th Conference on Control and Fault Tolerant Systems (SysTol), celebrada en Casablanca (Marruecos), del 18 al 20 de septiembre de 2019, In this paper, the control problem of a service robot under exogenous forces that might be exerted on the system is addressed, considering them as faults. Proposed approach is based on using a fault estimation scheme together with a Robust Unknown-Input Observer (RUIO) that allows to estimate the fault as well as the robot state, along with a state feedback control strategy. After the fault occurrence, from the fault estimation, a feedforward control action is added to the feedback control action to compensate the fault effect. To cope with the robot non-linearity, its model is transformed into a quasi-LPV model. Both the state-feedback and RUIO are designed using an LMI-based approach considering a gain-scheduling alike scheme, where global conditions have been considered. To illustrate the proposed fault-tolerant approach, a mobile service robot TIAGo, developed by PAL robotics, is used.

Trabajo presentado en la IEEE International Conference on Robotics and Automation (ICRA), conferencia virtual celebrada del 31 de mayo al 31 de agosto de 2020, In this paper we present a Deep Reinforcement Learning approach to solve dynamic cloth manipulation tasks. Differing from the case of rigid objects, we stress that the followed trajectory (including speed and acceleration) has a decisive influence on the final state of cloth, which can greatly vary even if the positions reached by the grasped points are the same. We explore how goal positions for non-grasped points can be attained through learning adequate trajectories for the grasped points. Our approach uses few demonstrations to improve control policy learning, and a sparse reward approach to avoid engineering complex reward functions. Since perception of textiles is challenging, we also study different state representations to assess the minimum observation space required for learning to succeed. Finally, we compare different combinations of control policy encodings, demonstrations, and sparse reward learning techniques, and show that our proposed approach can learn dynamic cloth manipulation in an efficient way, i.e., using a reduced observation space, a few demonstrations, and a sparse reward., This work has been supported by the ERC project Clothilde (ERC-2016- ADG-741930), the HuMoUR project TIN2017-90086-R (AEI/FEDER, UE) and by the AEI through the María de Maeztu Seal of Excellence to IRI (MDM-2016-0656)