Search

Aquest projecte centra la seva atenció en el desenvolupament d’una aplicació per a poder entrenar un robot per a que pugui dur a terme una tasca de Pick and Place. En el món de l’aprenentatge automàtic, existeixen diferents algorismes d’implementació per a poder fer exercicis d’aquest tipus, en aquest treball s’ha triat el Deep Deterministic Policy Gradient (DDPG) fusionat amb el Hindsight Experience Replay (HER). S’ha utilitzat el gym d’OpenAI per a provar i executar els diferents fulls de codi que s’han desenvolupat amb el llenguatge de programació de Python. Al llarg de l’estudi s’exposen les bases teòriques a seguir i s’expliquen les diferents estructures dels scripts desenvolupats per a entrenar al model programat. D’aquesta manera es pot seguir el fil del disseny la IA de principi a final adquirint també, paral·lelament, coneixement del món de la intel·ligència artificial. El proyecto presente centra su atención en el desarrollo de una aplicación para poder entrenar un robot para que pueda realizar una tarea de Pick and Place. En el mundo del aprendizaje autónomo, existen diferentes algoritmos de implementación para poder hacer ejercicios de este tipo, en este trabajo se ha escogido a Deep Deterministic Policy Gradient (DDPG) mezclado con el Hindsight Experience Replay (HER). Se ha utilizado el gym de OpenAI para probar y ejecutar las distintas hojas de código que se han desarrollado con el lenguaje de programación de Python. A lo largo del estudio se explican las bases teóricas a seguir y se exponen las diferentes estructures de los scripts creados para entrenar al modelo programado. De esta forma, se puede seguir el hilo del diseño de la IA de principio a fin, adquiriendo también, paralelamente, conocimientos del mundo de la inteligencia artificial. This project focuses on the development of an application to train a robot to perform a Pick and Place task. In the world of autonomous learning, there are different implementation algorithms to be able to do exercises of this type, in this work Deep Deterministic Policy Gradient (DDPG) has been chosen and mixed with Hindsight Experience Replay (HER). The OpenAI gym has been used to test and execute the different code sheets that have been developed with the Python programming language. Throughout the study, the theoretical bases to be followed are explained and the different structures of the scripts created to train the programmed model are presented. In this way, it is possible to follow the AI design thread from start to finish, acquiring, in parallel, knowledge of the world of artificial intelligence.

A dia d’avui trobem intel·ligència artificial aplicada a gran diversitat de camps, la podem trobar als telèfons mòbils, presents a la butxaca de qualsevol persona fins als hospitals o als bancs. L’aplicació de la intel·ligència artificial amb combinació de la robòtica ha representat una millora de la qualitat de vida. Des-de feines tedioses i monòtones fins a tasques perilloses o complicades que abans es duien a terme per persones ara es fan amb robots. Els robots autònoms són capaços de reconèixer i aprendre sobre el seu entorn i prendre decisions de forma independent sense la necessitat de control humà. Varien amb dimensions, funcionalitat, mobilitat, etc. en funció de les seves aplicacions. Es poden trobar als habitatges com els robot Roomba, processos industrials o destinats a l’exploració. La percepció es molt important en el robots autònoms, ja que abans de prendre decisions han d’avaluar informació sobre l’entorn. L’objectiu d’aquest treball de final grau serà dotar un robot quadrúpede, el Bittle, amb percepció visual a traves de l’entrenament d’una xarxa neuronal que utilitza el model Mask RCNN. Controlat per una Raspberry Pi, es buscara que el Bittle sigui capaç de detectar pilotes i actuar amb conseqüència

Robots are usually equipped with 3D range sensors such as laser line scanners (LLSs) or lidars. These sensors acquire a full 3D scan in a line by line manner while the robot is in motion. All the lines can be referred to a common coordinate frame using data from inertial sensors. However, errors from noisy inertial measurements and inaccuracies in the extrinsic parameters between the scanner and the robot frame are also projected onto the shared frame. This causes a deformation in the final scan containing all the lines, which is known as motion distortion. Rigid point cloud registration with methods like ICP is therefore not well suited for such distorted scans. In this paper we present a non-rigid registration method that finds the rigid transformation to be applied to each line in the scan in order to match an existing model. We fully leverage the continuous and relatively smooth robot motion with respect to the scanning time to formulate our method reducing the computational complexity while improving accuracy. We use synthetic and real data to benchmark our method against a state-of-the-art non-rigid registration method. Finally, the source code for the algorithm is made publicly available., IEEE Robotics and Automation Letters, 7 (3), ISSN:2377-3766

This paper proposes a 3D object recognition method for non-coloured point clouds using point features. The method is intended for application scenarios such as Inspection, Maintenance and Repair (IMR) of industrial sub-sea structures composed of pipes and connecting objects (such as valves, elbows and R-Tee connectors). The recognition algorithm uses a database of partial views of the objects, stored as point clouds, which is available a priori. The recognition pipeline has 5 stages: (1) Plane segmentation, (2) Pipe detection, (3) Semantic Object-segmentation and detection, (4) Feature based Object Recognition and (5) Bayesian estimation. To apply the Bayesian estimation, an object tracking method based on a new Interdistance Joint Compatibility Branch and Bound (IJCBB) algorithm is proposed. The paper studies the recognition performance depending on: (1) the point feature descriptor used, (2) the use (or not) of Bayesian estimation and (3) the inclusion of semantic information about the objects connections. The methods are tested using an experimental dataset containing laser scans and Autonomous Underwater Vehicle (AUV) navigation data. The best results are obtained using the Clustered Viewpoint Feature Histogram (CVFH) descriptor, achieving recognition rates of 51.2%, 68.6% and 90%, respectively, clearly showing the advantages of using the Bayesian estimation (18% increase) and the inclusion of semantic information (21% further increase).

One of the problems that service robotics deals with is to bring mobile manipulators to work in semi-structured human scenarios, which requires an efficient and flexible way to execute every-day tasks, like serve a cup in a cluttered environment. Usually, for those tasks, the combination of symbolic and geometric levels of planning is necessary, as well as the integration of perception models with knowledge to guide both planning levels, resulting in a sequence of actions or skills which, according to the current knowledge of the world, may be executed. This paper proposes a planning and execution framework, called SkillMaN, for robotic manipulation tasks, which is equipped with a module with experiential knowledge (learned from its experience or given by the user) on how to execute a set of skills, like pick-up, put-down or open a drawer, using workflows as well as robot trajectories. The framework also contains an execution assistant with geometric tools and reasoning capabilities to manage how to actually execute the sequence of motions to perform a manipulation task (which are forwarded to the executor module), as well as the capacity to store the relevant information to the experiential knowledge for further usage, and the capacity to interpret the actual perceived situation (in case the preconditions of an action do not hold) and to feed back the updated state to the planner to resume from there, allowing the robot to adapt to non-expected situations. To evaluate the viability of the proposed framework, an experiment has been proposed involving different skills performed with various types of objects in different scene contexts

Autonomous indoor service robots are supposed to accomplish tasks, like serve a cup, which involve manipulation actions. Particularly, for complex manipulation tasks which are subject to geometric constraints, spatial information and a rich semantic knowledge about objects, types, and functionality are required, together with the way in which these objects can be manipulated. In this line, this paper presents an ontological-based reasoning framework called Perception and Manipulation Knowledge (PMK) that includes: (1) the modeling of the environment in a standardized way to provide common vocabularies for information exchange in human-robot or robot-robot collaboration, (2) a sensory module to perceive the objects in the environment and assert the ontological knowledge, (3) an evaluation-based analysis of the situation of the objects in the environment, in order to enhance the planning of manipulation tasks. The paper describes the concepts and the implementation of PMK, and presents an example demonstrating the range of information the framework can provide for autonomous robots.

Nowadays, it is quite common to have one unmanned aerial vehicle (UAV) working on a task but having a team of UAVs is still rare. One of the problems that prevent us from using teams of UAVs more frequently is flight planning. In this work, we present the first open-source solution ( https://pypi.org/project/pode/ ) for splitting any complex area into multiple parts. The area of interest can be convex or nonconvex and can include any number of no-flight zones. Four solutions, based on the algorithm of Hert and Lumelsky, are tested with the aim of improving the compactness of the partitions. We also show how the shape of the partitions influences flight performance in a real case scenario. This work was supported by Ministerio de Economía, Industria y Competitividad, and Gobierno de España under grants/award numbers BES-2017-079798 and TRA2016-77012-R.

Action recognition is a challenging task that plays an important role in many robotic systems, which highly depend on visual input feeds. However, due to privacy concerns, it is important to find a method which can recognise actions without using visual feed. In this paper, we propose a concept for detecting actions while preserving the test subject&rsquo, s privacy. Our proposed method relies only on recording the temporal evolution of light pulses scattered back from the scene. Such data trace to record one action contains a sequence of one-dimensional arrays of voltage values acquired by a single-pixel detector at 1 GHz repetition rate. Information about both the distance to the object and its shape are embedded in the traces. We apply machine learning in the form of recurrent neural networks for data analysis and demonstrate successful action recognition. The experimental results show that our proposed method could achieve on average 96.47 % accuracy on the actions walking forward, walking backwards, sitting down, standing up and waving hand, using recurrent neural network.

EU Marine Robots project grant agreement no. 731103 The marine robotics community is lacking a high quality simulator for doing scientific research, especially when it comes to testing control and vision algorithms in realistic underwater intervention tasks. All of the solutions used today are either outdated or try to combine different software tools, which often results in bad performance, stability issues and lack of important features. This paper presents a new software tool, focused on, but not limited to, simulation of intervention autonomous underwater vehicles (I-AUV). It delivers advanced hydrodynamics based on actual geometry, simulation of underwater sensors and actuators, as well as realistic rendering of underwater environment and ocean surface. It consists of a library written in C++ and a Robot Operating System (ROS) package Patryk Ciésslak has received funding from the European Community H2020 Programme under the Marie Skłodowska-Curie grant agreement no. 750063 and under EU Marine Robots project grant agreement no. 731103

This paper describes a pose-based underwater 3D Simultaneous Localization and Mapping (SLAM) using a multibeam echosounder to produce high consistency underwater maps. The proposed algorithm compounds swath profiles of the seafloor with dead reckoning localization to build surface patches (i.e., point clouds). An Iterative Closest Point (ICP) with a probabilistic implementation is then used to register the point clouds, taking into account their uncertainties. The registration process is divided in two steps: (1) point-to-point association for coarse registration and (2) point-to-plane association for fine registration. The point clouds of the surfaces to be registered are sub-sampled in order to decrease both the computation time and also the potential of falling into local minima during the registration. In addition, a heuristic is used to decrease the complexity of the association step of the ICP from O(n2) to O(n). The performance of the SLAM framework is tested using two real world datasets: First, a 2.5D bathymetric dataset obtained with the usual down-looking multibeam sonar configuration, and second, a full 3D underwater dataset acquired with a multibeam sonar mounted on a pan and tilt unit This work was supported by the Spanish project DPI2014-57746-C3-3-R (ARCHROV) and two European Commission’s Seventh Framework Program projects: FP7-ICT-2011-7-288704 (MORPH) and FP7-INFRASTRUCTURES-2012-312762 (EUROFLEETS2)

We describe a model-based objects recognition system which is part of an image interpretation system intended to assist autonomous vehicles navigation. The system is intended to operate in man-made environments. Behavior-based navigation of autonomous vehicles involves the recognition of navigable areas and the potential obstacles. The recognition system integrates color, shape and texture information together with the location of the vanishing point. The recognition process starts from some prior scene knowledge, that is, a generic model of the expected scene and the potential objects. The recognition system constitutes an approach where different low-level vision techniques extract a multitude of image descriptors which are then analyzed using a rule-based reasoning system to interpret the image content. This system has been implemented using CEES, the C++ embedded expert system shell developed in the Systems Engineering and Automatic Control Laboratory (University of Girona) as a specific rule-based problem solving tool. It has been especially conceived for supporting cooperative expert systems, and uses the object oriented programming paradigm.

Research article compendium : peace and conflict research The University of Barcelona, in collaboration with different Spanish and foreign institutions and universities, hosted an international workshop on the ‘Sense and Scope of Autonomy in Emerging Military and Security Technologies’, which took place in February 2017. This academic event paved the way to the elaboration of the present report, which contains a compendium of innovative articles written by experts on scientific, legal, diplomatic and military matters. All papers reflect a broad spectrum of views on how ‘autonomy’ can be understood in emerging military and security technologies and what their legal, technical, political and societal impacts are by building on discussions unfolded along a variety of reflective approaches.

Treball desenvolupat dins el marc del programa 'European Project Semester'. The robotics facilities at Oslo and Akershus University College of Applied Sciences (HiOA) have at their disposal an autonomous humanoid robot by the name of Nao. It is a commercial product developed by SoftBank Robotics for research and educational purposes, and an international project team was assembled for the European Project Semester (EPS) program as a means of assessing its competencies and bolstering them wherever possible. The team set forth with a mission to establish meaningful interactions between the robot and his physical surroundings while placing an emphasis on humanlike behavior to the fullest extent. A goal was soon initiated for Nao to play a rule-based game against a human opponent such that his vast array of available faculties could be tested and put to use in one cohesive demonstration. Achieving this goal meant making advances in artificial intelligence,computer vision and motion planning, first as separate entities that would later converge into a singular yet multifaceted display of abilities. At the end of the project, computer vision has been implemented to the point where pattern recognition can be performed with accuracy amidst awide variety of error-prone detection scenarios. Several artificial intelligence schemes have been explored and compared quantitatively to derive performance metrics for each. Proof-ofconcept movement strategies have also been derived, and best practices for each focus area are provided as a resource for any future researchers or project groups who stand to benefit from the discoveries made in this endeavor

This work introduces new results on the modeling of early-vocal development using artificial intelligent cognitive architectures and a simulated vocal tract. The problem is addressed using intrinsically-motivated learning algorithms for autonomous sensorimotor exploration, a kind of algorithm belonging to the active learning architectures family. The artificial agent is able to autonomously select goals to explore its own sensorimotor system in regions where its competence to execute intended goals is improved. We propose to include a somatosensory system to provide a proprioceptive feedback signal to reinforce learning through the autonomous discovery of motor constraints. Constraints are represented by a somatosensory model which is unknown beforehand to the learner. Both the sensorimotor and somatosensory system are modeled using Gaussian mixture models. We argue that using an architecture which includes a somatosensory model would reduce redundancy in the sensorimotor model and drive the learning process more efficiently than algorithms taking into account only auditory feedback. The role of this proposed system is to predict whether an undesired collision within the vocal tract under a certain motor configuration is likely to occur. Thus, compromised motor configurations are rejected, guaranteeing that the agent is less prone to violate its own constraints.

This paper addresses the problem of obtaining the required motions for a humanoid robot to perform grasp actions trying to mimic the coordinated hand–arm movements humans do. The first step is the data acquisition and analysis, which consists in capturing human movements while grasping several everyday objects (covering four possible grasp types), mapping them to the robot and computing the hand motion synergies for the pre-grasp and grasp phases (per grasp type). Then, the grasp and motion synthesis step is done, which consists in generating potential grasps for a given object using the four family types, and planning the motions using a bi-directional multi-goal sampling-based planner, which efficiently guides the motion planning following the synergies in a reduced search space, resulting in paths with human-like appearance. The approach has been tested in simulation, thoroughly compared with other state-of-the-art planning algorithms obtaining better results, and also implemented in a real robot.

This paper proposes a pose-based algorithm to solve the full simultaneous localization and mapping problem for autonomous underwater vehicle (AUV) navigating in unknown and possibly unstructured environments. The proposed method first estimates the local path traveled by the robot while forming the acoustic image (scan) with range data coming from a mono-beam rotating sonar head, providing position estimates for correcting the distortions that the vehicle motion produces in the scans. Then, consecutive scans are cross-registered under a probabilistic scan matching technique for estimating the displacements of the vehicle including the uncertainty of the scan matching result. Finally, an augmented state extended Kalman filter estimates and keeps the registered scans poses. No prior structural information or initial pose are considered. The viability of the proposed approach has been tested reconstructing the trajectory of a guided AUV operating along a 600 m path within a marina environment This research work was partially sponsored by the Spanish project DPI2011-27977-C03-02 (COMAROB) and two European Commission's Seventh Framework Program 2007-2013 Projects: ICT-248497 (TRIDENT) and Marie Curie PERG-GA-2010-276778 (Surf3DSLAM). The dataset was acquired with the help of the members (staff and students) of the Computer Vision and Robotics research group at the University of Girona

In this letter, we propose a method to automate the exploration of unknown underwater structures for autonomous underwater vehicles (AUVs). The proposed algorithm iteratively incorporates exteroceptive sensor data and replans the next-best-view in order to fully map an underwater structure. This approach does not require prior environment information. However, a safe exploration depth and the exploration area (defined by a bounding box, parameterized by its size, location, and resolution) must be provided by the user. The algorithm operates online by iteratively conducting the following three tasks: (1) Profiling sonar data are first incorporated into a 2-D grid map, where voxels are labeled according to their state (a voxel can be labeled as empty, unseen, occluded, occplane, occupied, or viewed). (2) Useful viewpoints to continue exploration are generated according to the map. (3) A safe path is generated to guide the robot toward the next viewpoint location. Two sensors are used in this approach: a scanning profiling sonar, which is used to build an occupancy map of the surroundings, and an optical camera, which acquires optical data of the scene. Finally, in order to demonstrate the feasibility of our approach, we provide real-world results using the Sparus II AUV This work was supported by the EXCELLABUST and ARCHROV Projects under Grants H2020-TWINN-2015, CSA, ID: 691980, and DPI2014-57746-C3-3-R. The work of E. Vidal was supported by the Spanish Government through Ph.D. grant FPU14/05493

Treball desenvolupat dins el marc del programa 'European Project Semester'. The robotics facilities at Oslo and Akershus University College of Applied Sciences (HiOA) have at their disposal an autonomous humanoid robot by the name of Nao. It is a commercial product developed by SoftBank Robotics for research and educational purposes, and an international project team was assembled for the European Project Semester (EPS) program as a means of assessing its competencies and bolstering them wherever possible. The team set forth with a mission to establish meaningful interactions between the robot and his physical surroundings while placing an emphasis on humanlike behavior to the fullest extent. A goal was soon initiated for Nao to play a rule-based game against a human opponent such that his vast array of available faculties could be tested and put to use in one cohesive demonstration. Achieving this goal meant making advances in artificial intelligence,computer vision and motion planning, first as separate entities that would later converge into a singular yet multifaceted display of abilities. At the end of the project, computer vision has been implemented to the point where pattern recognition can be performed with accuracy amidst awide variety of error-prone detection scenarios. Several artificial intelligence schemes have been explored and compared quantitatively to derive performance metrics for each. Proof-ofconcept movement strategies have also been derived, and best practices for each focus area are provided as a resource for any future researchers or project groups who stand to benefit from the discoveries made in this endeavor

The use of commercially available autonomous underwater vehicles (AUVs) has increased during the last fifteen years. While they are mainly used for routine survey missions, there is a set of applications that nowadays can be only addressed by manned submersibles or work-class remotely operated vehicles (ROVs) equipped with teleoperated arms: the intervention applications. To allow these heavy vehicles controlled by human operators to perform intervention tasks, underwater structures like observatory facilities, subsea panels or oil-well Christmas trees have been adapted, making them more robust and easier to operate. The TRITON Spanish founded project proposes the use of a light-weight intervention AUV (I-AUV) to carry out intervention applications simplifying the adaptation of these underwater structures and drastically reducing the operational cost. To prove this concept, the Girona 500 I-AUV is used to autonomously dock into an adapted subsea panel and once docked perform an intervention composed of turning a valve and plugging in/unplugging a connector. The techniques used for the autonomous docking and manipulation as well as the design of an adapted subsea panel with a funnel-based docking system are presented in this article together with the results achieved in a water tank and at sea This work was supported by the Spanish project DPI2014-57746-C3 (MERBOTS Project) and by Generalitat Valenciana under Grant GVA-PROMETEO/2016/066. The University of Girona wants to thank the SARTI group for their collaboration with the TRITON project

We present an approach for navigating in unknown environments while, simultaneously, gathering information for inspecting underwater structures using an autonomous underwater vehicle (AUV). To accomplish this, we first use our pipeline for mapping and planning collision-free paths online, which endows an AUV with the capability to autonomously acquire optical data in close proximity. With that information, we then propose a reconstruction pipeline to create a photo-realistic textured 3D model of the inspected area. These 3D models are also of particular interest to other fields of study in marine sciences, since they can serve as base maps for environmental monitoring, thus allowing change detection of biological communities and their environment over time. Finally, we evaluate our approach using the Sparus II, a torpedo-shaped AUV, conducting inspection missions in a challenging, real-world and natural scenario This work was supported by MORPH, Excellabust, and Roboacademy European projects (FP7-ICT-2011-7-288704, H2020-TWINN-2015 (CSA)-691980, and FP7-PEOPLE-2013-ITN-608096), the ARCHROV Spanish project (DPI2014-57746-C3-3-R), the Generalitat de Catalunya through the ACCIO/TecnioSpring program (TECSPR14-1-0050), and partially supported by the Colombian Government through its Predoctoral Grant Program (No. 568) offered by Colciencias

Intervention autonomous underwater vehicles (I-AUVs) have the potential to open new avenues for the maintenance and monitoring of offshore subsea facilities in a cost-effective way. However, this requires challenging intervention operations to be carried out persistently, thus minimizing human supervision and ensuring a reliable vehicle behaviour under unexpected perturbances and failures. This paper describes a system to perform autonomous intervention—in particular valve-turning—using the concept of persistent autonomy. To achieve this goal, we build a framework that integrates different disciplines, involving mechatronics, localization, control, machine learning and planning techniques, bearing in mind robustness in the implementation of all of them. We present experiments in a water tank, conducted with Girona 500 I-AUV in the context of a multiple intervention mission. Results show how the vehicle sets several valve panel configurations throughout the experiment while handling different errors, either spontaneous or induced. Finally, we report the insights gained from our experience and we discuss the main aspects that must be matured and refined in order to promote the future development of intervention autonomous vehicles that can operate, persistently, in subsea facilities This work has been supported by the FP7-ICT-2011-7 project PANDORA-Persistent Autonomy through Learning, Adaptation, Observation and Re-planning (Ref. 288273) funded by the European Commission

Nowadays have already passed more than two years since PATRICIA research project about using pet robots to reduce pain and anxiety in hospitalized children was started and the advances made are more than significant. Patients, parents, nurses, psychologists, engineers... all of them have adopted with illusion Pleo robot, a kind of robotic pet, working hard looking for new procedures and new solutions to the current pediatrics diseases. From this work, a technological contribution is provided going one step beyond to what seems a long path. Concretely, it is wanted to develop a system to wirelessly communicate with Pleo in order to help the coordinator who leads the therapy with the kid, to understand and control Pleo’s behaviour at any moment. This article explains how this technological part is being developed and obtained technical results.

Simultaneous localization and mapping (SLAM) is an important problem to solve in robotics theory in order to build truly autonomous mobile robots. This work presents a novel method for implementing a SLAM system based on a single camera sensor. The SLAM with a single camera, or monocular SLAM, is probably one of the most complex SLAM variants. In this case, a single camera, which is freely moving through its environment, represents the sole sensor input to the system. The sensors have a large impact on the algorithm used for SLAM. Cameras are used more frequently, because they provide a lot of information and are well adapted for embedded systems: they are light, cheap and power-saving. Nevertheless, and unlike range sensors, which provide range and angular information, a camera is a projective sensor providing only angular measurements of image features. Therefore, depth information (range) cannot be obtained in a single step. In this case, special techniques for feature system-initialization are needed in order to enable the use of angular sensors (as cameras) in SLAM systems. The main contribution of this work is to present a novel and robust scheme for incorporating and measuring visual features in filtering-based monocular SLAM systems. The proposed method is based in a two-step technique, which is intended to exploit all the information available in angular measurements. Unlike previous schemes, the values of parameters used by the initialization technique are derived directly from the sensor characteristics, thus simplifying the tuning of the system. The experimental results show that the proposed method surpasses the performance of previous schemes.

We present a seabed profile estimation and following method for close proximity inspection of 3D underwater structures using autonomous underwater vehicles (AUVs). The presented method is used to determine a path allowing the AUV to pass its sensors over all points of the target structure, which is known as coverage path planning. Our profile following method goes beyond traditional seabed following at a safe altitude and exploits hovering capabilities of recent AUV developments. A range sonar is used to incrementally construct a local probabilistic map representation of the environment and estimates of the local profile are obtained via linear regression. Two behavior-based controllers use these estimates to perform horizontal and vertical profile following. We build upon these tools to address coverage path planning for 3D underwater structures using a (potentially inaccurate) prior map and following cross-section profiles of the target structure. The feasibility of the proposed method is demonstrated using the GIRONA 500 AUV both in simulation using synthetic and real-world bathymetric data and in pool trials This research was sponsored by the Spanish government (COMAROB Project, DPI2011-27977-C03-02) and the MORPH EU FP7-Project under the Grant agreement FP7-ICT-2011-7-288704. We are grateful for this support

Starting in January 2009, the RAUVI (Reconfigurable Autonomous Underwater Vehicle for Intervention Missions) project is a 3-year coordinated research action funded by the Spanish Ministry of Research and Innovation. In this paper, the state of progress after 2 years of continuous research is reported. As a first experimental validation of the complete system, a search and recovery problem is addressed, consisting of finding and recovering a flight data recorder placed at an unknown position at the bottom of a water tank. An overview of the techniques used to successfully solve the problem in an autonomous way is provided. The obtained results are very promising and are the first step toward the final test in shallow water at the end of 2011 This research was partly supported by Spanish Ministry of Research and Innovation DPI2008-06548-C01-02-03 (RAUVI Project), by the European Commissions Seventh Framework Programme FP7/2007-2013 underGrant agreement 248497 (TRIDENT Project), by Fundació Caixa Castelló-Bancaixa P1-1B2009-50, and by the Spanish Ministry of Education and Science (MCINN Grant CTM2010-15216), Ramon Y Cajal Program (N. Gracias) and the FPI program (R. Campos)

When unmanned underwater vehicles (UUVs) perform missions near the ocean floor, optical sensors can be used to improve local navigation. Video mosaics allow to efficiently process the images acquired by the vehicle, and also to obtain position estimates. We discuss in this paper the role of lens distortions in this context, proving that degenerate mosaics have their origin not only in the selected motion model or in registration errors, but also in the cumulative effect of radial distortion residuals. Additionally, we present results on the accuracy of different feature-based approaches for self-correction of lens distortions that may guide the choice of appropriate techniques for correcting distortions.

This paper presents a hybrid behavior-based scheme using reinforcement learning for high-level control of autonomous underwater vehicles (AUVs). Two main features of the presented approach are hybrid behavior coordination and semi on-line neural-Q/spl I.bar/learning (SONQL). Hybrid behavior coordination takes advantages of robustness and modularity in the competitive approach as well as efficient trajectories in the cooperative approach. SONQL, a new continuous approach of the Q/spl I.bar/learning algorithm with a multilayer neural network is used to learn behavior state/action mapping online. Experimental results show the feasibility of the presented approach for AUVs.

This paper proposes a field application of a high-level reinforcement learning (RL) control system for solving the action selection problem of an autonomous robot in cable tracking task. The learning system is characterized by using a direct policy search method for learning the internal state/action mapping. Policy only algorithms may suffer from long convergence times when dealing with real robotics. In order to speed up the process, the learning phase has been carried out in a simulated environment and, in a second step, the policy has been transferred and tested successfully on a real robot. Future steps plan to continue the learning process on-line while on the real robot while performing the mentioned task. We demonstrate its feasibility with real experiments on the underwater robot ICTINEU AUV.

The Robot Operating System (ROS) is a set of software libraries and tools used to build robotic systems. ROS is known for a distributed and modular design. Given a model of the environment, task planning is concerned with the assembly of actions into a structure that is predicted to achieve goals. This can be done in a way that minimises costs, such as time or energy. Task planning is vital in directing the actions of a robotic agent in domains where a causal chain could lock the agent into a dead-end state. Moreover, planning can be used in less constrained domains to provide more intelligent behaviour. This paper describes the ROSPLAN framework, an architecture for embedding task planning into ROS systems. We provide a description of the architecture and a case study in autonomous robotics. Our case study involves autonomous underwater vehicles in scenarios that demonstrate the flexibility and robustness of our approach.

Treball desenvolupat dins el marc del programa 'European Project Semester'. The robotics facilities at Oslo and Akershus University College of Applied Sciences (HiOA) have at their disposal an autonomous humanoid robot by the name of Nao. It is a commercial product developed by SoftBank Robotics for research and educational purposes, and an international project team was assembled for the European Project Semester (EPS) program as a means of assessing its competencies and bolstering them wherever possible. The team set forth with a mission to establish meaningful interactions between the robot and his physical surroundings while placing an emphasis on humanlike behavior to the fullest extent. A goal was soon initiated for Nao to play a rule-based game against a human opponent such that his vast array of available faculties could be tested and put to use in one cohesive demonstration. Achieving this goal meant making advances in artificial intelligence,computer vision and motion planning, first as separate entities that would later converge into a singular yet multifaceted display of abilities. At the end of the project, computer vision has been implemented to the point where pattern recognition can be performed with accuracy amidst awide variety of error-prone detection scenarios. Several artificial intelligence schemes have been explored and compared quantitatively to derive performance metrics for each. Proof-ofconcept movement strategies have also been derived, and best practices for each focus area are provided as a resource for any future researchers or project groups who stand to benefit from the discoveries made in this endeavor

This work presents a complete navigation architecture for an autonomous aerial robot. The proposed scheme consist of: i) a low-level controller for establishing the attitude and position of the vehicle, ii) a Simultaneous Localization and Mapping (SLAM) system, based in bearing (angular) measurements, which gives the robot the ability for navigating in unknown environments, and iii) a high-level motion control system which generates online trajectories. The high-level motion control system (MCS), which represents the main contribution of this work, is inspired by the behavior-based control strategies. The MCS takes as input a very high level mission target (e.g. “explore as much as you can”) and generates online trajectories according to the mission, but at the same time minimizing uncertainty in the estimations in order to maintain the integrity of the robot. The proposed architecture is explained for simplified 3DOF dynamics, but it could be extended in a straightforward manner in order to be applied to full dynamics. Several simulations are included in order to show the performance of the proposed scheme.

In recent years, many researchers have addressed the issue of making Unmanned Aerial Vehicles (UAVs) more and more autonomous. In this context, the state estimation of the vehicle position is a fundamental necessity for any application involving autonomy. However, the problem of position estimation could not be solved in some scenarios, even when a GPS signal is available, for instance, an application requiring performing precision manoeuvres in a complex environment. Therefore, some additional sensory information should be integrated into the system in order to improve accuracy and robustness. In this work, a novel vision-based simultaneous localization and mapping (SLAM) method with application to unmanned aerial vehicles is proposed. One of the contributions of this work is to design and develop a novel technique for estimating features depth which is based on a stochastic technique of triangulation. In the proposed method the camera is mounted over a servo-controlled gimbal that counteracts the changes in attitude of the quadcopter. Due to the above assumption, the overall problem is simplified and it is focused on the position estimation of the aerial vehicle. Also, the tracking process of visual features is made easier due to the stabilized video. Another contribution of this work is to demonstrate that the integration of very noisy GPS measurements into the system for an initial short period of time is enough to initialize the metric scale. The performance of this proposed method is validated by means of experiments with real data carried out in unstructured outdoor environments. A comparative study shows that, when compared with related methods, the proposed approach performs better in terms of accuracy and computational time.

This paper addresses the problem of object recognition from colorless 3D point clouds in underwater environments. It presents a performance comparison of state-of-the-art global descriptors, which are readily available as open source code. The studied methods are intended to assist Autonomous Underwater Vehicles (AUVs) in performing autonomous interventions in underwater Inspection, Maintenance and Repair (IMR) applications. A set of test objects were chosen as being representative of IMR applications whose shape is typically known a priori. As such, CAD models were used to create virtual views of the objects under realistic conditions of added noise and varying resolution. Extensive experiments were conducted from both virtual scans and from real data collected with an AUV equipped with a fast laser sensor developed in our research centre. The underwater testing was conducted from a moving platform, which can create deformations in the perceived shape of the objects. These effects are considerably more difficult to correct than in above-water counterparts, and therefore may affect the performance of the descriptor. Among other conclusions, the testing we conducted illustrated the importance of matching the resolution of the database scans and test scans, as this significantly impacted the performance of all descriptors except one. This paper contributes to the state-of-the-art as being the first work on the comparison and performance evaluation of methods for underwater object recognition. It is also the first effort using comparison of methods for data acquired with a free floating underwater platform This work was supported by the Spanish Government through a FPI Ph.D. grant to K. Himri, as well as by the Spanish Project DPI2017-86372-C3-2-R (TWINBOT-GIRONA1000) and the H2020-INFRAIA-2017-1-twostage-731103 (EUMR)

This paper presents a method to build a semantic map to assist an underwater vehicle-manipulator system in performing intervention tasks autonomously in a submerged man-made pipe structure. The method is based on the integration of feature-based simultaneous localization and mapping (SLAM) and 3D object recognition using a database of a priori known objects. The robot uses Doppler velocity log (DVL), pressure, and attitude and heading reference system (AHRS) sensors for navigation and is equipped with a laser scanner providing non-coloured 3D point clouds of the inspected structure in real time. The object recognition module recognises the pipes and objects within the scan and passes them to the SLAM, which adds them to the map if not yet observed. Otherwise, it uses them to correct the map and the robot navigation if they were already mapped. The SLAM provides a consistent map and a drift-less navigation. Moreover, it provides a global identifier for every observed object instance and its pipe connectivity. This information is fed back to the object recognition module, where it is used to estimate the object classes using Bayesian techniques over the set of those object classes which are compatible in terms of pipe connectivity. This allows fusing of all the already available object observations to improve recognition. The outcome of the process is a semantic map made of pipes connected through valves, elbows and tees conforming to the real structure. Knowing the class and the position of objects will enable high-level manipulation commands in the near future This work was supported by the Spanish Government through a FPI Ph.D. grant to K. Himri, as well as by the Spanish Project DPI2017-86372-C3-2-R (TWINBOT-GIRONA1000) and the European project H2020-INFRAIA-2017-1-twostage-731103 (EUMarineRobots)