ROBOT 2017: Third Iberian Robotics Conference

The aim of this paper is to introduce an innovative machinery result of combining cable suspended robot technology based on parallel kinematics with a traditional gantry crane. This machinery has been named Cablecrane. Its purpose is to keep the same load capabilities as in traditional gantry cranes while enabling full 6 degrees of freedom (DOF) control of the payload.This means that the payload is fully controlled in position and orientation while it is being manipulated. Thus, precision load handling and movement without oscillations are possible in any direction, in any orientation.In addition, combined with appropriate calibration, sensors and integrated CNC controller, most of manipulation tasks in plant can be programmed and automatized.The final result is an increase in production, full plant automation and enhanced plant safety.

To realize human-robot collaboration in manufacturing, industrial robots need to share an environment with humans and to work hand in hand. This introduces safety concerns but also provides the opportunity to take advantage of human-robot interactions to control the robot. The main objective of this work is to provide HRI without compromising safety issues in a realistic industrial context. In the paper, a region-based filtering and reasoning method for safety has been developed and integrated into a human-robot collaboration system. The proposed method has been successfully demonstrated keeping safety during the showcase evaluation of the European robotics challenges with a real mobile manipulator.

There is a lack of quality indexes to evaluate grasps that are more likely to allow a hand-to-hand transfer of an object during a manipulation task. In order to overcome it, this paper presents a proposal of grasp transfer quality measures to evaluate how easy or feasible is that an object grasped by one hand could be grasped by another hand to perform a hand-to-hand transfer. Experiments were conducted to evaluate the proposed grasp transfer quality measures using different objects and the model of a real robotic hand.

In this paper, a customized interface developed in the framework of the ROBOT FASE II project is described. This project aimed at improving the productivity of two FANUC ARCMate 100iC MIG welding robots with R-30iA controllers mounted in an 8 meters-high mobile gantry crane at Navantia company in Puerto Real, Spain. The solution designed for welding application by the University of Cadiz consists of four parts (1) a library of piece templates including relevant information for each piece to be welded, including obviously the typical piece geometry shape and dimensions, but also all parameters needed for welding(sequence, intensity of the arc, waving description,…) and optimized to get a perfect result by a professional welder team (2) a coordinate measuring arm used to capture 3D information from the real world, (3) a software to generate automatically the optimized FANUC welding program using both the template and the 3D information captured by the arm, adapting the template to real-world coordinates and orientation, (4) and finally an FTP interface to transmit the optimized welding program to each robot for immediate welding operation. The use of this solution for welding operation has reduced robot programming time from hours to minutes for a typical structure allowing an important increase of productivity at Navantia company.

This paper presents an automated robotic welding system adapted for shipbuilding in large shipyards. This solution has been devised in the shipyard of Navantia located in the south of Spain, in the context of ROBOT FASE II R&D project. It also presents the human teams that have developed this welding system. The article explains the 3 parts that make up the welding system. The location of the robotic welding arm of the Fanuc brand is detailed in the first part. In addition, the gantry and spatial coordinate axes are described and indicated where the robotic arm is housed. The second part contains the system capture of the coordinates in the space for the reading of the singular points to be soldered. These points are measured by a portable measuring arm. The last part is composed of the system of communication between the different parts throughout the computer. The computer is responsible for synchronizing the measuring arm and the robotic welding arm by translating the points to be soldered.

Cybersecurity in autonomous systems is a growing concern. Currently most research robotic systems are built using ROS framework, along with other commercial software. The goal of this paper is to improve ROS security features by using encrypted communications and semantic rules to ensure a correct behavior. To encrypt communications, Advanced Encryption Standard algorithm has been applied. Then, the framework ROSRV has been used to define semantic rules for ROS messages. In order to test this proposal, two experiments have been carried out: in the first one, plain-text messages are not allowed and must be blocked; in the second one, rules for detecting denial of service attacks are tested against a real attack performed on a Real-Time Locating System, used by a mobile robot to estimate its location.

The calibration of an accelerometer, and a gyroscope is performed. A linear sensor model have been used. The triaxial sensor calibration algorithm is based on the minimization of a cost function, and is performed offline. Calibration hardware has been designed, and used to get the calibration data. The algorithm has been tested with the acquired data, and the calibration results are presented. Although the sensor model is linear, some experiences about dealing with non-linearities are exposed.

This paper proposes a methodology for automatic characterization of phase resetting controllers for quick balance recovery after loss of it during biped locomotion. The system allows to easily characterize and design useful phase resetting controllers using a simulation environment. Several experiments have been performed using a NAO humanoid robot in order to automatically characterize and test the phase resetting mechanism. Notwithstanding, it can be implemented by using any humanoid robot with a similar kinematic structure. Once the controllers are characterized, the proposed system detects the robot’s current state through the information provided by its inertial sensors and then applies the correct phase resetting in a short period of time in order to quickly recover the robot’s balance. The proposed control scheme reacts quickly whenever unknown external perturbations are applied to the robot’s body by using the proposed phase resetting mechanism.

The use of teleoperation systems can provide substantial advantages in various fields such as remote explosive dismantling and rescue operations, as well as improve the accuracy on certain tasks like surgery. These systems offer feedback mainly based on visual information of the task performed showing relevant data of the systems used as well as their surroundings. Our proposal involves the use of haptic devices to remotely control a robot in order for the user to receive haptic feedback alongside the visual information. In a first design that several experts evaluated, it was concluded that the interface used had to be substantially improved. This communication describes the process developed to design and evaluate an interface to improve user interaction with the teleoperation system.

Visual attention plays a central role in natural and artificial systems to control perceptual resources. The classic artificial visual attention systems uses salient features of the image obtained from the information given by predefined filters. Recently, deep neural networks have been developed for recognizing thousands of objects and autonomously generate visual characteristics optimized by training with large data sets. Besides being used for object recognition, these features have been very successful in other visual problems such as object segmentation, tracking and recently, visual attention. In this work we propose a biologically inspired object classification and localization framework that combines Deep Convolutional Neural Networks with foveal vision. First, a feed-forward pass is performed to obtain the predicted class labels. Next, we get the object location proposals by applying a segmentation mask on the saliency map calculated through a top-down backward pass. The main contribution of our work lies in the evaluation of the performances obtained with different non-uniform resolutions. We were able to establish a relationship between performance and the different levels of information preserved by each of the sensing configurations. The results demonstrate that we do not need to store and transmit all the information present on high-resolution images since, beyond a certain amount of preserved information, the performance in the classification and localization task saturates.

There are many open issues and challenges in the reinforcement learning field, such as handling high-dimensional environments. Function approximators, such as deep neural networks, have been successfully used in both single- and multi-agent environments with high dimensional state-spaces. The multi-agent learning paradigm faces even more problems, due to the effect of several agents learning simultaneously in the environment. One of its main concerns is how to learn mixed policies that prevent opponents from exploring them in competitive environments, achieving a Nash equilibrium. We propose an extension of several algorithms able to achieve Nash equilibriums in single-state games to the deep-learning paradigm. We compare their deep-learning and table-based implementations, and demonstrate how WPL is able to achieve an equilibrium strategy in a complex environment, where agents must find each other in an infinite-state game and play a modified version of the Rock Paper Scissors game.

In the context of assistive robotics, robots need to make multiple decisions. We explore the problem where a robot has multiple choices to perform a task and must select the action that maximizes success probability among a repertoire of pre-trained actions. We investigate the case in which sensory data is only available before making the decision, but not while the action is being performed. In this paper we propose to use a Gaussian Mixture Model (GMM) as decision-making system. Our adaptation permits the initialization of the model using only one sample per component. We also propose an algorithm to use the result of each execution to update the model, thus adapting the robot behavior to the user and evaluating the effectiveness of each pre-trained action. The proposed algorithm is applied to a robotic shoe-dressing task. Simulated and real experiments show the validity of our approach.

Pose invariant object detection and classification plays a critical role in robust image recognition systems and can be applied in a multitude of applications, ranging from simple monitoring to advanced tracking. This paper analyzes the usage of the Bag of Words model for recognizing objects in different scales, orientations and perspective views within cluttered environments. The recognition system relies on image analysis techniques, such as feature detection, description and clustering along with machine learning classifiers. For pinpointing the location of the target object, it is proposed a multiscale sliding window approach followed by a dynamic thresholding segmentation. The recognition system was tested with several configurations of feature detectors, descriptors and classifiers and achieved an accuracy of 87% when recognizing cars from an annotated dataset with 177 training images and 177 testing images.

This paper presents the application of machine learning to tactile sensing for rescue robotics. Disaster situations often exhibit low-visibility scenarios where haptic feedback provides a valuable information for the search of potential victims. To extract haptic information from the environment, a tactile sensor attached to a lightweight robotic arm is used. Then, methods based on the SURF descriptor, support vector machines (SVM), Deep Convolutional Neural Networks (DCNN) and transfer learning are implemented to classify the data. Besides, experiments have been carried out, to compare those procedures, using different contact elements, such as human parts and objects that could be found in catastrophe scenarios. The best achieved accuracy of $$92.22\%$$, results from the application of the transfer learning procedure using a pre-trained DCNN and fine-tuning the classification layer of the network.

Missing tools is a problem in aircraft factories. It may reduce the productivity of the assembly line and missing items may cause FOD (Foreign Object Damage) if they are lost inside the aerostructure. This paper proposes a method which uses aerial robots to search and locate missing tools. Each tool will be equipped with a radio tag with an ID that can listen and respond to request messages from the aerial robot. Thus, the robot can take range measurements to the missing tools from different locations while performing other tasks in the factory. The range measurements are used to estimate the location of every missing tool using a Particle Filter (PF) which will eventually converge to an Extended Kalman Filter (EKF). The proposed method was evaluated and validated in real experiments performed in an emulated scenario very similar to the real factory. Preliminary tests were also performed in the Airbus DS CBC factory with good results.

This paper addresses preliminary results of integration of a mobile node in a wireless sensor network for an urban information system. The proposed mobile node has been designed with a modular and scalable architecture that allows changing the set of sensors in short time. An implementation of the mobile node, including sensors for gas concentrations and environmental parameters, has been installed on an electric vehicle and tested in real scenarios in the city of Malaga.

This paper presents the design and initial experimentation of a novel robot-sensor network security architecture that exploits the synergies between robots and sensor networks to provide high security level with moderate resource consumption. The robot implements security functionalities that in traditional schemes are assumed by the sensor nodes and the Base Station. In contrast to traditional sensor network security schemes, it is not sensor nodes but the robot who discovers other sensor nodes and establishes the network topology, involving important security advantages. This paper presents the design of the architecture, its main advantages and shows its validity in initial field experiments.

A Robust Model Predictive Control (MPC) scheme for the control of formations of Autonomous Underwater Vehicles (AUVs) is presented and discussed. This application domain is extremely relevant and exhibits very difficult control challenges: (i) slow, low data-rate acoustic communications, (ii) significant perturbations inherent to the hydrodynamic environment, (iii) unexpected emergence of obstacles, and (iv) severe onboard computation constraints. While the later aspect is discussed by the Reach Set MPC scheme implementation which maximizes the a priori off-line computation as enabled by taking into account, as much as possible, invariant data, the other challenges are addressed by increasing the robustness of the proposed basic MPC scheme by considering a number of intermediate steps which, in spite of the increase of the on-line computational burden, this remains strongly lower than the one associated with typical standard MPC schemes.

This paper is focused on creating semantic 3D models of unstructured warehouses from coloured point clouds. Several scans from different locations of a laser scanner are integrated into a unique 3D dataset that is afterwards processed. The paper presents an efficient 3D processing algorithm that is able to segment, recognize and locate the existing materials of a storage place. The obtained 3D model provides to logistic managers precise and valuable information, such as: the current location of the stock, the free space for coming merchandises or the occupied volume variations between two scanning sessions taken at different times. The method has been tested under noise conditions in simulated scenarios and the extracted model has been compared with a ground truth model. The good results demonstrate that this approach could be useful in the logistic field.

In this work, a crop inspection system is presented. A mobile platform, based on a commercial electric vehicle, is equipped with different on-board sensors to inspection annual crops (maize, cereal, etc.) and multi-annual crops (orchards, vineyards, etc.). The use of a low-cost RGB-D sensor, the Microsoft Kinect v2 sensor, for the inspection of woody crops is tested. A method to generate automatic 3D reconstructions of large areas, such as a complete crop row, from the information directly supplied by the RGB-D sensor is shown as well as a procedure to correct the drift that appears in the reconstruction of crop rows. All these methods were tested and validated in real fields at different times throughout 2016. The development presented in this paper is a promising technology to achieve better crop management, which will increase crop yield.

This work proposes an automatic method to detect ancient symbols written in stone. The proposed method takes into account well-known techniques used in computer vision to identify the contour of the symbols in the image. The two-stage method consists of segmentation and localization processes. Segmentation process includes a pre-processing step, edge detection and thresholding. Localization process is based on two conditions that take into account several parameters, like the distance between points, and the orientation and the continuity of the edges. This proposal has been applied to localize Egyptian cartouches (borders enclosing the name of a king) and stonemason’s marks from images obtained under varying lighting conditions (controlled and natural lighting). The proposed method is compared favorably against other methods based on chain coding, neural networks and statistical correlation. The promising results give new possibilities to identify and recognize complex symbols and ancient texts.

In this paper we outline the euRathlon 2015 and the ERL Emergency 2017 Grand Challenge robotics competitions, and the results and lessons learned from euRathlon 2015. Staged at Piombino, Italy in September 2015, euRathlon 2015 was the world’s first multi-domain (air, land and sea) multi-robot search and rescue competition. In a mock disaster scenario inspired by the 2011 Fukushima NPP accident the euRathlon 2015 Grand Challenge required teams of robots to cooperate to map the area, and missing workers and stem a leak. The second edition of the competition will be held also in Piombino in September 2017, under the name of ERL Emergency and as part of the new European Robotics League initiative.

This paper proposes the whole system scheme designed for the autonomous landing of a multicopter on a moving platform. The technology used for the tracking and landing is the visual detection and recognition of a marker placed on the platform. Both the hardware and software architecture are explained and also the results of some succesful tests are shown. In addition, the proposed system was validated and compared during the MBZIRC robotics competition in March 2017.

The use of maps allows mobile robots to navigate between known points in an environment. Using maps allows to calculate routes avoiding obstacles and not being stuck in dead ends. This paper shows how to integrate 3D perceptions on a map to obtain obstacle-free paths when obstacles are not at the level of 2D sensors, but elevated. Chairs and tables usually pose a problem when one can only see the legs with a 2D laser, although they present a high hurdle with a much larger area. This approach builds a static map starting from the construction plans of a building. A long-term map is started from the static map, and updated when adding and removing furniture, or when doors are opened or closed. A short-term map represents dynamic obstacles such as people. Obstacles are perceived by merging all available information, both 2D laser and RGB-D cameras, into a compact 3D probabilistic representation. This approach is appropriate for fast deployment and long-term operations in office or domestic environments, able to adapt to changes in the environment. This work is designed for domestic environments, and has been tested in the RoboCup@home competition, where robots must navigate in an environment that changes during the tests.

In this paper, we describe a navigation system requiring very few computational resources, but still providing performance comparable with commonly used tools in the ROS universe. This lightweight navigation system is thus suitable for robots with low computational resources and provides interfaces for both ROS and NAOqi middlewares. We have successfully evaluated the software on different robots and in different situations, including SoftBank Pepper robot for RoboCup@Home SSPL competitions and on small home-made robots for RoboCup@Home Education workshops. The developed software is well documented and easy to understand. It is released open-source and as Debian package to facilitate ease of use, in particular for the young researchers participating in robotic competitions and for educational activities.

In this paper it is proposed the use of a Total Station as odometry helper to improve the localization of UAVs for 3D reconstruction of underside of bridges where typically lacks of GPS signal and steel structures might produce interferences on the measures on the UAVs. The information from the Total Station is sent to the UAV through a WIFI network and is fused with data from the IMU as odometry. Robot is also provided with a RGB-D camera which provides pointclouds for building the map.

Fruit harvesting is a topic of intereset in agricultural industries. In order to perform this task, robots should be able to recognise and segment fruit in their perceptual environment. Particularly, apple trees are often arranged as planar trellis structures in commercial orchards. The vine-like branches have leaves that can occlude fruit and produce noise in typical depth sensor data that also populates the scene with objects that are not of interest. In this paper, we present a method that uses a Dirichlet mixture of Gaussian processes and a Gibbs-Sampler for segmenting clusters of apples to support selective autonomous harvesting. Furthermore, the model provides probabilistic reconstruction of the entire apple which can be used for better grasping of the fruit.

This paper proposes the application of a stereo vision system for estimating and controlling the Cartesian and joint deflection in an anthropomorphic, compliant and ultra-lightweight dual arm designed for aerial manipulation. Each arm provides four degrees of freedom (DOF) for end-effector positioning in a human-like kinematic configuration. A simple and compact spring-lever mechanism introduced in all joints provides mechanical compliance to the arms. A color marker attached at the end effector of the arms is visually tracked by a stereo pair installed over the shoulders. The Cartesian position and velocity of the markers is estimated with an Extended Kalman Filter (EKF), while the corresponding points in an equivalent stiff-joint manipulator are obtained from the kinematic model and the position of the servos. The Cartesian deflection is defined as the difference between these two measurements, obtaining the joint deflection from the inverse kinematics. The vision-based deflection estimator is validated in test bench experiments: position estimation accuracy, impact response, passive/active compliance and contact force control.

We propose a novel 3D navigation system for autonomous vehicle path-planning. The system processes a point-cloud data from an RGB-D camera and creates a 3D occupancy grid with adaptable cell size. Occupied grid cells contain normal distribution characterizing the data measured in the area of the cell. The normal distributions are then used for cell classification, traversability, and collision checking. The space of traversable cells is used for path-planning. The ability to work in three-dimensional space allows autonomous robots to operate in highly structured environments with multiple levels, uneven surfaces, and various elevated and underground crossings. That is important for the usage of robots in real-world scenarios such as in urban areas and for disaster rescue missions.

This paper presents Robobo in the context of higher education. Robobo is a low-cost educational mobile robot that combines a simple wheeled base with a smartphone, which provides the latest technology to the robot. With Robobo, students can develop their own projects through ROS using cameras, microphones or high-resolution screens, bringing teaching closer to the real requirements of the market they will find when they finish their studies. In this work, the hardware and software development that has been carried out is described in detail. Furthermore, it is presented an exemplifying case of student project that shows the potentiality of Robobo in this context.

ROS Industrial (ROS-I) is an effort to deploy the Robot Operating System (ROS) for industrial manufacturing applications. The ROS-I activities are organised by the ROS Industrial consortium (RIC). With the EU-funded project ROSIN, which started in 2017, the ROS-I activities are further supported. The project will give out funds for developing ROS-I components. As a further important measure, the ROSIN project focuses on education measures for training a large number of students and industry professionals to become specialists in ROS-I. In this paper, we outline the broad ROSIN education programme, which consists of a series of summer schools, a professional academy and intends to provide the course contents in Massive Open Online Courses as well.

Engineering degrees require a strong background in Physical Sciences and Mathematics, demanding a high level of conceptualization and abstract reasoning that many students do not possess at the entry level of their high education studies. This can cause students demotivation and dropout, a situation that Higher Education institutions have felt the need to cope with. One methodology to address this problem is to introduce the use of robots in the classes. This tool has unique characteristics that may potentially contribute to increase students’ motivation and engagement, which are key factors on their academic success. This paper presents the rationale, challenges and methodology used to introduce robots as a tool to teach introductory electronics to first year students in a Electronics and Telecommunications Engineering Masters degree. The paper also reports evaluation indicators that result from two different surveys, one generic, carried out in the scope of the Quality Assurance System of the University, and another one developed specifically to evaluate the course. The results confirm that there is a clear and overall positive impact. Particularly significant are the gains on the students motivation and subject comprehension, without a noticeable impact on the course difficulty and required effort. It is also specially relevant that students are strongly in favour of keeping robot’s usage due to its impact on both knowledge and motivation.

In 58 Italian Public Comprehensive Institutes (Istituti Comprensivi), that include Primary and Elementary schools, 2911 students experimented the use of a mobile robot, Sapientino Doc by Clementoni, to learn curricula matters such as Mathematics, Geometry and Geography (MGG). The project “A scuola di coding con Sapientino” was developed during the 2016/2017 regular school year for about 3 months (April–June 2017). The schools were distributed throughout Italy and involved 2911 students from 5 to 8 years old, 155 classes, and 163 teachers. The aim of the research is to demonstrate a learning gain in Mathematics, Geometry and Geography, after the students use a mobile robot during regular lessons held by their own teachers in their classrooms. In this paper, we present the methodology used to develop the project and the results of data analysis.

This paper describes an estimator architecture for a Formula Student Prototype, based on data from an inertial measurement unit (IMU), a global positioning system (GPS), and from the underlying dynamic model of the car. A non-linear dynamic model of the car and realistic models for the sensors are presented. The estimates of attitude, rate-gyro bias, position, velocity and sideslip are based on Kalman filtering techniques. The resulting system is validated on a Formula Student prototype and assessed given ground truth data obtained by a set of differential GPS receivers installed onboard.

Torque Vectoring (TV) has the objective to substitute the need of a mechanical differential, while improving the handling and response of the wheeled vehicle. This work addresses the design of a torque vectoring system in an rear wheel driven formula student prototype. The proposed solution resorts to a PID controller for yaw rate tracking with an evenly distributed torque to each wheel. Also an LQR scheme is discussed, for tracking the yaw rate and the lateral velocity. To assess and design, first a 7 degree of freedom (DOF) non linear model is constructed, followed by a linear 2 DOF model, both validated with real data. The linear model, is used to design and simulate the proposed controllers. When the controller is within the desired parameters it is tested in the non linear model. Tests with the vehicle are performed to verify the contribution of the controller to the overall performance of the vehicle.

This paper presents an approach for plan generation, selection and pruning of trajectories for autonomous driving, capable of dealing with dynamic complex environments, such as driving in urban scenarios. The planner first discretizes the plan space and searches for the best trajectory and velocity profile of the vehicle. The main contributions of this work are the use of $$G^{2}$$-splines for path generation and a method that takes into account accelerations and passenger comfort for generating and pruning velocity profiles based on 3rd order splines, both fulfilling kinodynamic constraints. The proposed methods have been implemented in a motion planner in MATLAB and tested through simulation in different representative scenarios, involving obstacles and other moving vehicles. The simulations show that the planner performs correctly in different dynamic scenarios, maintaining the passenger comfort.

Battery Electric Vehicles have high potentials for the modern transportations, however, they are facing limited cruising range. To address this limitation, we present a semi-autonomous ecological driver assistance system to regulate the velocity with energy-efficient techniques. The main contribution of this paper is the design of a real-time nonlinear receding horizon optimal controller to plan the online cost-effective cruising velocity. Instead of conventional $$\ell _2$$-norms, a deadzone-quadratic penalty function for the nonlinear model predictive controller is proposed. Obtained field experimental results demonstrate the effectiveness of the proposed method for a semi-autonomous electric vehicle in terms of real-time energy-efficient velocity regulation and constraints satisfaction.

The use of the odometry and SLAM visual methods in autonomous vehicles has been growing. Optical sensors provide valuable information from the scenario that enhance the navigation of autonomous vehicles. Although several visual techniques are already available in the literature, their performance could be significantly affected by the scene captured by the optical sensor. In this context, this paper presents a comparative analysis of three monocular visual odometry methods and three stereo SLAM techniques. The advantages, particularities and performance of each technique are discussed, to provide information that is relevant for the development of new research and novel robotic applications.

This paper addresses the problem of vehicle detection using a little explored LIDAR’s modality: the reflection intensity. LIDAR reflection measures the ratio of the received beam sent to a surface, which depends upon the distance, material, and the angle between surface normal and the ray. Considering a 3D-LIDAR mounted on board a robotic vehicle, which is calibrated with respect to a monocular camera, a Dense Reflection Map (DRM) is generated from the projected sparse LIDAR’s reflectance intensity, and inputted to a Deep Convolutional Neural Network (ConvNet) object detection framework for the vehicle detection. The performance on the KITTI is superior to some of the approaches that use LIDAR’s range-value, and hence it demonstrates the usability of LIDAR’s reflection for vehicle detection.

Object identification in images taken from moving vehicles is still a complex task within the computer vision field due to the dynamism of the scenes and the poorly defined structures of the environment. This research proposes an efficient approach to perform recognition on images from a stereo camera, with the goal of gaining insight of traffic scenes in urban and road environments. We rely on a deep learning framework able to simultaneously identify a broad range of entities, such as vehicles, pedestrians or cyclists, with a frame rate compatible with the strong requirements of onboard automotive applications. The results demonstrate the capabilities of the perception system for a wide variety of situations, thus providing valuable information to understand the traffic scenario.

In the Intelligent Transportation Systems Society (ITSS), the research interest on intelligent vehicles is increasing during the last few years. Accordingly, this paper presents the advances in the development of the ROS-based (Robot Operating System) software architecture for intelligent vehicles. The main contribution of the architecture is its powerfulness, flexibility, and modularity, which allows the researchers to develop and test different algorithms. The architecture has been tested on different platforms, autonomous ground vehicles from the iCab (Intelligent Campus Automobile) project and in the intelligent vehicle based on Advanced Driver Assistance Systems (ADAS) incorporated from IvvI 2.0 (Intelligent Vehicle based on Visual Information) project.

The emerging technologies that are conforming the Industry 4.0 are also impacting on health. Artificial intelligence, 3D printing, robotics, big data, Internet of Things, augmented reality, among others, are adding a layer of digitization on classical processes, allowing to increase the effectiveness and efficiency in the processes related to health and opening a new space of possibilities. In this article, some examples will show the state of art of Health 4.0 in the non-surgical field.

This works presents the impact that collaborative robotic technologies can offer for surgical applications, with emphasis on the tracking and execution steps. In particular, a new workflow for spine and trauma surgery is presented, in which a miniature mechanical tracker is attached directly to the patients’ bony structure (Patent pending). The tracker is capable of following the patients’ motion with high precision, measuring the deviation with respect to the trajectories defined in the surgical plan and providing a feedback channel to a robot which assists the surgeon holding the surgical tools in place. The clinical application of vertebral fusion has been chosen as testing scenario and preliminary results are presented to demonstrate the feasibility of this concept.

The progress of technology and robotics in industry has not yield to an equivalent development in the medical field. This paper analyses surgical procedures from the point of view of industrial processes looking for analogies in both fields so as to evaluate the possibilities of using equivalent technologies in both of them. After analyzing surgical specialties from the mechanical point of view, the actions to be performed, and the main requirements as precision and working conditions, a look at the main challenges that surgical robotics should face is presented.

Hand-assisted laparoscopic surgery is a Minimally Invasive Surgery technique that is based on the insertion of one surgeon’s hand inside the abdominal cavity. In this scenario, a robotic assistant can properly collaborate with the surgeon, working side by side with him/her. This paper presents a robotic system for this kind of technique, based on a cognitive architecture that makes possible an efficient collaboration with the surgeon, thanks to a better understanding of the environment and the learning mechanisms included. This architecture includes a hand gesture recognition module and two different autonomous movement of the robotic arms, one for the camera motion and the other for the tool movement. All of these modules take advantage of the cognitive learning mechanisms of the architecture, fitting their behavior to the current user and procedure.

This paper presents the mechanical design of a novel hand exoskeleton for assistance and rehabilitation therapies. As a solution for the movement transmission, the proposed device uses modular linkage that are attached to each finger by means of snap-in fixations. The linkage is kinematically and dynamically analyzed by means of simulations with AnyBody Simulation Software to obtain an estimation of the range of motion and admissible forces. In order to check the deviations of the real performance respect to the simulated results, due to uncertain variables, a first prototype is built and tested.

Recent projects highlight how motor learning and a high level of attention control can potentially improve submaximal force production during recovery after stroke. This study focuses on the assessment of detailed metrics of force production and position control -healthy subjects- and their correlation with submaximal force production control learning during a new task consisting in maintaining the position for early rehabilitation after stroke.We used a Motorized Ankle Foot Orthosis (MAFO) with zero-torque control together with a visual paradigm interface to exert controlled torque profiles to the ankle of the subject. The subject is asked to follow the trajectories in the visual interface, while the robot disturbs the movement. The aim of the exercise is to improve the motor control by learning how to maintain the position to follow the trajectory, compensating the perturbations, in three possible training paradigms: (1) fixed torque, (2) progressive increase of torque, and (3) modulated torque based on score on the task.All training paradigms led to an improvement in the score comparing pre and post-training performances, so we concluded that this platform induces learning on healthy subjects.To sum up, we conclude that this tool is useful to induce learning in healthy subjects, and thus will keep improving the training paradigms, for the translation into a rehabilitative tool.

Human Papilloma Virus (HPV) could develop precancerous lesions and invasive cancer, as it is the main cause of nearly all cases of cervical cancer. There are many strains of HPV and current vaccines can only protect against some of them. This makes the detection and genotyping of HPV a research area of utmost importance. Several biomedical systems can detect HPV in DNA samples; however, most of them do not have a procedure as fast, automatic or precise as it is actually needed in this field. This manuscript presents a novel XML-based hierarchical protocol architecture for biomedical robots to describe each protocol step and execute it sequentially, along with a robust and automatic robotic system for HPV DNA detection capable of processing from 1 to 24 samples simultaneously in a fast (from 45 to 162 min), efficient (100% markers effectiveness) and precise (able to detect 36 different HPV genotypes) way. It includes an efficient artificial vision process as the last step of the diagnostic.

Hand-guiding is a main functionality of collaborative robots, allowing to rapidly and intuitively interact and program a robot. Many applications require end-effector precision positioning during the teaching process. This paper presents a novel method for precision hand-guiding at the end-effector level. From the end-effector force/torque measurements the hand-guiding force/torque (HGFT) is achieved by compensating for the tools weight/inertia. Inspired by the motion properties of a passive mechanical system, mass subjected to coulomb/viscous friction, it was implemented a control scheme to govern the linear/angular motion of the decoupled end-effector. Experimental tests were conducted in a KUKA iiwa robot in an assembly operation.

The current state of technology permits very accurate 3D reconstructions of real scenes acquiring information through quite different sensors altogether. A high precision modelling that allows simulating any element of the environment on virtual interfaces has also been achieved. This paper illustrates a methodology to correctly model a 3D reconstructed scene, with either a camera RGB-D or a laser, and how to integrate and display it in virtual reality environments based on Unity, as well as a comparison between both results. The main interest regarding this line of research consists in the automation of all the process from the map generation to its visualisation with the VR glasses, although this first approach only managed to get results using several programs manually. The long-term objective would be indeed a real-time immersion in Unity interacting with the scene seen by the camera.

This paper proposes a logistic planner with supplementary 3D spatial representations to enhance and interact with traditional logistic systems on the context of mobile manipulators performing internal logistics operations. By defining a hierarchical structure, the logistic world model, as the central entity synchronized between multiple system components, the reliability and accuracy of the logistic system is strengthened. The proposed approach aims at implementing a robust and intuitive solution for the set-up of mobile manipulator based logistic systems. The logistic planner includes a web based interface for fast setup of the warehouse layout based on robot sensing, as well as the definition of missions for the fleet of robotic systems.

The rapid growth of e-commerce increases the need for larger warehouses and their automation, thus using robots as assistants to human workers becomes a priority. In order to operate efficiently and safely, robot assistants or the supervising system should recognize human intentions. Theory of mind (ToM) is an intuitive conception of other agents’ mental state, i.e., beliefs and desires, and how they cause behavior. In this paper we present a ToM-based algorithm for human intention recognition in flexible robotized warehouses. We have placed the warehouse worker in a simulated 2D environment with three potential goals. We observe agent’s actions and validate them with respect to the goal locations using a Markov decision process framework. Those observations are then processed by the proposed hidden Markov model framework which estimated agent’s desires. We demonstrate that the proposed framework predicts human warehouse worker’s desires in an intuitive manner and in the end we discuss the simulation results.

The new paradigms of Industry 4.0 demand the collaboration between robot and humans. They could help and collaborate each other without any additional safety unlike other manipulators. The robot should have the ability of acquire the environment and plan (or re-plan) on-the-fly the movement avoiding the obstacles and people. This paper proposes a system that acquires the environment space, based on a kinect sensor, performs the path planning of a UR5 manipulator for pick and place tasks while avoiding the objects, based on the point cloud from kinect. Results allow to validate the proposed system.

Some industries have critical areas (dangerous or hazardous) where the presence of a human must be reduced or avoided. In some cases, there are areas where humans should be replaced by robots. The present work uses a robot with differential drive to scan an environment with known and unknown obstacles, defined in 3D simulation. It is important that the robot be able to make the right decisions about its way without the need of an operator. A solution to this challenge will be presented in this paper. The control application and its communication module with a simulator or a real robot are proposed. The robot can perform the scan, passing through all the waypoints arranged in a grid. The results are presented, showcasing the robot’s capacity to perform a viable trajectory without human intervention.

The need for efficient automation methods has prompted the fast development in the field of Robotics. However, most robotic solutions found in industrial environments lack in both flexibility and adaptability to be applied to any generic task. A particular problem arises when robots are integrated in work cells with extra degrees of freedom, such as external axis or positioners. The specification/design of high redundancy systems, including robot selection, tool and fixture design, is a multi-variable problem with strong influence in the final performance of the work cell. This work builds on top of optimisation techniques to deal with the optimal poses reachability for high redundancy robotic systems. In this paper, it will be proposed a poses optimisation approach to be applicable within high redundancy robotic systems. The proposed methodology was validated by using real environment existent infrastructures, namely, the national CoopWeld project.

The use of robotic palletizing systems has been increasing in the so-called Fast Moving Consumer Goods (FMCG) industry. However, because of the type of solutions developed, focused on high performance and efficiency, the degree of adaptability of packaging solutions from one type of product to another is extremely low. This is a relevant problem, since companies are changing their production processes from low variety/high volume to high variety/low volume. This environment requires companies to perform the setup of their robots more frequently, which has been leading to the need to use offline programming tools that decrease the required robot stop time. This work addresses these problems and, in this paper, is described the solution proposed for the automated offline development of collision free robot programs for palletizing applications.

The estimation of object’s weight is a very challenging problem due to limitations on tactile technology and robust and fast controllers that can adapt to friction state changes. In this article we study the weight estimation using skin-like tactile sensors, which provide accurate 3 dimensional force measurements on the finger tips of Vizzy, a robot with human-like hands. The sensor reading from the fingertips are analyzed and segmented in order to find the most adequate stages of the movement for the computation of the weight. The analysis is based on the difference of the load force between the grasping and holding up stages, which provides a good estimation of the object’s weight for different objects and various repetitions of the experiments.

Control of flexible link parallel manipulators is still an open area of research. The flexibility and deformations of the limbs make the estimation of the Tool Center Point (TCP) position a non-trivial area, being one of the main challenges on this type of robots. In the literature different approaches to estimate this deformation and determine the location of the TCP have been proposed. However, most of these approaches require the use of high computational cost integration methods or expensive measurement systems. This work presents a novel approach which can not only estimate precisely the deformation of the flexible links (less than 3% error), but also its derivatives (less than 4% error). The validity of the developed estimator is tested in a Delta Robot, resulting in less than 0.025% error in the estimation of the TCP position in comparison with the results obtained with ADAMS Multibody software.

This paper presents a method to estimate the pose of an object inside a robotic hand by exploiting contact and joint position information. Once an initial visual estimation is provided, a Bootstrap Particle Filter is used to evaluate multiple hypothesis for the object pose. The function used to score the hypothesis considers feasibility and physical meaning of the contacts between the object and the hand. The method provides a good estimation of in-hand pose for different 3D objects.

This paper presents a study about gait patterns for hexapod robots with extremities called C-legs. The study analyses several modes of gait that different animals use to move through the terrestrial environment, and another new ones that arise when looking at the limitations that present the existing ones. The whole study is reinforced with a series of simulations carried out, where the obtained results are analysed to select the best gait pattern for a specific situation.

Developing an optimized omnidirectional walking for biped robots is a challenging task due to their complex dynamics and kinematics. This paper proposes a hierarchical walk engine structure to generate fast and stable walking. In particular, this structure provides a closed-loop CPG-based omnidirectional walking that takes into account two human-inspired push recovery strategies. In addition, this structure is fully parametric and allows using a policy search algorithm to find the optimum parameters for the walking. To show the performance of the proposed structure, a set of experiments on a simulated NAO robot has been carried out. Experimental results demonstrate that the proposed structure is able to generate fast and stable omnidirectional walking. The maximum speed of forward walking that we have achieved was 59 cm/s.

Due to the difficulty of building and making control tests in real robots, it is usual to first have a simulated model that provides a good approach of a real robot’s behaviour. The importance of a good control system in execution of a planned trajectory inspired this work, whose purpose is to design a control system for a quadruped robot and test its performance.

The work presented focuses in the simulations and real experiments of perimeter surveillance under communication constraints, performed by teams of UAVs using a Bluetooth communication framework. When UAVs work in a colaborative manner, communication among them is essential to properly perform their task. Moreover, energy consumption and weight of the devices equipped in a UAV are important to be reduced at minimum possible, particularly in micro-UAVs. A coordination variables strategy is implemented to perform the perimeter division.

In this paper we present an approach to connect a multi-robot system of unmanned ground and aerial vehicles inside a mobile ad-hoc network to exchange sensor data requiring a high bandwidth. The introduced communication layer extends the messaging system of an underlying middleware (ROS) to work with non-ideal wireless links. Every connection can be configured with a user defined data rate, protocol and priority to control the overall traffic load in the network. If required, messages can be bufferd during communication outages or due to low available bandwidth. Moreover, the message transport system contains mechanisms for lossless and lossy data compression as well as an user interface to quickly react to different application conditions.

In the present paper we address the problem of trajectory planning for scenarios in which some robot has to exchange information with other moving robots for at least a certain time, determined by the amount of information. We are particularly focused on scenarios where a team of robots must be deployed, reaching some locations to make observations of the environment. The information gathered by all the robots must be shared with an operation center (OP), thus some robots are devoted to retransmit to the OP the data of their teammates. We develop a trajectory planning method called Time-Constrained RRT (TC-RRT). It computes trajectories to reach the assigned primary goals, but subjected to the constraint determined by the need of communicating with another robot acting as moving relay, just during the time it takes to fulfill the data exchange. Against other methods in the literature, using this method it is not needed a task allocator to assign beforehand specific meeting points or areas for communication exchange, because the planner finds the best area to do it, simultaneously minimizing the time to reach the goal. Evaluation and limitations of the technique are presented for different system parameters.

Unmanned Aerial Vehicles (UAVs) are becoming an important tool for facilities monitoring, target tracking, surveillance, etc. In many of these contexts, a UAV needs to reach an area of interest (AOI) while streaming data to a ground station (GS) where one or more drones are controlled by an operator. In remote areas, intermediate UAVs can act as relays and form a line network to extend range. Interactive control requires a live video stream where both throughput and delay are important, which limits the number of relays and thus maximum range. However, throughput also depends on the quality of the links. It is known that in open space scenarios, where links can be considered similar, maximum throughput is achieved with equal spacing of the relays. This is not the case near large obstacles or high interference areas, in which links can be rather asymmetric. In this paper, we address this scenario and show that maximum throughput is achieved with uneven relay spacing but balanced packet delivery ratios. We propose a new distributed protocol fit to high throughput streams on line networks that tracks and balances the quality of neighboring links by moving the relay nodes accordingly. Real world experiments with multiple AR.Drone 2.0 UAVs show that variable link length in a scenario with asymmetric links generates gains in packet delivery up to $$40\%$$ and throughput up to $$300\%$$.

Tunnel-like environments represent a challenge in terms of robot localization due to the special conditions of this type of scenarios. Traditional robot localization systems based on laser, camera or odometry sensors do not perform well due to the hostile conditions and the general lack of distinctive features. In this paper we present a discrete localization system which takes advantage of the periodicity nature of the RF signal fadings that appears inside tunnels under certain configurations. Experimental results show the feasibility of the proposed method in order to periodically correct the robot position during its displacement along the tunnel.

Sensor networks can monitor wide areas to gather information and relay alerts about concerning events. Response robotic missions in confined scenarios where such a network existed, like tunnels or mines, could take advantage of it as a backbone. This paper addresses challenges arising from the combined characteristics of nodes, typically of low power and bandwidth, and signal propagation in such scenarios, that exhibits extended range but also blind spots due to waveguide self-interference. Firstly, a measurement campaign is reported that characterized RSSI and PDR performance of XBee nodes in the Somport tunnel, enabling improved placement of nodes. Then, a teleoperation mission is demonstrated in which a mobile robot that relays rangefinder readings is commanded thought a backbone multi-hop network, within the restrictions of the extremely limited bandwidth of the IEEE 802.15.4 protocol.

In tunnel-like environments such as sewers, road tunnels and mines, a robot or team of networked mobile robots can provide monitoring services, surveillance or search and rescue. However, these scenarios pose multiple challenges from the robotics and from the communication points of view. Structurally, sewers are GPS-denied and narrow scenarios with lack of illumination and presence of sediments, and in the communication context, the multipath propagation causes strong fading phenomena. In this work we characterize a sewer scenario from the communications point of view, based on a measuring campaign carried out in the sewers of the city of Barcelona, Spain, in the context of an ECHORD++ project towards future inspection with drones.