Robot 2015: Second Iberian Robotics Conference

Relative positioning systems play a vital role in current multi-robot systems. We present a self-contained detection and tracking approach, where a robot estimates a distance (range) and an angle (bearing) to another robot using measurements extracted from the raw data provided by two laser range finders. We propose a method based on the detection of circular features with least-squares fitting and filtering out outliers using a map-based selection. We improve the estimate of the relative robot position and reduce its uncertainty by feeding measurements into a Kalman filter, resulting in an accurate tracking system. We evaluate the performance of the algorithm in a realistic indoor environment to demonstrate its robustness and reliability.

The development of an assistive robotic platform poses exciting engineering and design challenges due to the diversity of possible applications. This article introduces Vizzy, a wheeled humanoid robot with an anthropomorphic upper torso, that combines easy mobility, grasping ability, human-like visual perception, eye-head movements and arm gestures. The humanoid appearance improves user acceptance and facilitates interaction. The lower body mobile platform is able to navigate both indoors and outdoors. We describe the requirements, design and construction of Vizzy, as well as its current cognitive capabilities and envisaged domains of application.

The paper addresses terrain modeling for mobile robots with fuzzy elevation maps by improving computational speed and performance over previous work on fuzzy terrain identification from a three-dimensional (3D) scan. To this end, spherical sub-sampling of the raw scan is proposed to select training data that does not filter out salient obstacles. Besides, rule structure is systematically defined by considering triangular sets with an unevenly distributed standard fuzzy partition and zero order Sugeno-type consequents. This structure, which favors a faster training time and reduces the number of rule parameters, also serves to compute a fuzzy reliability mask for the continuous fuzzy surface. The paper offers a case study using a Hokuyo-based 3D rangefinder to model terrain with and without outstanding obstacles. Performance regarding error and model size are compared favorably with respect to a solution that uses quadric-based surface simplification (QSlim).

Motion planning has evolved from coping with simply geometric problems to physics-based ones that incorporate the kinodynamic and the physical constraints imposed by the robot and the physical world. Therefore, the criteria for evaluating physics-based motion planners goes beyond the computational complexity (e.g. in terms of planning time) usually used as a measure for evaluating geometrical planners, in order to consider also the quality of the solution in terms of dynamical parameters. This study proposes an evaluation criteria and analyzes the performance of several kinodynamic planners, which are at the core of physics-based motion planning, using different scenarios with fixed and manipulatable objects. RRT, EST, KPIECE and SyCLoP are used for the benchmarking. The results show that KPIECE computes the time-optimal solution with heighest success rate, whereas, SyCLoP compute the most power-optimal solution among the planners used.

Nowadays, the utilization of multiple sensors on every robotic platform is a reality due to their low cost, small size and light weight. Multi Sensor Fusion (MSF) algorithms are required to take advance of all the given measurements in a robust and complete manner. These high demanding developed algorithms need to be tested under real and challenging situations and environments. To the knowledge of the authors, none of the available datasets fulfills are fully suitable to be used for MSF applied to Robotics, due to the lack of multiple sensor measurements provided by light weight, small size and low cost sensors; due to their restricted motions (typically planar movements); or to their limited environmental conditions.The contributions of the paper are twofold. First, a low-cost portable and versatile testbed has been developed for MSF research with various types of sensors. Second, a group of datasets for MSF research for Robotics have been made public as a common framework for algorithm testing after a comparison with the existing databases in the state of the art, highlighting the differences and advantages of the one presented in this paper, that are: low-cost sensors for general use on Robotics, fully 3 dimensional movements (six degrees of freedom), as well as challenging indoor and outdoor small and large environments.

To cope with the growing complexity of manipulation tasks, the way to combine and access information from high- and low-planning levels has recently emerged as an interesting challenge in robotics. To tackle this, the present paper first represents the manipulation problem, involving knowledge about the world and the planning phase, in the form of an ontology. It also addresses a high-level and a low-level reasoning processes, this latter related with physics-based issues, aiming to appraise manipulation actions and prune the task planning phase from dispensable actions. In addition, a procedure is contributed to run these two-level reasoning processes simultaneously in order to make task planning more efficient. Eventually, the proposed planning approach is implemented and simulated through an example.

Grasping and manipulating objects with robotic hands depend largely on the features of the object to be used. Especially softness and deformability are crucial features to take into account during the manipulation tasks. Positions of the fingers and forces to be applied when manipulating an object are adapted to the caused deformation. For unknown objects, a previous recognition stage is needed to set features of the object, and manipulation strategies can be adapted depending on that recognition stage. This paper presents an adaptable tactile servoing control scheme that can be used in manipulation tasks of deformable objects. Tactile control is based on maintaining a force value at the contact points which changes according to the object softness, a feature estimated in an initial stage.

This paper presents the application of the Fast Marching Method, with or without an external vectorial field, to the path planning problem of robots in difficult outdoors environments. The resulting trajectory has to take into account the obstacles, the slope of the terrain (gradient of the height), the roughness (spherical variance) and the type of terrain (presence of sand) that can lead to slidings. When the robot is in sandy terrain with a certain slope, there is a landslide (usually small) that can be modelled as a lateral current or vectorial field in the direction of the negative gradient. Besides, the method calculates a weight matrix W that represents difficulty for the robot to move in certain terrain and is built based on the information extracted from the surface characteristics. Then, the Fast Marching Method is applied with matrix W being a velocities map. Finally, the algorithm has been modified to incorporate the effect of an external vectorial field.

Autonomous mobile service robots move in our buildings, carrying out different tasks across multiple floors. While moving and performing their tasks, these robots find themselves in a variety of states. Although speech is often used for communicating the robot’s state to humans, such communication can often be ineffective. We investigate the use of lights as a persistent visualization of the robot’s state in relation to both tasks and environmental factors. Programmable lights offer a large degree of choices in terms of animation pattern, color and speed. We present this space of choices and introduce different animation profiles that we consider to animate a set of programmable lights on the robot. We conduct experiments to query about suitable animations for three representative scenarios of our autonomous symbiotic robot, CoBot. Our work enables CoBot to make its state persistently visible to humans.

This paper deals with accurate attitude estimation in unmanned ground vehicles using low-cost inertial measurement units. Using Euler angles representation, direct estimations are firstly performed from a single sensor, accelerometer or gyroscope. Then, the low frequency components of the first one and the high frequency components of the second one are fused through an explicit complementary filter (ECF), which uses the quaternion representation. A feedback control structure implements the ECF whose controller parameters determine the filter cut-off frequencies. Finally, a scheduling of controllers in the ECF structure overcomes the shortcomings of accelerometer direct estimation at low frequencies. It provides reliable attitude, although the vehicle movement is accelerated. Illustrative experiments are driven with a Traxxas Car equipped with an Ardupilot Mega 2.5 board.

Laser scanners are state-of-the-art devices used for mapping in service, industry, medical and rescue robotics. Although a lot of work has been done in laser-based SLAM, maps still suffer from interferences caused by objects like glass, mirrors and shiny or translucent surfaces. Depending on the surface’s reflectivity, a laser beam is deflected such that returned measurements provide wrong distance data. At certain positions phantom-like objects appear. This paper describes a specular reflectance detection approach applicable to the emerging technology of multi-echo laser scanners in order to identify and filter reflective objects. Two filter stages are implemented. The first filter reduces errors in current scans on the fly. A second filter evaluates a set of laser scans, triggered as soon as a reflective surface has been passed. This makes the reflective surface detection more robust and is used to refine the registered map. Experiments demonstrate the detection and elimination of reflection errors. They show improved localization and mapping in environments containing mirrors and large glass fronts is improved.

This paper presents a study on various types of attacks on the security of a robotic system. The work focuses on hardware attacks, and particularly considers vulnerable features of the I2C protocol. An analysis of the effects of these attacks, when they are applied on a differential driving mobile robot that aims to track a path, is presented. The paper shows how, with such actions, it is possible to modify the behavior of the robot without leaving traces of the attack and also maintaining the system without any damage.

The humanoid robot Nao is a great platform for robotics research, in particular it provides an important testbed for computer vision, machine learning and human robot interface. Nevertheless, its limited sensorization and computation power reduces its autonomy severely. To overcome some of these limitations, in this paper we describe the integration of a RGB-D camera together with a mini-PC, into the Nao robot. Our objective is to get a Nao robot being able to carry out autonomously (onboard) all the tasks that involve 3D environment perception. As an example we used two applications: (1) mimic of human movements, (which will be used on learning by demonstration), and (2) RTABmap SLAM algorithm. Finally, we also tested the Nao’s walking stability when it was equipped with all the new elements.

Processing natural language instructions for execution of robotic tasks has been regarded as a means to make more intuitive the interaction with robots. This paper is focused on the applications of natural language processing in manipulation, specifically on the problem of recovering from the instruction the information missing for the manipulation planning, which has been traditionally assumed to be available for instance via pre-computed grasps or pre-labeled objects. The proposed approach includes a clustering process that discriminates areas on the object that can be used for different types of tasks (therefore providing valuable information for the grasp planning process), the extraction and consideration of task information and grasp constraints for solving the manipulation problem, and the use of an integrated grasp and motion planning that avoids relying on a predefined grasp database.

In this paper we present a new concept of robot-dependent reachability map (RDReachMap) for mobile platforms. In heterogeneous multi-robot systems, the reachability limit of robots motion and actuation must be considered when assigning tasks to them. We created an algorithm that generates those reachability maps, separating regions that can be covered by a robot from the unreachable ones, using morphological operations. Our method is dependent on the robot position, and is parameterized with the robot’s size and actuation radius. For this purpose we introduce a new technique, the partial morphological closing operation. The algorithm was tested both in simulated and real environment maps. We also present a common problem of multi robot routing, which we solve with a planner that uses our reachability maps in order to generate valid plans. We contribute a heuristic that generates paths for two robots using the reachability concept.

This paper describes the challenges that involve playing with the Nao humanoid robot on a tablet. For that purpose, an inverse kinematic solver that allows the robot to move it’s limbs, and a computer vision algorithm that allows the robot to understand the items displayed on the tablet, are needed. The presented solution uses NAOqi’s Cartesian Control and OpenCV’s Hough Transform respectively. To overcome the lack of force and tactile sensors on Nao’s hand, we propose a touch movement based on visual feedback. As an initial approach, we chose the Tic-Tac-Toe game and we introduced interaction mechanisms to make it more pleasant and enjoyable, with the objective of creating a template for HRI and machine learning integration. The experimental results show the robustness of the proposed architecture.

This paper deals with the stability analysis of a Cooperative Adaptive Cruise Control system (CACC) in a convoy of vehicles, in order to solve traffic congestion. The differences between the Adaptive Cruise Control (ACC) and its cooperative version will be studied from the string stability point of view, highlighting the advantages of adding Vehicle-to-Vehicle (V2V) communications. The system will be tested in simulation using MATLAB/Simulink for a convoy of three vehicles and, after that, the control system will be translated to real experimentation for a convoy of three Pioneer robots.

This paper describes the design and development of a biologically inspired flying robot prototype (a machine able to fly by beating its wings, as birds do). For its implementation, the flight of biological beings was analysed, as well as the techniques involved in ornithopter’s construction. Some parameters adopted by biological beings to maintain a stable flight were studied, and the prototype was designed based on these values. To conclude the project, an ornithopter was built, aiming to perform a stabilized flight and some preliminary experiments were performed to check if its behaviour meets the design expectations.

Building inspection is a critical issue for designing rehabilitation projects, which are recently gaining importance for environmental and energy efficiency reasons. Image sensors on-board unmanned aerial vehicles are a powerful tool for building inspection, given the diversity and complexity of façades and materials, and mainly, their vertical disposition. The UBRISTES (UAV-based Building Rehabilitation with vISible and ThErmal infrared remote Sensing) system is proposed as an effective solution for façade inspection in urban areas, validating a method for the simultaneous acquisition of visible and thermal aerial imaging applied to the detection of the main types of façade anomalies/pathologies, and showcasing its possibilities using a first principles analysis. Two public buildings have been considered for evaluating the proposed system. UBRISTES is ready to use in building inspection and has been proved as a useful tool in the design of rehabilitation projects for inaccessible, complex building structures in the context of energy efficiency.

In this paper we present a reliable motion planner that takes into account the kinematic restrictions, the shape of the robot and the motion uncertainty along the path. Our approach is based on a state lattice that predicts the uncertainty along the paths and obtains the one which minimizes both the probability of collision and the cost. The uncertainty model takes into account the stochasticity in motion and observations and the corrective effect of using a Linear Quadratic Gaussian controller. Moreover, we introduce an adaptive multi-resolution lattice that selects the most adequate resolution for each area of the map based on its complexity. Experimental results, for several environments and robot shapes, show the reliability of the planner and the effectiveness of the multi-resolution approach for decreasing the complexity of the search.

This paper presents a two folded solution to facilitate and improve the teleoperation of unmanned vehicles in unknown scenarios. The first part of the solution regards increasing the operators perception of the vehicle surroundings by means of a new vibration feedback transmitted by a haptic controller. The second part concerns the implementation of new anti-collision methods that take into account both vehicle and environment constraints through a spatial representation of the allowed vehicle velocities. The solution was tested and validated by 28 subjects tele-operating an omnidirectional ground vehicle through an unseen maze. The experiment results show a reduction of the human operator workload and of the time taken to complete the task. The vibration feedback was compared by the subjects with other haptic feedbacks in an experiment to identify the direction of a single obstacle, outperforming these in the effective indication of the presence and direction of the existing obstacle.

If robots are to invade our homes and offices, they will have to interact more naturally with humans. Natural interaction will certainly include the ability of robots to plan their motion, accounting for the social norms enforced. In this paper we propose a novel solution for Human-Aware Navigation resorting to external omnidirectional static cameras, used to implement a vision-based person tracking system. The proposed solution was tested in a typical domestic indoor scenario in four different experiments. The results show that the robot is able to cope with human-aware constraints, defined after common proxemics rules.

A great variety of human gesture recognition methods exist in the literature, yet there is still a lack of solutions to encompass some of the challenges imposed by real life scenarios. In this document, a gesture recognition for robotic search and rescue missions in the high seas is presented. The method aims to identify shipwrecked people by recognizing the hand waving gesture sign.We introduce a novel motion descriptor, through which high recognition accuracy can be achieved even for low resolution images. The method can be simultaneously applied to rigid object characterization, hence object and gesture recognition can be performed simultaneously.The descriptor has a simple implementation and is invariant to scale and gesture speed. Tests, preformed on a maritime dataset of thermal images, proved the descriptor ability to reach a meaningful representation for very low resolution objects. Recognition rates with 96.3% of accuracy were achieved.

Manual design of self-organized behavioral control for swarms of robots is a complex task. Neuroevolution has proved a viable alternative given its capacity to automatically synthesize controllers. In this paper, we introduce the concept of Genome Variations (GV) in the neuroevolution of behavioral control for robotic swarms. In an evolutionary setup with GV, a slight mutation is applied to the evolving neural network parameters before they are copied to the robots in a swarm. The genome variation is individual to each robot, thereby generating a slightly heterogeneous swarm. GV represents a novel approach to the evolution of robust behaviors, expected to generate more stable and robust individual controllers, and benefit swarm behaviors that can deal with small heterogeneities in the behavior of other members in the swarm. We conduct experiments using an aggregation task, and compare the evolved solutions to solutions evolved under ideal, noise-free conditions, and to solutions evolved with traditional sensor noise.

This paper addresses the problem of real-time adaptive sampling using a coordinated fleet of Autonomous Underwater Vehicles (AUVs). The system setup consists of one leader AUV and one or more follower AUVs, all equipped with conductivity, temperature and depth (CTD) sensor devices and capable of running in real-time an on-line unsupervised learning computer algorithm that uses and updates Gaussian Mixture Models (GMMs) to model the CTD data that is being acquired in real-time. The path to be traced by the leader is predefined. The followers path will depend on the CTD data. More precisely, during each resurfacing of the AUVs (and this has to be done in a coordinated fashion), every follower AUV receives the GMM hypothesis of the leader and computes the variational distance error between its own GMM and the received one. This error, that provides a notion of how different is the CTD data of each follower from the leader, is used to reconfigure the formation by scaling the distance between the AUVs in the formation (making a zoom-in and zoom-out), in order to improve the efficiency of the CTD data acquisition in a given region. The simulation results show the feasibility of the proposed strategy in uniform and more complex environments.

In hilly terrains, the exploitation of (semi-)autonomous systems able to travel nimbly and safely on different terrains and perform agricultural operations is still far from reality.In this perspective, the articulated 4-wheeled system, that shows an optimal steering capacity and the possibility to adapt to uneven terrains thanks to a passive degree of freedom on the central joint, is one of the most promising mobile wheeled-robot architectures. In this work, the instability of this robotic platform is evaluated in the two different conditions, i.e. phase I and phase II [1], and the effect of blocking the passive DoF of the central joint investigated in order to highlight possible stabilizing conditions and best manoeuvring practices for overturning avoidance. In order to do so, a quasi-static model of the robotic platform has been developed and implemented in a Matlab™ simulator thanks to which the different conditions have been studied.

Coverage path planning (CPP) is a fundamental agricultural field task required for autonomous navigation systems. It is also important for resource management, increasingly demanding in terms of reducing costs and environmental polluting agents as well as increasing productivity. Additional problems arise when this task involves irregular agricultural terrains where the crop follows non-uniform configurations and extends over steep rocky slopes. For mountain vineyards, finding the optimal path to cover a restricted set of terraces, some of them with dead ends and with other constraints due to terrain morphology, is a great challenge. The problem involves other variables to be taken into account such as speed, direction and orientation of the vehicle, fuel consumption and tank capacities for chemical products. This article presents a decision graph-based approach, to solve a Rural Postman Coverage like problem using A* and Dijkstra algorithms simultaneously to find the optimal sequence of terraces that defines a selected partial coverage area of the vineyard. The decision structure is supported by a graph that contains all the information of the Digital Terrain Model (DTM) of the vineyard. In this first approach, optimality considers distance, cost and time requirements. The optimal solution was represented in a graphical user OpenGL application developed to support the path planning process. Based on the results, it was possible to prove the applicability of this approach for any vineyards which extend like routes. Near optimal solutions based on other specific criteria could also be considered for future work.

Crop monitoring and harvesting by ground robots in steep slope vineyards is an intrinsically complex challenge, due to two main reasons: harsh conditions of the terrain and reduced time availability and unstable localization accuracy of the Global Positioning System (GPS). In this paper the use of agricultural wireless sensors as artificial landmarks for robot localization is explored. The Received Signal Strength Indication (RSSI), of Bluetooth (BT) based sensors/technology, has been characterized for distance estimation. Based on this characterization, a mapping procedure based on Histogram Mapping concept was evaluated. The results allow us to conclude that agricultural wireless sensors can be used to support the robot localization procedures in critical moments (GPS blockage) and to create redundant localization information.

To develop robust algorithms for agricultural navigation, different growth stages of the plants have to be considered. For fast validation and repeatable testing of algorithms, a dataset was recorded by a 4 wheeled robot, equipped with a frame of different sensors and was guided through maize rows. The robot position was simultaneously tracked by a total station, to get precise reference of the sensor data. The plant position and parameters were measured for comparing the sensor values. A horizontal laser scanner and corresponding total station data was recorded for 7 times over a period of 6 weeks. It was used to check the performance of a common RANSAC row algorithm. Results showed the best heading detection at a mean growth height of 0.268 m.

Autonomous vehicles rely on the accurate estimation of their pose, speed and direction of travel to perform basic navigation tasks. Although GPSs are very useful, they have some drawbacks in urban applications. Visual odometry is an alternative or complementary method, because it uses a sensor already available in many vehicles for other tasks and provides the ego motion of the vehicle with enough accuracy. In this paper, a new method is proposed that detects and tracks features available on the surface of the ground, due to the texture of the road or street and road markings. This way it is assured only static points are taking into account in order to obtain the relative movement between images. A Kalman filter is used taking into account the Ackermann steering restrictions. Some results in real urban environments are shown in order to demonstrate the good performance of the algorithm.

Convolutional Neural Networks (CNNs) surpassed the human performance on the German Traffic Sign Benchmark competition. Both the winner and the runner-up teams trained CNNs to recognize 43 traffic signs. However, both networks are not computationally efficient since they have many free parameters and they use highly computational activation functions. In this paper, we propose a new architecture that reduces the number of the parameters $$27\%$$ and $$22\%$$ compared with the two networks. Furthermore, our network uses Leaky Rectified Linear Units (Leaky ReLU) activation function. Compared with 10 multiplications in the hyperbolic tangent and rectified sigmoid activation functions utilized in the two networks, Leaky ReLU needs only one multiplication which makes it computationally much more efficient than the two other functions. Our experiments on the German Traffic Sign Benchmark dataset shows $$0.6\%$$ improvement on the best reported classification accuracy while it reduces the overall number of parameters and the number of multiplications $$85\%$$ and $$88\%$$, respectively, compared with the winner network in the competition. Finally, we inspect the behaviour of the network by visualizing the classification score as a function of partial occlusion. The visualization shows that our CNN learns the pictograph of the signs and it ignores the shape and color information.

Particle filters are sequential Monte Carlo estimation methods with applications in the field of mobile robotics for performing tasks such as tracking, simultaneous localization and mapping (SLAM) and navigation, by dealing with the uncertainties and/or noise generated by the sensors as well as with the intrinsic uncertainties of the environment. This work presents a field programmable gate arrays (FPGA) implementation of a particle filter applied to SLAM problem based on a low cost Neato XV-11 laser scanner sensor. Post processing is performed on data provided by a realistic simulation of a differential robot, equipped with a hacked Neato XV-11 laser scanner, that navigates in the Robot@Factory competition maze. The robot was simulated using SimTwo, which is a realistic simulation software that can support several types of robots. The simulator provides the robot ground truth, odometry and the laser scanner data. The results achieved from this study confirmed the possible use such low cost laser scanner for different robotics applications.

Laser scanners are widely used in mobile robotics localization systems but, despite the enormous potential of its use, their high price tag is a major drawback, mainly for hobbyist and educational robotics practitioners that usually have a reduced budget. This paper presentes the modeling and simulation of a hacked Neato XV-11 Laser Scanner, having as motivation the fact that it is a very low cost alternative, when compared with the current available laser scanners. The modeling of a hacked Neato XV-11 Laser Scanner allows its realistic simulation and provides valuable information that can promote the development of better designs of robot localization systems based on this sensor. The sensor simulation was developed using SimTwo, which is a realistic simulation software that can support several types of robots.

This paper proposes a novel method for appearance based vehicle detection by employing stereo vision system and radar units. In the vein of utilizing advanced driver assistance systems, detection and tracking of moving objects or particularly vehicles, represents an essential task. For the merits of such application, it has often been suggested to combine multiple sensors with complementary modalities. In accordance, in this work we utilize a stereo vision and two radar units, and fuse the corresponding modalities at the level of detection. Firstly, the algorithm executes the detection procedure based on stereo image solely, generating the information about vehicles’ position. Secondly, the final unique list of vehicles is obtained by overlapping the radar readings with the preliminary list obtained by stereo system. The stereo vision–based detection procedure consists of (i) edge processing plugging in also the information about disparity map, (ii) shape based vehicles’ contour extraction and (iii) preliminary vehicles’ positions generation. Since the radar readings are examined by overlapping them with the list obtained by stereo vision, the proposed algorithm can be considered as high level fusion approach. We analyze the performance of the proposed algorithm by performing the real-world experiment in highly dynamic urban environment, under significant illumination influences caused by sunny weather.

This paper presents improvements made to previous method for monocular teach-and-repeat navigation of mobile robots. The method is based on recording the position of image features in camera image, and moving the robot so their position matches during the recall. The method has shown good reliability, though requires odometry to perform well. This paper targets improvements of the method by replacement of a simple odometry by visual pose recognition approach. Thus, localization becomes independent of preceding pose computation. This prevents accumulation of error during the run of the algorithm.A pose recognition method based on angle differences is presented herein. The substitution of odometry implies necessary adjustments to the aforementioned method to be used. Suitability of the method for pose recognition is evaluated experimentally. The method has shown to be feasible for the nav task, although the achieved accuracy is lower than the original method.

Perception is the process by which an intelligent system translates sensory data into an understanding of the world around it. Perception of dynamic environments is one of the key components for intelligent vehicles to operate in real-world environments. This paper proposes a method for static/dynamic modeling of the environment surrounding a vehicle. The proposed system comprises two main modules: (i) a module which estimates the ground surface using a piecewise surface fitting algorithm, and (ii) a voxel-based static/dynamic model of the vehicle’s surrounding environment using discriminative analysis. The proposed method is evaluated using KITTI dataset. Experimental results demonstrate the applicability of the proposed method.

The number of LIDAR sensors installed in robotic vehicles has been increasing, which is a situation that reinforces the concern of sensor calibration. Most calibration systems rely on manual or semi-automatic interactive procedures, but fully automatic methods are still missing due to the variability of the nearby objects with the point of view. However, if some simple objects could be detected and identified automatically by all the sensors from several points of view, then automatic calibration would be possible on the fly. This is indeed feasible if a ball is placed in motion in front of the set of uncalibrated sensors allowing them to detect its center along the successive positions. This set of centers generates a point cloud per sensor, which, by using segmentation and fitting techniques, allows the calculation of the rigid body transformation between all pairs of sensors. This paper proposes and describes such a method with encouraging preliminary results.

This paper presents an algorithm to perform pedestrian pose estimation using a stereo vision system in the Advanced Driver Assistance Systems (ADAS) context. The proposed approach isolates the pedestrian point cloud and extracts the pedestrian pose using a visibility based pedestrian 3D model. The model accurately predicts possible self occlusions and uses them as an integrated part of the detection. The algorithm creates multiple pose hypotheses that are scored and sorted using a scheme reminiscent of the Monte Carlo techniques. The technique performs a hierarchical search of the body pose from the head position to the lower limbs. In the context of road safety, it is important that the algorithm is able to perceive the pedestrian pose as quickly as possible to potentially avoid dangerous situations, the pedestrian pose will allow to better predict the pedestrian intentions. To this end, a single pedestrian model is used to detect all pertinent poses and the algorithm is able to extract the pedestrian pose based on a single stereo depth point cloud and minimal orientation information. The algorithm was tested against data captured with an industry standard motion capture system. Accurate results were obtained, the algorithm is able to correctly estimate the pedestrian pose with acceptable accuracy. The use of stereo setup allows the algorithm to be used in many varied contexts ranging from the proposed ADAS context to surveillance or even human-computer interaction.

In this paper, we present a novel methodology to compute a 3D scene representation. The algorithm uses macro scale polygonal primitives to model the scene. This means that the representation of the scene is given as a list of large scale polygons that describe the geometric structure of the environment. Results show that the approach is capable of producing accurate descriptions of the scene. In addition, the algorithm is very efficient when compared to other techniques.

The manuscript evaluates the performance of a monocular visual odometry approach when images from different spectra are considered, both independently and fused. The objective behind this evaluation is to analyze if classical approaches can be improved when the given images, which are from different spectra, are fused and represented in new domains. The images in these new domains should have some of the following properties: i) more robust to noisy data; ii) less sensitive to changes (e.g., lighting); iii) more rich in descriptive information, among other. In particular in the current work two different image fusion strategies are considered. Firstly, images from the visible and thermal spectrum are fused using a Discrete Wavelet Transform (DWT) approach. Secondly, a monochrome threshold strategy is considered. The obtained representations are evaluated under a visual odometry framework, highlighting their advantages and disadvantages, using different urban and semi-urban scenarios. Comparisons with both monocular-visible spectrum and monocular-infrared spectrum, are also provided showing the validity of the proposed approach.

This paper represents research in progress in Simultaneous Localization and Mapping (SLAM) for Micro Aerial Vehicles (MAVs) in the context of rescue and/or recognition navigation tasks in indoor environments. In this kind of applications, the MAV must rely on its own onboard sensors to autonomously navigate in unknown, hostile and GPS denied environments, such as ruined or semi-demolished buildings. This article aims to investigate a new SLAM technique that fuses laser and visual information, besides measurements from the inertial unit, to robustly obtain the 6DOF pose estimation of a MAV within a local map of the environment. Laser is used to obtain a local 2D map and a footprint estimation of the MAV position, while a monocular visual SLAM algorithm enlarges the pose estimation through an Extended Kalman Filter (EKF). The system consists of a commercial drone and a remote control unit to computationally afford the SLAM algorithms using a distributed node system based on ROS (Robot Operating System). Some experimental results show how sensor fusion improves the position estimation and the obtained map under different test conditions.

In this paper, a single DOF anthropomorphic finger module specifically designed for aerial manipulation and grasping is presented, where low weight and compliance are required features for safer performance of the aerial platform. The under-actuated mechanism consists of a high torque micro motor with a small reel for driving a tendon that moves the three joints of the finger. An elastic element maintains the finger bones tied together and extended by default, providing a compliant response against collisions with walls, the floor or any object in the environment during the grasping operation. Compliance provided by the elastic joints will also be exploited for stable object grasping and for collision detection. The development of a specific electronics for finger control instead of employing conventional servos make possible the design and application of a wider range of control strategies, including position, velocity or open-loop force control. The proposed design is validated through different grasping and control experiments.

The work presented in this paper is a part of research work on autonomous navigation for Micro Aerial Vehicles (MAVs). Simultaneous Localization and Mapping (SLAM) is crucial for any task of MAV navigation. The limited payload of the MAV makes the single camera as best solution for SLAM problem. In this paper the Large Scale Dense SLAM (LSD-SLAM) pose is fused with inertial data using Smooth Variable Structure Filter which is a robust filter. Our MAV-SVSF-SLAM application is developed under Linux using Robotic Operating System (ROS) so that the code can be distributed in different nodes, to be used in other applications of guidance, control and navigation. The proposed approach is validated first in simulation, then experimentally using the Bebop Quadrotor in indoor and outdoor environment and good results have been obtained.

Increasing flight endurance of Unmanned Aerial Vehicles (UAVs) is a main issue for many applications of these aircrafts. This paper deals with air to air refueling between UAVs. Relative estimation using only INS/GPS system is not sufficiently accurate to accomplish an autonomous dock for aerial refueling using a boom system in the tanker.In this paper we propose a quaternion based relative state estimator to fuse GPS and INS sensor data of each UAV with vision pose estimation of the receiver obtained from the tanker. Simulated results validate the approach and are the starting point for ground and flight tests in the next months.

The work presented in this paper is part of the autonomous planning architecture of a team of aerial robots equipped with on-board robotic arms. The mission of the team is the construction of structures in places where the access is difficult by conventional means, which is the scenario considered in the framework of the ARCAS European Project. This paper presents a planning engine for this context. From the 3D CAD model of the structure an assembly planner computes the required assembly tasks, which are the inputs for the system. These tasks are assigned to the available aerial robots by a task allocation planner, which computes an assignment and improves it step by step by communicating with a symbolic planner. The symbolic planner estimates the cost of the sequence of actions needed in the mission execution for the given assignment. This paper presents simulation results that show the feasibility of the approach and a comparison between different solvers.

Multi-UAV missions are complex systems that may include a fleet of UAVs, a crew of operators and different computers and interfaces. Currently, an important challenge is the reduction of the number of operators that is required for performing a multi-UAV mission. This challenge can be addressed by increasing the autonomy of fleets and providing capabilities of operators to the interfaces. This paper presents a proposal of control architecture for multi-UAV missions. This architecture shares some elements with centralized and distributed approaches and it has three layers: mission, task and action. The mission layer is implemented in the GCS and performs the mission planning and operator interfacing. Meanwhile, the task and action layers are located in the UAVs and perform respectively the task planning and executing. This architecture is applied to a simulation environment that reproduce a competitive scenario with two fleets of UAVs.

This paper addresses the problem of trajectory tracking control of autonomous rotorcraft relative to pole-shaped structures using LiDAR sensors. The proposed approach defines an alternative kinematic model, directly based on LiDAR measurements, and uses a trajectory-dependent error space to express the dynamic model of the vehicle. An LPV representation with piecewise affine dependence on the parameters is adopted to describe the error dynamics over a set of predefined operating regions. The synthesis problem is stated as a continuous-time $$\mathcal {H}_2$$ control problem, solved using LMIs and implemented within the scope of gain-scheduling control theory. The performance of the proposed control method is validated with comprehensive simulation results.

In this paper we address the path-following problem for fixed-wing Unmanned Aerial Vehicles (UAVs) in presence of wind disturbances. Given a desired path with a specified airspeed profile assigned on it, the goal is to follow the desired maneuver while minimizing the control effort. We propose a predictive path following control scheme based on trajectory optimization techniques, to compute feasible UAV trajectories, combined with a sample-data Model Predictive Control (MPC) approach, to handle the wind field. By explicitly addressing the wind field, the UAV exploits the surrounding environment thus extending its capabilities in executing the desired maneuver. We provide numerical computations based on straight line and circular paths under various wind conditions. The computations allow us to highlight the benefits of the proposed control scheme.

This paper presents a efficent conflict detection and resolution (CDR) method for an aerial vehicle sharing airspace with other aerial vehicles. It is based on a conflict detection (CD) algorithm (axis-aligned minimum bounding box) and conflict resoluction (CR) algorithm (genetic algorithms) to find safe trajectories. Monte-Carlo estimation is used to evaluate the best predicted trajectories considering different sources of uncertainty such as the wind, the inaccuracies in the vehicle model and limitations of on-board sensors and control system. Simulations are performed in different scenarios and conditions of wind to test the method.

In this work we present the integration between a robot cooperative control strategy and a wireless network simulated with OMNeT++. We use a consensus control strategy to carry out a rendez-vous task where information is shared among a group of robots. These robots are then simulated in a MANET environment with a TDMA-based protocol to minimize message collisions. We consider two cases in this work: a fixed pre-determined topology, which does not accept new links, and a dynamic topology that creates new links as robots get within communication range. We show the impact of the network on the control strategy performing a rendez-vous task, considering both topologies. In particular, there is a considerable degradation of the rendez-vous task if care is not taken when deploying the cooperative control strategy, e.g. the initial message collisions due to desynchronized slot start. Finally, we compare these simulation results with those from a Matlab implementation of the control strategy using a typical simplified network model. The difference reveals the importance of using more accurate network models such as those of OMNeT++.

This paper presents a high-level communication infrastructure to deal with dynamically changing reconfigurable multi-robot systems. The infrastructure builds upon official standards of the Foundation for Intelligent Physical Agents (FIPA). FIPA standards have been successfully applied in a variety of multi-agent frameworks, but they have found little application in the domain of robotics. This paper introduces an implementation that can complement existing robotic communication frameworks and allows the robotics community to take better advantage of multi-agent research efforts. We present the essential components of the infrastructure and show its interoperability using the widely known multi-agent framework JADE.

In this paper, we address the problem of optimal multi-robot team deployment while maintaining communication for all the robots. The objective is to execute the mission of reaching several goals with minimal number of robots, as well as reducing the total distance travelled to reach the goals. Therefore, we develop an algorithm that computes some secondary or virtual goals to move robots enhancing the coverage over the map. Due to the presence of obstacles, we study the use of different criteria in order to add more flexibility to the optimization in terms of travelled distance or relay nodes saving.

In tunnel-like environments such as road tunnels or mines, a team of networked mobile robots can provide surveillance, search and rescue, or monitoring services. However, these scenarios pose multiple challenges from the robotics and from the communication points of view. Structurally, tunnels are much more longer than they are wide, and in the communication context, the multipath propagation causes strong fading phenomena. While the former can be addressed implementing routing techniques that allow multi-hop communication, fadings are unavoidable and represent a serious issue. However, under certain transmitter-receiver configurations, these fadings are predictable and periodic. On this basis, in this work we present a set of solutions to improve the communications between a base station and a robot taking advantage of spatial diversity and link-quality-aware navigation. These proposals are tested by means of simulations and real-world experiments carried out in a long railway tunnel, involving mobile and fixed nodes towards an application for monitoring purposes.

In this paper, it is proposed a visual surveillance system for multiple UAVs under communication constrains. In previous works, a dynamic task allocation algorithm was designed for assigning patrolling and tracking tasks between multiple robots. The idea was to assign the intruders dynamically among the robots using one-to-one coordination technique. However due to communication constrains, every UAVs could store different information. In this paper, local information about targets is obtained by a visual algorithm that detects moving objects during surveillance tasks using fixed low-cost monocular RGB-camera connected to an on-board computer. This system was tested in a urban surveillance scenario, implemented in an indoor test-bed, under EC-SAFEMOBIL project.

The increase of the use of Autonomous Vehicles in different types of environments leads to an improvement of the Localization and Navigation algorithms. The goal is to increase the levels of efficiency, security and robustness of the system, minimizing the tasks completion time.The application of cleaning robots in domestic environments have several advantages however some improvements should be performed in order to develop a robust system. Also in large spaces one robot doesn’t achieve the desired performance in terms of robustness to faults and efficiency in the cleaning process. Considering a fleet of autonomous robots, this process could be improved. The purpose of our paper is the presentation of an architecture for management a fleet of cleaning robots, considering a complete coverage path planning for large and structured environments. Compartments are found in a grid-like decomposition and an area coverage strategy are evolved (optimized) by using Genetic Algorithms. The Task allocation module is based on Auctions strategy, thus obtaining cooperation under dynamic constraints in complex environments. The case study optimizes the number of robots involved in the cooperative cleaning of a full building in the campus, based on its real architectural plans.

The involvement of children and adolescents in robotics is on demand by the many robotics events and competitions all over the world. This non-deterministic world is more attractive, fun, hands-on and with real results than computer virtual simulations and 3D worlds. It is important, by different reasons, to involve people of all ages in an area that some consider the future of mankind and an opportunity to increase the low rate of engineers globally. Robotics competitions at this level are essentially based on teaching motion and programming skills by using Lego™ based robots and a set of challenges to overcome. This paper presents a different approach that is being used by Minho University in order to attract STEM candidates into these fields, with visible success and excellent results. The event is called RoboParty® and teaches children, adolescents and adults, from any area, how to build a robot from scratch, using electronics, mechanics and programming during three non-stop days.

A multi-robot stage production is novel and challenging as different robots have to communicate and coordinate to produce a smooth performance. We made a multi-robot stage production possible using the NAO humanoid robots and the Lego Mindstorms NXT robots with a group of undergraduate women who had programming experience, but little experience with robots. The undergraduates from around the world were participating in a three day workshop – Opportunities for Undergraduate Research in Computer Science (OurCS), organized by the School of Computer Science from Carnegie Mellon University that provide opportunities for these undergraduates to work on computing-related research problems. They were given twelve and a half hours over a span of three days to familiarize themselves with the robots, plan the storyboard of the performance, program the robots, generate a multi-robot performance and create a presentation on what they learned and did. In this paper, we describe the tools and infrastructure we created to support the creation of a multi-robot stage production within the allocated time and explain how the time in the workshop was allocated to enable the undergraduates to complete the multi-robot stage production.

In this article, a course that explores the potential of learning ROS using a collaborative game world is presented. The competitive mindset and its origins are explored, and an analysis of a collaborative game is presented in detail, showing how some key design features lead participants to overcome the challenges proposed through cooperation and collaboration. The data analysis is supported through observation of two different game simulations: the first, where all competitors were playing solo, and the second, where the players were divided in groups of three. Lastly, the authors reflect on the potentials that this course provides as a tool for learning ROS.

Taking distributed robotic system research from simulation to the real world often requires the use of small robots that can be deployed and managed in large numbers. This has led to the development of a multitude of these devices, deployed in the thousands by researchers worldwide. This paper looks at the Khepera IV mobile robot, the latest iteration of the Khepera series. This full-featured differential wheeled robot provides a broad set of sensors in a small, extensible body, making it easy to test new algorithms in compact indoor arenas. We describe the robot and conduct an independent performance evaluation, providing results for all sensors. We also introduce the Khepera IV Toolbox, an open source framework meant to ease application development. In doing so, we hope to help potential users assess the suitability of the Khepera IV for their envisioned applications and reduce the overhead in getting started using the robot.