Search

When oil spills occur at sea, effective detection and monitoring are required to establish a successful response plan, and remote sensing emerges as an appropriate technology to support this process. Furthermore, Earth observation is progressively relying on low-cost multispectral sensors developed for monitoring particular features across various scenarios. However, despite these advancements, the ongoing challenge lies in reducing computational costs, resource requirements, and energy consumption. This work aims to select the spectral index that best detects coastal spills among those documented. The long-term goal is to develop a low-cost multispectral sensor with suitable bands. For this purpose, this study uses data from different sensors that acquired data of the Deepwater Horizon accident in the Gulf of Mexico. The confusion matrices, accuracy, and F1-score generated from the kNN pixel classification based on the indices values are measures of its performance. For this study, the recently introduced normalized difference oil index (NDOI) proves to be the best option for identifying coastal spills since it minimizes the false positives related to suspended sand and is quick to calculate. In addition, it has demonstrated that it helps distinguish different spill thicknesses and estimates the oil volume. Therefore, future research will focus on developing and validating a low-cost multispectral system that uses the NDOI bands to detect spills.

Electromagnetic waves propagation through the ionosphere is subject to several effects including refraction, absorption, signal delay, or Faraday rotation. A ray tracer propagator that simulates these effects is an important tool for all satellite missions relying on transionospheric communications and Earth observation. This study presents an update to the 1975 Jones and Stephenson IONORT's code by implementing the use of up-to-date ionospheric, atmospheric, and geomagnetic models (International Reference Ionosphere or NeQuick, NRLMSISE-00, and International Geomagnetic Reference Field), and a new 3-D model for equatorial plasma bubbles. The developed code is part of the Ionosphere Modular Software Package in the context of the European Space Agency project SIMIONO. High-frequency radar sounders, synthetic aperture radars, and GNSS reflectometry missions constitute the main application of this software. The ray tracer has been validated with respect to the original software and by comparing real versus simulated vertical ionograms in different locations and dates.

Text processing is a fundamental aspect of Natural Language Processing (NLP) and is crucial for various applications in fields such as artificial intelligence, data science, and information retrieval. It plays a core role in language models. Most text-processing approaches focus on describing and synthesizing, to a greater or lesser degree, lexical, syntactic, and semantic properties of text in the form of numerical vectors that induce a metric space, in which, it is possible to find underlying patterns and structures related to the original text. Since each approach has strengths and weaknesses, finding a single approach that perfectly extracts representative text properties for every task and application domain is hard. This paper proposes a novel approach capable of synthesizing information from heterogeneous state-of-the-art text processing approaches into a unified representation. Encouraging results demonstrate that using this representation in popular machine-learning tasks not only leads to superior performance but also offers notable advantages in memory efficiency and preservation of underlying information of the distinct sources involved in such a representation.

This paper presents the methodology followed for the design of a quad-ridged flared horn (QRFH) feed for the VLBI Global Observing System (VGOS) antennas. The study demonstrates the potential undesired effects caused by the presence of non-fundamental modes at the throat of the horn antenna, and how they can be avoided following some design considerations. The approach consists of a separate design of the feeding transition and the flared section of the horn, and it can be extrapolated to other implementations. Firstly, the geometry of the coaxial to quad-ridged waveguide (QRWG) transition is defined. The proposed design technique ensures negligible propagation of higher order modes at the waveguide interface. Secondly, the ridges and sidewall profiles of the flared part are optimized to achieve the desired performance in terms of input reflection and aperture efficiency on the VGOS 13.2-meter displaced-axis two-reflector radio telescope. Following the optimized design, five QRFH antenna units have been manufactured and tested, all showing reflection coefficients better than -10 dB and predicted aperture efficiencies better than 55 % in the band from 2.2 to 14 GHz. The excellent agreement between simulations and experimental results demonstrates the validity of the proposed methodology.

Invasive Coronary Angiography (ICA) images are considered the gold standard for assessing the state of the coronary arteries. Deep learning classification methods are widely used and well-developed in different areas where medical imaging evaluation has an essential impact due to the development of computer-aided diagnosis systems that can support physicians in their clinical procedures. In this paper, a new performance analysis of deep learning methods for binary ICA classification with different lesion degrees is reported. To reach this goal, an annotated dataset of ICA images containing the ground truth, the location of lesions, and seven possible severity degrees ranging between 0% and 100% was employed. The ICA images were divided into “lesion” or “non-lesion” patches. We aim to study how binary classification performance is affected by the different lesion degrees considered in the positive class. Therefore, five Convolutional Neural Network architectures – DenseNet-201, MobileNet-V2, NasNet-Mobile, ResNet-18, and ResNet-50 – were trained with different input images where different lesion degree ranges were gradually incorporated until considering the seven lesion degrees. Besides, four types of experiments with and without data augmentation were designed, whose F-measure and Area Under Curve (AUC) were computed. Reported results achieved an F-measure and AUC of 92.7% and 98.1%, respectively. However, lesion classification is highly affected by the degree of the lesion intended to be classified, with 15% less accuracy when < 99% lesion patches are present.

Hospitality institutions are significant consumers of water, electricity and gas resources. These resources may be scarce or critical in several locations. Authors have investigated strategies for efficient building operation, considering various energy carriers. Despite reported methods to enhance efficiency, nowadays they commonly have to be integrated into hotels from the design process through regular operation. This study aims to provide a detailed analysis of sustainable architecture in hotel design to decrease resource usage without compromising the quality of services provided. The considered design process is based on using a structured methodology focused on a energy management system that guarantees an efficient use of resources. The present study also considers incorporating renewable energy sources for the nominal operation of the hotel to achieve a net-zero energy building classification. The design of the building resulted in the proposed architecture that can maintain a comfortable temperature range of $17^{\circ } C$ to $24^{\circ } C$ in the guest rooms throughout the year. The energy generated from the proposed photovoltaic, thermal solar systems and rainwater harvesting can guarantee a nearly negligible net consumption of these resources for the hotel.

The technological advancements in port operations over the last decade have enabled the incorporation of significant elements into the supply chain, contributing to the progress of Industry 4.0. Through Big Data, the Internet of Things, and Artificial Intelligence, strides have been made in technical, operational, and procedural management for maritime ports. The literature on this subject is diverse and spans various areas of focus. This study aims to outline the principal trajectories of digital maritime transport by applying scientometrics and the ToS metaphor. Employing an exploratory methodology, a scientometric mapping was conducted based on Scopus and WoS, analyzing the most relevant scientific production by authors, countries, journals, and affiliations. Three trends are presented: Smart Ports: A Technological Necessity for the 21st Century; Greening the Digital Harbor; and Sustainable Port Management. The study concludes that the dynamics of scientific production have seen significant growth in the influence of Industry 4.0 in the maritime field, and it also evidences the applicability of technological innovations in the port sector.

Wine aging is a fundamental stage in the creation of special high-quality wines, such as sherries and similar wines. To undergo the process of “biological aging”, some of these wines are stored in oak casks for long periods. During this process, a thin layer of yeast is spontaneously formed on the surface of the wine. This layer protects the wine from oxidation and contributes to its unique organoleptic characteristics. This work presents the prototype of an instrumented bung capable of monitoring, in real-time, the absorbance evolution of the Fino Sherry wine during its biological aging process. The results of the tests carried out in the laboratory showed that the device can directly measure the absorbance at wavelengths of 420 nm and 520 nm, detecting small changes produced by the wine oxidation processes. This work has significant implications for winery technicians and oenologists as they could analyze the evolution of the wine in real time, detect problems in advance, and avoid work-related accidents. These conclusions solidify the enduring relevance of this research and offer a robust foundation for future developments in the field.

From 1980 to 2021, earthquakes have caused more than 846 000 casualties and about US$ 661 billion in economic losses. At present, there are no reliable earthquake precursors to generate alerts. Currently, the link between earthquakes and total electron content variations measured by global navigation satellite systems (GNSSs) monitoring ground stations has been studied. However, GNSS ground-based monitoring stations are irregularly disseminated around the globe with significant gaps, particularly in the ocean's regions. In this article, we analyze ionospheric intensity scintillation anomalies as potential proxies of earthquakes. NASA CYGNSS GNSS-R (reflectometry) acquired by CYGNSS/delay Doppler mapping instrument from 2017 to 2021 has been used to calculate and analyze the anomalies in the S4 parameter over ocean areas affected by earthquakes. More than 30 000 ocean earthquakes within ±40° in latitude and with a magnitude larger than M4 have been examined. The daily planetary geomagnetic index Kp was used to discard data that may be disturbed due to space weather conditions. In addition, a daily sea wind speed mask was used to eliminate sea states that impact the signal reflectivity. The standard deviation and the interquartile time series methods have been used to detect these S4 anomalies. The confusion matrix, the receiver operating curve, and some other figures of merit have been used for the first time to evaluate and improve the performance of the prediction parameters and identify the optimum configuration to be used as a potential proxy of earthquake occurrence. As a result, a small, but detectable positive S4 anomaly was detected between 1 and 7 days before the earthquakes under study.

We analyze a point-to-point MIMO communication system over a deterministic channel with two paths: line-of-sight (LoS) and a scattered path over a rough surface. We characterize the MIMO channel capacity as a function of the root-mean-square (RMS) surface roughness and provide three important findings: 1) capacity is monotonically decreasing with the surface roughness, 2) capacity is lower bounded by the SISO case and upper bounded by the smooth surface case, and 3) capacity has a breaking point corresponding to the maximum surface roughness that supports two spatial streams. When the roughness is below the maximum, the proposed LoS-MIMO system assigns power to two spatial streams. If surface roughness is above the maximum, then capacity is attained by assigning power to a single spatial stream. We derive a closed-form expression for the breaking point in terms of the transmitted power, antenna gains, wavelength, and path geometry parameters. We analyze the practical implications of these findings, including the potential multiplexing gains (from 68% to 86% with respect to the SISO channel) in mmWave systems using compact linear arrays. The channel capacity is evaluated numerically in three scenarios typical of point-to-point millimeter wave links to confirm our theoretical findings. We also study the sensitivity of the capacity and maximum surface roughness to the incidence angle, number of antennas, polarization, frequency, and surface material. Finally, we discuss implementation issues with fully-digital and hybrid analog/digital hardware. Our simulations show that hybrid hardware achieves 92% of the capacity of a fully-digital solution in the simulated scenarios.

This research is based on achieving modular solutions for photovoltaic solar systems (PVSS) in buildings. Several studies tend to inadequately address the analysis of photovoltaic (PV) module orientation, often relying on the conventional south-facing orientation. The objective is to analyze and compare the influence of PV module orientation. Therefore, two proposed PVSS were analyzed, incorporating module orientation and inclination variations. A hotel electrical consumption study was conducted in 2018-2019. The incidence of shadows over the roofed surfaces of the building was analyzed from a three-dimensional model using Sketchup software. The annual and monthly generation, as well as the performance of the installation, were estimated using PVsyst software for each projected arrangement. An economic and environmental study was carried out for these proposals to remark on their feasibility and influence on environmental protection. As a result, the east-west oriented PV system is the most efficient and feasible with a peak power of 113.24 kWp and a Performance Ratio (PR) of 0.80. The east-west oriented proposal allows avoiding emissions of 301 421 TCO2 into the atmosphere. These promising results were due to both PV modules physical orientation and their lower daily working temperature. Also, an increase in electrical production per module and an optimization of the surface are achieved. These results highlight the significance of this research as they encourage the adoption of these systems, leading to improved investment recovery periods and enhanced environmental conservation efforts.

Causal Reinforcement Learning (CRL) is an emerging field where two essential areas for the development of artificial intelligence are integrated. Existing works in the area have shown how causality can contribute to mitigate some of the limitations of reinforcement learning (RL), ranging from data-inefficiency, lack of interpretability, and long learning times, among others. However, how to use reinforcement learning to support causal discovery (CD) has so far been less explored. In this article, we introduce CARL, a Causality-Aware Reinforcement Learning framework for simultaneously learning and using causal models to speed-up the policy learning in online Markov decision process (MDP) settings. In a synergistic way, our method alternates between: (i) (RL for CD), where it promotes the selection of actions to obtain better causal models in fewer episodes than traditional methods of obtaining data in RL, (ii) (CD), where a score-based algorithm is used to learn causal models and (iii) (RL using CD), where the learned models are used to select actions that speed up the learning of the optimal policy by reducing the number of interactions with the environment. Experiments in simulated environments show that our method achieves better results in policy learning than traditional model-free and model-based algorithms while it is also able to learn the underlying causal models. We also demonstrate how the learned causal models can be directly transferred to a similar task of greater complexity, significantly reducing the number of episodes required to learn an optimal policy Finally, the method’s scalability to high-dimensional states, where the action-value function needs to be represented with deep neural networks, was verified.

Leading a healthy lifestyle significantly reduces the risk of developing Non-Communicable Diseases (NCD). However, defining and monitoring healthy eating behaviour depends on multiple factors and usually requires the intervention of experts. On the other hand, nowadays millions of images are shared on social media and web platforms. In particular, many of them are food images taken from a smartphone over time, providing information related to the individual’s diet. Consequently, exploiting recent advances in image processing and Artificial Intelligence (AI), this scenario represents an excellent opportunity to: i) create new methods that analyse the individuals’ health from what they eat, and ii) develop personalised recommendations to improve nutrition and diet under specific circumstances (e.g., obesity or COVID). This article introduces the AI4Food-NutritionFW, a framework that facilitates the creation of food image datasets tailored to configurable eating behaviours. The framework considers various aspects such as region and lifestyle, simulating a user-friendly scenario where individuals capture food images using their smartphones. The study also presents a novel food image dataset comprising 4,800 diverse weekly diets from 15 distinct profiles, ranging from healthy eating habits to unhealthy ones. Finally, we evaluate the healthy eating behaviours through a score based on the Normalised Mahalanobis Distance (NMD), achieving promising results (99.53% and 99.60% accuracy and sensitivity, respectively). We also release to the research community a software implementation of our proposed AI4Food-NutritionFW (https://github.com/BiDAlab/AI4Food-NutritionFW) and the mentioned food image dataset created with it.

Perfusion cardiovascular magnetic resonance imaging is used to quantify the heart’s blood flow, which requires the segmentation of the myocardium, a laborious task. Deep learning-based methods, the most accurate to accomplish this task, still rely on expensive motion correction steps and require large labeled datasets. This paper presents an innovative, efficient approach to myocardial perfusion segmentation, utilizing deep learning techniques without motion correction and with minimal data requirements. Through transfer learning, this methodology leverages the wealth of information available from large, publicly accessible cine magnetic resonance datasets, which provide anatomically analogous images to perfusion ones. This methodology includes normalization and cropping of cine images using a Region-of-Interest detector based on a Markovian, graph-based visual saliency algorithm improved by a sequence of morphological operations. After pretraining a U-net convolutional neural network, a special fine-tuning scheme optimizes its performance. The parameters learned are the starting point for training on a smaller perfusion dataset from the Clinical Hospital of the University of Chile. After an ablation study, the best model is obtained when using both cropping and fine-tuning from the cine dataset, segmenting the left ventricle endocardium with Dice, IoU, and Hausdorff distance of 92.2%, 85.9%, and 5.1 mm respectively, and 95.6%, 91.7%, and 4.6 mm for the left ventricle epicardium. Notably, fine-tuning achieves a Dice of 91.8% for endocardium and 95.2% for epicardium when only 289 perfusion training images are available. These are promising results for developing targeted implementations in real healthcare settings when only small datasets are available.

Wind turbine blades have been constantly increasing since wind energy becomes a popular renewable energy source to generate electricity. Therefore, the wind sector requires a more efficient and representative characterization of vertical wind speed profiles to assess the potential for a wind power plant site. This paper proposes an alternative characterization of vertical wind speed profiles based on Ward's agglomerative clustering algorithm, including both wind speed module and direction data. This approach gives a more accurate incoming wind speed variation around the rotor swept area, and subsequently, provides a more realistic and complete wind speed vector characterization for vertical profiles. Real wind database collected for 2018 in the Forschungsplattformen in Nordund Ostsee (FINO) research platform is used to assess the methodology. A preliminary pre-processing stage is proposed to select the appropriated number of heights and remove missing or incomplete data. Finally, two locations and four heights are selected, and 561588 wind data are characterized. Results and discussion are also included in this paper. The methodology can be applied to other wind database and locations to characterize vertical wind speed profiles and identify the most likely wind data vector patterns.

We present RGB-D-Fusion, a multi-modal conditional denoising diffusion probabilistic model to generate high resolution depth maps from low-resolution monocular RGB images of humanoid subjects. Accurately representing the human body in 3D is a very active research field given its wide variety of applications. Most 3D reconstruction algorithms rely on depth maps, either coming from low-resolution consumer-level depth sensors, or from monocular depth estimation from standard images. While many modern frameworks use VAEs or GANs for monocular depth estimation, we leverage recent advances in the field of diffusion denoising probabilistic models. We implement a multi-stage conditional diffusion model that first generates a low-resolution depth map conditioned on an image and then upsamples the depth map conditioned on a low-resolution RGB-D image. We further introduce a novel augmentation technique, depth noise augmentation, to increase the robustness of our super-resolution model. Lastly, we show how our method performs on a wide variety of humans with different body types, clothing and poses.

Microgrids (MGs) have emerged throughout the world as the major means of integrating a variety of distributed energy resource (DER) technologies into the distribution system. These DER technologies are mostly coupled through electronic power converters (EPCs). This fact has created new operation challenges as these converter-interfaced DERs could compromise the MG stability, reliability, and resilience margins. On the other hand, induction motor-type loads (IMs) are dynamic loads with the greatest incidence in distribution grids. Air conditioning (A/C) type loads have IMs among their components. IMs demand considerable amounts of reactive power for their correct operation. In particular, when there are disturbances in MGs with high penetration of these types of loads, the phenomenon of fault-induced delay voltage recovery (FIDVR) occurs. This phenomenon causes problems with the dynamic voltage stability (DVS) due to the deficit and inability to supply reactive power of the DERs generator park composing the MG, which could lead to a voltage collapse. To face these new issues, this work proposes an approach to incorporate flexible ac transmission system (FACTS) controllers, such as an SVC and a DSTATCOM, into the grid as reactive power supply resources for improving the microgrid dynamic voltage stability. The proposal includes the design, optimal location, and dynamic modeling of FACTS devices aiming at enhancing MG operation resilience. For validation of the developed methodology, an MG test system reported in the literature and a real microgrid system of the Galapagos Islands in Ecuador have been used. The results demonstrated the superiority of DSTATCOM over SVC in enhancing DVS, FIDVR mitigation, and consequently operational resilience.

This paper presents the development of a virtual rehabilitation game and mental load with different difficulty blocks. The game was controlled with body tracking and physiological signals - acrlong ECG and acrlong EDA - were recorded throughout the session. Several parameters - acrlong HR (HR), acrlong HRV (HRV), acrlong SCL (SCL), acrlong SCR (SCR), energy expenditure...- were extracted from these signals to check mental load influence on them. Mental load was found to affect the variation in kinetic power and instantaneous heart rate; a Support Vector Machine with linear kernel was trained with these two variables and an 82.3% accuracy rate was obtained. Furthermore, the mental load was reflected in the number of errors made by the volunteers, in the selection time and in the number of rounds in the game.

The existing clearance in the mechanical joints plays a crucial role in the assembly of mechanical systems, allowing the mobility of its components. However, few studies have explored models that consider joint clearance in the case of electromechanical energy harvesting systems. This paper examines the effect of existing clearance in an electromagnetic energy harvester attached to the human lower limb. The dynamic response of the system and an estimate of its generated power are obtained by developing a lumped model, in which clearance is included by adding a dead band and assigning a stiffness coefficient during contact between elements. The natural motion of the lower limb is the input to the formulated model, which takes into account the nonlinear interaction of the electromagnetic device and the power conditioning circuit. Central composite design is used to study the influence of two selected factors on the dynamics of the system; joint clearance size and contact stiffness. The results suggest that the presence of clearance between the clamping mechanism and the human body positively affects the performance of the analyzed electromagnetic energy harvesting system. It was revealed that an increase of around 27% of output power could be intentionally achieved by adding larger clearance sizes.

Sensor fusion is a critical aspect in autonomous drone navigation as several tasks, such as object detection and self-pose estimation, require combining information from heterogeneous sources. The performance of these solutions depends on several factors, such as the characteristics of the sensors and the environment, as well as the computing platforms, which can heavily impact their accuracy and response time. Carrying out such performance evaluations through real flight tests can be a resource-demanding, time-consuming, and, at times, risky process, which is why researchers often rely on simulation environments for testing and validating sensor fusion algorithms. The simulation environment should provide photorealistic environmental features, as well as a comprehensive set of sensors, in order to allow to test the most extensive set of sensor fusion algorithms. This paper presents AirLoop, an AirSim-based flight simulator for Hardware-in-the-Loop and Software-in-the-Loop algorithm testing and validation. AirLoop extends the sensor setup provided by AirSim with an FMCW RADAR sensor simulation, which has been evaluated based on the Infineon Technologies BGT60TR13C RADAR. Furthermore, this work provides several Software-in-the-Loop (SITL) and Hardware-in-the-Loop (HITL) demonstrations, including interfacing with the Pixhawk 2 flight controller and an extensive evaluation of the communication of the engine with the NVIDIA Jetson Nano, which has been evaluated in various use cases, including dataset creation, object detection, Path Planning, and Simultaneous Localization and Mapping (SLAM).

Given the pervasiveness of web tracking practices on the Internet, many countries are developing and enforcing new privacy regulations to ensure the rights of their citizens. However, discovering websites that do not comply with those regulations is becoming very challenging, given the dynamic nature of the web or the use of obfuscation techniques. This work presents ePrivo, a new online service that can help Internet users, website owners, and regulators inspect how privacy-friendly a given website is. The system explores all the content of the website, including traffic from third parties and dynamically modified content. The ePrivo service combines different state-of-the-art tracking detection and classification methods, including TrackSign, to discover both previously known and zero-day tracking methods. After 6 months of service, ePrivo detected the largest browsing history trackers and more than 40k domains including cookies with a lifespan longer than one year, which is forbidden in some countries.

Machine learning (ML) methods often require large volumes of labeled data to achieve meaningful performance. The expertise necessary for labeling data in medical applications like pathology presents a significant challenge in developing clinical-grade tools. Crowdsourcing approaches address this challenge by collecting labels from multiple annotators with varying degrees of expertise. In recent years, multiple methods have been adapted to learn from noisy crowdsourced labels. Among them, Gaussian Processes (GPs) have achieved excellent performance due to their ability to model uncertainty. Deep Gaussian Processes (DGPs) address the limitations of GPs using multiple layers to enable the learning of more complex representations. In this work, we develop Deep Gaussian Processes for Crowdsourcing (DGPCR) to model the crowdsourcing problem with DGPs for the first time. DGPCR models the (unknown) underlying true labels, and the behavior of each annotator is modeled with a confusion matrix among classes. We use end-to-end variational inference to estimate both DGPCR parameters and annotator biases. Using annotations from 25 pathologists and medical trainees, we show that DGPCR is competitive or superior to Scalable Gaussian Processes for Crowdsourcing (SVGPCR) and other state-of-the-art deep-learning crowdsourcing methods for breast cancer classification. Also, we observe that DGPCR with noisy labels obtains better results ( $\text{F}1=81.91$ %) than GPs ( $\text{F}1=81.57$ %) and deep learning methods ( $\text{F}1=80.88$ %) with true labels curated by experts. Finally, we show an improved estimation of annotators’ behavior.

This article addresses the problem of energy and code allocation to many users accessing, under spreading-based nonorthogonal multiple access, a wireless node set up with a successive interference cancellation architecture aided by redundancy-check error control. As an application, we consider the asynchronous access of a delay-tolerant satellite system, where users employ finite-length channel codes and are subject to a known power unbalance induced by the known distribution of the channel’s attenuation. The article develops, as a mathematically tractable approximation to massively populated systems, a unified framework to compute the best energy and code allocation rules that maximize the spectral efficiency of a network that handles asymptotically many users. Concretely, the presented approach circumvents the exponential complexity in the number of users when modeling the propagation of packet decoding failures through the receiver’s decoding scheme. It also enables a deterministic analysis of the more complex features affecting the receiver, making the related performance optimization problem amenable to systematic tools from differential and variational calculus. The derived expressions evidence the most favorable three-way unbalance between energy, rate, and reliability for receiver performance. Low-level system simulations are carried out for validation.

Whole Slide Images (WSI) are widely used in histopathology for research and the diagnosis of different types of cancer. The preparation and digitization of histological tissues leads to the introduction of artifacts and variations that need to be addressed before the tissues are analyzed. WSI preprocessing can significantly improve the performance of computational pathology systems and is often used to facilitate human or machine analysis. Color preprocessing techniques are frequently mentioned in the literature, while other areas are usually ignored. In this paper, we present a detailed study of the state-of-the-art in three different areas of WSI preprocessing: Artifacts detection, color variation, and the emerging field of pathology-specific data augmentation. We include a summary of evaluation techniques along with a discussion of possible limitations and future research directions for new methods.

The possibilities that Distributed Ledger Technologies (DLTs) offer for cooperation, development, and achievement of the Sustainable Development Goals (SDGs) are remarkable. This is because DLTs enable several key features, such as sharing complete information about every data transaction in the distributed system that participants belong to, the immutability of the recorded data transactions, or consensus among what data can be regarded as true, of great usefulness for the implementation of the SDGs. As far as developing countries are concerned, this information could be useful in trading locally produced goods, as it could enhance the reputation and profitability of Small-Scale Producers (SSPs). Unfortunately, it is rare to find a digitalized marketplace that has been specifically implemented for this application domain. This paper puts forward a blockchain-based marketplace that makes use of Smart Contracts and offers information about how the sold goods were produced and can be traced to their very origin. Besides, cloud computing has been conceived to be used in this development from the beginning to reduce the computational resources required by end user operations. An implementation with cloud computing facilities, software components running on the Ethereum blockchain and a web front end have been tested with satisfactory performance results.

The main objective of this work is to determine whether social networks can be used to improve self-management skills in people with Intellectual or Developmental Disability (IDD) given the high rate of acceptance of such tools among them. The proposed methodology applies an employment-based intervention to evaluate whether social network services help improve self-management skills. A fit-for-purpose social networking service is built, and different evaluation mechanisms are applied to determine the degree to which users with IDD could use the tool without assistance. The working hypothesis is that in order to ensure the usability of the implemented tool, users (with and without IDD) should be involved throughout the whole software development cycle, following a co-design methodological approach. Results demonstrate that this methodological approach leads to high satisfaction and motivation rates. In terms of effectiveness and efficiency, results demonstrate that the tool is useful and usable for people with IDD. It can therefore be concluded that social network services built having users with IDD at its center will help improve their self-management skills. The main implication of this research is that when using social network services as a tool to support interventions, special attention should be paid to previously evaluating the usability and motivation rate as a measure of how autonomously the person with IDD can use the tool. Nevertheless, further research is needed to determine the impact that such tools can have on other types of intervention.

Micro aerial vehicles (MAVs) can make explorations in 3D environments using technologies capable of perceiving the environment to map and estimate the location of objects that could cause collisions, such as Simultaneous Localization and Mapping (SLAM). Nevertheless, the agent needs to move during the environment mapping, reducing the flying time to employ additional activities. It has to be noted that adding more devices (sensors) to MAVs implies more power consumption. Since more energy to perform tasks is required, growing the dimensions of MAVs limits the flying time. Contrarily, Generative Adversarial Networks (GAN) have demonstrated the usefulness of creating images from one domain to another, but the GAN domain changes require a large number of samples. Therefore, an interoperability coefficient is employed to determine a minimum number of samples to connect the different domains. In order to prove the coefficient, the performance to estimate the depth and semantic mask between authentic and virtual samples with the number limited of samples is analyzed. Consequently, an RGB-D sensor can be replaced by a few samples of a real scenario based on GANs. Although GAN allows creating images with depth and semantic mask information, there is an additional problem to be tackled: the presence of intrinsic noise, where a simple GAN architecture is not enough. In this proposal, the performance of this solution against a physical RGB-D sensor (Microsoft Kinect V1) and other state-of-the-art approaches is compared. Experimental results allow us to affirm that this proposal is a viable option to replace a physical RGB-D sensor with limited information.

In this paper we present a new approach to cooperative overhead object transportation with obstacle avoidance for multi-robot systems. This approach considers situations in which there are obstacles that only pose a risk to the robots involved in the transport and not necessarily to the group of robots with the load. Thus, we propose to separate the workspace into two layers in a way that the obstacles which only pose risk of collision to the robots can be transposed by a local planner. This new approach to the load transportation problem allows the system to find shorter paths or even find a solution in situations which would not be possible with the classic approach. Furthermore, the developed method has a configuration space independent of the number of robots and it is robust to environmental variation, not reducing its performance in more complex environments as in narrow passages or long corridors. The proposed approach was tested by simulation and the results are compared to the classic one.

The annotation of large datasets is often the bottleneck in the successful application of artificial intelligence in computational pathology. For this reason recently Multiple Instance Learning (MIL) and Semi Supervised Learning (SSL) approaches are gaining popularity because they require fewer annotations. In this work we couple SSL and MIL to train a deep learning classifier that combines the advantages of both methods and overcomes their limitations. Our method is able to learn from the global WSI diagnosis and a combination of labeled and unlabeled patches. Furthermore, we propose and evaluate an efficient labeling paradigm that guarantees a strong classification performance when combined with our learning framework. We compare our method to SSL and MIL baselines, the state-of-the-art and completely supervised training. With only a small percentage of patch labels our proposed model achieves a competitive performance on SICAPv2 (Cohen’s kappa of 0.801 with 450 patch labels), PANDA (Cohen’s kappa of 0.794 with 22,023 patch labels) and Camelyon16 (ROC AUC of 0.913 with 433 patch labels). Our code is publicly available at https://github.com/arneschmidt/ssl_and_mil_cancer_classification.

Managing and evolving a chatbot’s content is a laborious process and there is still a lack of standardization. In this context of standardization, the absence of a management process can lead to bad user experiences with a chatbot. This work proposes the Chatbot Management Process, a methodology for content management on chatbot systems. The proposed methodology is based on the experiences acquired with the development of Evatalk, the chatbot for the Brazilian Virtual School of Government. The focus of this methodology is to evolve the chatbot content through the analysis of user interactions, allowing a cyclic and human-supervised process. We divided the proposed methodology into three distinct phases, namely, manage, build, and analyze. Moreover, the proposed methodology presents a clear definition of the roles of the chatbot team. We validate the proposed methodology along with the creation of the Evatalk chatbot, whose amount of interactions was of 22,771 for the 1,698,957 enrolled attendees in the Brazillian Virtual School of Government in 2020. The application of the methodology on Evatalk’s chatbot brought positive results: we reduced the chatbot’s human hand-off rate from 44.43% to 30.16%, the chatbot’s knowledge base examples increased by 160% whilst maintaining a high percentage of confidence in its responses and keeping the user satisfaction collected in conversations stable.

For robotics to become more accessible to people not specialized in the area, it is of fundamental importance to improve and simplify the way people interact with robots. Despite human-robot interaction (HRI) being an effervescent research area, most of the works published so far on the use of gesture interfaces for human-robot communication do not clearly describe how the used gestures were elicited, thus hindering the reproducibility of those works. Considering this, we propose a new and reproducible Frustration-Based Approach (FBA), scientifically established on previous research, which can be used to obtain an intuitive and robust gesture vocabulary for HRI. To accomplish this, we propose Intuitiveness Level (IL), a score to rank gestures according with its intuitiveness. Using IL, it is possible to conceive a complex vocabulary, allowing an increasing of robustness, since more than one gesture can be associated to a task. In a general sense, the proposed methodology is not limited only for HRI, and it can also be used for human-machine interaction in general. In short, the contributions of this work are: (i) A complete methodology to elicit gestures to be used as intuitive communication interface between humans and robots. (ii) A metric of intuitiveness which takes into account at least three different characteristics about the elicited gestures.

Hardware accelerators based on field programmable gate array (FPGA) and system on chip (SoC) devices have gained attention in recent years. One of the main reasons is that these devices contain reconfigurable logic, which makes them feasible for boosting the performance of applications. High-level synthesis (HLS) tools facilitate the creation of FPGA code from a high level of abstraction using different directives to obtain an optimized hardware design based on performance metrics. However, the complexity of the design space depends on different factors such as the number of directives used in the source code, the available resources in the device, and the clock frequency. Design space exploration (DSE) techniques comprise the evaluation of multiple implementations with different combinations of directives to obtain a design with a good compromise between different metrics. This paper presents a survey of models, methodologies, and frameworks proposed for metric estimation, FPGA-based DSE, and power consumption estimation on FPGA/SoC. The main features, limitations, and trade-offs of these approaches are described. We also present the integration of existing models and frameworks in diverse research areas and identify the different challenges to be addressed.

Machine learning (ML) models have demonstrated discriminative and representative learning capabilities over a wide range of applications, even at the cost of high-computational complexity. Due to their parallel processing capabilities, reconfigurability, and low-power consumption, systems on chip based on a field programmable gate array (SoC/FPGA) have been used to face this challenge. Nevertheless, SoC/FPGA devices are resource-constrained, which implies the need for optimal use of technology for the computation and storage operations involved in ML-based inference. Consequently, mapping a deep neural network (DNN) architecture to a SoC/FPGA requires compression strategies to obtain a hardware design with a good compromise between effectiveness, memory footprint, and inference time. This letter presents an efficient end-to-end workflow for deploying DNNs on an SoC/FPGA by integrating hyperparameter tuning through Bayesian optimization (BO) with an ensemble of compression techniques. [ABSTRACT FROM AUTHOR]

Unmanned aircraft have had a great impact on society, their applications in various fields are increasing day by day, so it becomes necessary to propose, develop or validate new designs and concepts that help minimize operating errors or improve the safety of aircraft. One way to understand the operating principle of these unmanned aircraft is through laboratory prototypes or test benches; these prototypes allow experimentation and testing in controlled environments to minimize errors in their operation. This article presents a novel configuration for a laboratory prototype of a tandem-rotor helicopter with two degrees of freedom (DOF), and the dynamic modeling, control, signal filtering, and implementation of the control and filtering system using a microcontroller and a low-cost prototype made with easily accessible materials. This is to promote education, research, and technological development in this area of engineering.

The aim of this work is to prove that it is possible to develop a system able to detect gestures based only on ultrasonic signals and Edge devices. A set of 7 gestures plus idle has been defined, being possible to combine them to increase the recognized gestures. In order to recognize them, Ultrasound transceivers will be used to detect the 2 dimensional gestures. The Edge device approach implies that the whole data is processed in the device at the network edge rather than depending on external devices or services such as Cloud Computing. The system presented in this paper has been proven to be able to measure Time of Flight (ToF) signals that can be used to recognize multiple gestures by the integration of two transceivers, with an accuracy between 84.18% and 98.4%. Due to the optimization of the preprocessing correlation technique to extract the ToF from the echo signals and our specific firmware design to enable the parallelization of concurrent processes, the system can be implemented as an Edge Device.

This paper develops on the process of inverse kinematics (IK) for manipulator robots based on a recent technique from the computer graphics area that has been highlighted by its simplicity and low computational cost. This IK solver is known as Forward And Backward Reaching Inverse Kinematics (FABRIK) suffers from singularities when applied in kinematic chains composed of one degree of freedom (1-DOF) joints. Some extensions of this method have been proposed to incorporate the constraints in some of the most common joints, however, their application over kinematic chains with only 1-DOF joints is still not possible. Since several manipulators have kinematic chains composed of 1-DOF joints, this work presents a new method, named FABRIK-R, to extend the original method for applications in the robotics manipulators field.

System monitors are applications used to monitor other systems (often mission critical) and take corrective actions upon a system failure. Rather than reactively take action after a failure, the potential of fuzzy logic to anticipate and proactively take corrective actions is explored here. Failures adversely affect a system’s non-functional qualities (e.g., availability, reliability, and usability) and may result in a variety of losses such as data, productivity, or safety losses. The detection and prevention of failures necessarily improves a critical system’s non-functional qualities and avoids losses. The paper is self-contained and reviews set and logic theory, fuzzy inference systems (FIS), explores parameterization, and tests the neighborhood of rule thresholds to evaluate the potential for anticipating failures. Results demonstrate detectable gradients in FIS state spaces and means fuzzy logic based system monitors can anticipate rule violations or system failures.

In this paper, a constant speed control of slider-crank mechanisms for machine tools is proposed. A joint-task space hybrid controller based on a second-order sliding mode control and time-base generator was used to guarantee a constant speed trajectory tracking and a complete turn of the mechanism crank. A switching criterion was implemented in order to avoid the singularities located at the two extreme positions of the slider stroke. A trapezoidal speed profile with parabolic blends was designed directly over task space slider trajectory considering a constant cutting speed, the workpiece dimensions and the slider stroke length. Stability of the second-order sliding mode control was validated with the Lyapunov stability theory. Simulations were carried out to verify this approach.

Hardware solutions based on blockchain used in peer-to-peer electricity trading operations have been on rise during the last years. It is expected that due to their usage, it will become easier for prosumers to participate in the power grid on more equal terms when compared to the traditional players that have been settled in this market for the last decades. However, devices used to fully integrate prosumers are scarce and often offer minimal functionalities to perform the task of becoming integrated in those markets. This manuscript puts forward a Constrained Hardware Device enhanced with several software elements related to blockchain and cloud infrastructures, which make possible for any electricity generator or storage system to perform major actions like executing smart contracts, requesting energy prices to a Transmission System Operator and replicating the interchanged data in a cloud computing environment in case there are blockchain node failures. In this way, Renewable Energy Sources can be integrated by means of inexpensive, reliable devices with all the required software components preinstalled, with prosumers being able to further intervene in energy markets.

The cryptanalytic resistance of modern block and stream encryption systems mainly depends on the substitution box (S-box). In this context, the problem is thus to create an S-box with higher value of nonlinearity because this property can provide some degree of protection against linear and differential cryptanalysis attacks. In this paper, we design a scheme built on a human behavior-based optimization algorithm, supported with Self-Organizing Maps to prevent premature convergence and improve the nonlinearity property in order to obtain strong $8 \times 8$ substitution boxes. The experiments are compared with S-boxes obtained using other metaheuristic algorithms such as Ant Colony Optimization, Genetic Algorithm and an approach based on chaotic functions and show that the obtained S-boxes have good cryptographic properties. The obtained S-box is investigated against standard tests such as bijectivity, nonlinearity, strict avalanche criterion, bit independence criterion, linear probability and differential probability, proving that the proposed scheme is proficient to discover a strong nonlinear component of encryption systems.

The rapid evolution of wind energy in reducing CO2 emissions worldwide is undeniable, which is, in fact, expected to continue or even increase its impressive yearly capacity growth. In this regard, optimizing operations and maintenance of wind turbines (WTs) and farms is considered to be one of the options for reducing the levelized cost of electricity of wind energy. This can be achieved by developing innovative condition monitoring methods. To this end, the use of the windowed scalogram difference (WSD) algorithm, based on wavelets, is proposed as an alternative solution, combined with current signature analysis (CSA). The electric generator is one of the major contributors to WT failure rates and downtime, and doubly-fed induction generators (DFIGs) are the dominant technology in variable-speed WTs. In the present work, operational data on an in-service WT DFIG are analyzed over a period of eight months, in contrast to the majority of the studies in this field, which rely on laboratory or simulated data. The evolution of the fault, namely rotor mechanical asymmetry, at an early stage, is analyzed and quantified implementing WSD to the stator current signals, supported by the previous diagnosis achieved through CSA. The combination of CSA and WSD shows strong potential for diagnosing and tracking, respectively, incipient faults in in-service WT DFIGs.