Abstract
Currently, distributed software systems have evolved at an unprecedented pace. Modern software-quality requirements are high and require significant staff support and effort. This study investigates the use of a supervised machine learning model, a Multi-Layer Perceptron (MLP), for anomaly detection in microservices. The study covers the creation of a microservices infrastructure, the development of a fault injection module that simulates application-level and service-level anomalies, the creation of a system monitoring dataset, and the creation and validation of the MLP model to detect anomalies. The results indicate that the MLP model effectively detects anomalies in both domains with higher accuracy, precision, recovery, and F1 score on the service-level anomaly dataset. The potential for more effective distributed system monitoring and management automation is highlighted in this study by focusing on service-level metrics such as service response times. This study provides valuable information about the effectiveness of supervised machine learning models in detecting anomalies across distributed software systems.

Abstract
The accelerated changes on our planet have led to a growing interest in climate change and its consequences: natural hazards and adverse socio-economic impacts. However, the development of climate research and the proliferation of datasets require an integrated and efficient approach to the analysis, investigation, and visualization of atmospheric meteorological data. Thus, we propose a literature review of existing systems viewing meteorological phenomena in four and three dimensions. Moreover, we evaluate meteorological occurrences to better understand the dynamics associated with a meteorological phenomenon and visualize different weather data. Based on the investigation of tools and methods, we consider the existence of different ways of representing meteorological data and methodologies. However, it was imperative to obtain knowledge and create our way of visualizing weather data. This article found eleven existing solutions for 4D meteorological visualization and meteorological phenomena.

Abstract
The advance of industrialization regarding the optimization of production to obtain greater productivity and consequently generate more profits has led to the emergence of Industry 4.0, which aims to create an environment called smart manufacturing. On the other hand, the Internet of Things is a global network of interrelated physical devices, such as sensors, actuators, intelligent applications, computers, mechanical machines, objects, and people, becoming an essential part of the Internet. These devices are data sources that provide abundant information on manufacturing processes in an industrial environment. A concern of this type of system is processing large sets of data and generating knowledge. These challenges often raise concerns about security, more specifically cybersecurity. Good cybersecurity practices make it possible to avoid damage to production lines and information. With the growing increase in threats in terms of security, this paper aims to carry out a review of existing technologies about cybersecurity in intelligent manufacturing and an introduction to the architecture of the IoT and smart manufacturing.

Abstract
The Industry 4.0 (I4.0) paradigm comes as a direct action by the German government to improve the industrial production process, by enhancing the ability to take action during the process and produce customized products, while maintaining the mass production industrial context. The I4.0 solutions rely on cybernetics systems that can enhance the users' decision-making. Some technologies are particularly suited for this purpose, including data science combined with context sensitive applications and virtual assistants (VA). These types of user application can provide information and features according to the user's context, thus acting proactively and foreseeing the user actions. In this work, we partnered with Continental Advanced Antenna - a manufacturer of radiofrequency devices for the auto industry, to further develop the concept of a VA to assist the production management. A prototype was built to interact and keep the production management team up to date regarding the ongoing execution of the production plan.

Abstract
Data and services are available anywhere at any time thanks to the Internet and mobile devices. Nowadays, there are new ways of representing data through trendy technologies such as augmented reality (AR), which extends our perception of reality through the addition of a virtual layer on top of real-time images. The great potential of unmanned aerial vehicles (UAVs) for carrying out routine and professional tasks has encouraged their use in the creation of several services, such as package delivery or industrial maintenance. Unfortunately, drone piloting is difficult to learn and requires specific training. Since regular training is performed with virtual simulations, we decided to propose a multiplatform cloud-hosted solution based in Web AR for drone training and usability testing. This solution defines a configurable trajectory through virtual elements represented over barcode markers placed on a real environment. The main goal is to provide an inclusive and accessible training solution which could be used by anyone who wants to learn how to pilot or test research related to UAV control. For this paper, we reviewed drones, AR, and human-drone interaction (HDI) to propose an architecture and implement a prototype, which was built using a Raspberry Pi 3, a camera, and barcode markers. The validation was conducted using several test scenarios. The results show that a real-time AR experience for drone pilot training and usability testing is achievable through web technologies. Some of the advantages of this approach, compared to traditional methods, are its high availability by using the web and other ubiquitous devices; the minimization of technophobia related to crashes; and the development of cost-effective alternatives to train pilots and make the testing phase easier for drone researchers and developers through trendy technologies.