Predictive maintenance is an approach based on technologies such as the Internet of Things (IoT), Artificial Intelligence (AI), Machine Learning (ML), and robust system architectures. It is widely used in industry and is increasingly adopted in the automotive sector, where its application is often limited to newer vehicles and trucks, excluding most older models. To fill this gap, we propose a solution for cars manufactured from 2010 onward with On-Board Diagnostics II (OBD-II). Our system has a low-cost onboard device, the Legacy Internet of Vehicles (IoV), which enables these vehicles to connect to the cloud server. The system’s initial goal is to send telemetry data to the cloud for analysis, identify anomalies, and predict potential future failures. The system alerts users through internet-connected devices when problems are detected, ensuring that notifications are sent directly to the user. By directing processing to the cloud, the system minimizes the complexity required onboard the vehicle, making it an accessible and scalable approach for a wider audience. To validate our solution, we conducted one of the application scenarios for this system. In this first stage, we performed tests with the system, identifying engine temperature alert problems that would immediately notify the user on the registered device, which in this case was the smartphone.

The use of Internet of Things (IoT) technologies has become more readily available with the advent of cyber-physical systems. This context motivates concerns about cybersecurity and the occurrence of malicious attacks in IoT networks. This paper investigates the problem of the automatic detection of cyberattacks in MQTT-based IoT networks by deploying artificial intelligence algorithms for processing traffic data and indicating whether an attack is occurring or not. Thus, this paper trains different artificial intelligence binary classifiers based on machine learning and compares their performance for malicious attack detection in cyber-physical systems with MQTT-based IoT networks. For training and testing the classifiers, we employ the MQTTset dataset which contains many labelled samples with both legitimate traffic and observations under attack occurrence. By analyzing data features and preprocessing, the algorithms achieved good performance in classifying network traffic, contributing to the security of cyber-physical systems.

This Innovative Practice research paper describes an intelligent digital twin platform designed to enhance the virtualization of practical activities in the field of mechatronics education. In recent years, the classic concept of digital twins has significantly evolved in academia and industry. In education, especially in engineering, digital twins have brought countless alternatives for building environments close to the reality of students' practical activities. To improve these activities in the mechatronics area, intelligent digital twins will add added resources and technologies that assist the virtualization process. Therefore, a platform was developed to access remote laboratories in the mechatronics area and help correct the experiments offered. This platform will enable teachers to insert practical activities into remote laboratories on the platform. This platform has three central areas: teaching plans, teachers, and students. In teaching plants, it is possible to create a register of plants that provides information on the characteristics necessary to connect with their equipment and the number of plant elements. In the teachers' area, it is essential to create an activity for students to conduct, and this activity must be conducted for the system to capture the data and images to form a correction model. The correction model uses these data to help intelligent digital twins digitize practical activities. In the student area, it is possible to check all available experiments and their virtualizations and automatically correct the practical experiment.

This innovative practice WIP is about using serious games in robotics education. In the realm of STEM education, the professional development of students is intrinsically linked to practical experiences, such as hands-on exercises, experiments, and laboratory classes. However, the financial implications of acquiring and maintaining mechatronic devices such as robots, automation-related equipment, and instruments for these activities are high. These excessive costs, which cover initial funding and ongoing maintenance, pose challenges for educational institutions in establishing and maintaining well-equipped laboratories. In addition, access to these resources depends on their availability, creating restrictions on student learning opportunities. The state of Amazonas, Brazil, needs to solve these problems; as a city of numerous factories, there is a growing demand for well-trained professionals and a dwindling number of resources to apply to education. To address these challenges, we propose an approach that integrates simulation tools for application in educational games. These tools used to conduct simulations are essential for developing shop floor scenarios without physical equipment, effectively reducing the costs associated with mechanical collisions and problems related to student programming errors. The importance of using simulation tools in the learning process lies in their ability to offer a risk-free environment where students can experiment, learn, and perfect their skills without incurring additional expenses for corrective maintenance. Innovatively, the incorporation of gamification-based learning by simulating real problems based on the shop floor is gaining momentum. We believe this is the most attractive way for trainees or students to deliver fast, high-quality professionalism. This approach represents an appealing alternative for educators looking to enhance their educational experience. From this perspective, this paper presents the application of a serious game in the industrial laboratory at UFAM. The aim is to offer students a dynamic and engaging platform for learning about collaborative robots, taking advantage of the principles of the learning factory, and promoting good practices in an economical and risk-controlled virtual environment.

This research-to-practice full paper focuses on the use of technology in education, which enables a wide variety of activities. For example, a computer with internet access allows students to read articles, make video calls, perform group work, and even take tests. In 2021, this tool was relevant to face-to-face teaching because the interruption caused by COVID-19 resulted in social distancing. In this way, remote teaching was adopted as an emergency measure in the teaching process. Although remote teaching uses technological innovation, it lacks the inclusion of educational theories and remote practices in virtual classes. Thus, this approach analyzes Lev Vygotsky's theories of learning, specifically the zone of proximal development and scaffolding, to understand the satisfaction of learning the content alone or with other people's help through the Industrial Internet of Things (IIoT) technology. It then investigated IIoT, intending to build a remote virtual environment. In this context, it was possible to develop a remote teaching method called Hobots. This remote teaching method comprises six collaborative modules (the integrating platform, interfaced, MQTT broker, communication module, server, and Robot-X) for the remote teaching of industrial robotics. In addition, the method was evaluated by ten students using the ISO 9241 - Part 10 - Dialog Systems standard, with a view to their satisfaction with using the method in the remote teaching process, and the results were admissible. Therefore, in addition to theory and simulation, the proposed method stands out for the possibility of remote implementation.

This work aims at developing a fault detection system for photovoltaic systems (PVSs) based on the Fuzzy C-Means (FCM) clustering technique. A simulator is developed based on a mathematical model of the PVS by considering three groups of faults: shading, short circuit, and junction box to investigate the fault effects in photovoltaic systems. Several tests are carried out and indicate that the FCM is able to detect these faults and separate the data of a faulty PVS from that of a healthy one based on uncertain data obtained from different weather conditions.

Concepts such as the Internet of Things, big data, cloud computing, and cyber-physical systems support new scenarios for further research and application on the Industry 4.0 initiative. One of the most promising ones is the digital twin applied to the shop floor. In this work-in-progress paper, we briefly review fundamental concepts and definitions and present a lasting solution to problems related to the reliability of industrial equipment. More precisely, our approach is based on an industrial twin model that, after proving its reliability, shall be used in more sophisticated applications. Finally, we present a case study of a storage process plant that uses the proposed architecture to perform a reliability study. The results of this work in progress may contribute to understanding the needs and gaps that may be filled before reaching a consolidated product, and we would like to discuss them with the community.

The COVID-19 pandemic has imposed numerous restrictions on face-to-face meetings, and one of the most impacted activities was engineering education. Most teaching professionals have adopted diverse solutions based on distance learning techniques. However, practical experiments that require laboratories could not be performed in most cases. One of the subjects studied in modern engineering that has been impacted is related to industrial automation systems. In April 2020, amidst tighter restrictions on social contact, we started cooperating with a local company to train our students on topics of interest to the corporation. Our research group was responsible for dealing with Industry 4.0 and its associated technologies. Over the past two years, we have conducted this interaction with the industry under the pandemic's constraints and obtained very interesting results. In this report, on an innovative experience in engineering education, we will describe our project’s syllabus, how the relationship was between students and teachers through online tools, and mainly how we managed to ensure that students had access to experiments close to the industrial reality.

Remote Access Labs (RAL) are hardware and software tools that allow students to remotely operate actual equipment in physical labs at schools or universities. This work presents the development of a system for students to use remote access laboratories in the mechatronics area without human supervision. This system verifies the activities the registered labs can offer through algorithms that use digital twin technology to learn the actual or virtual industrial plants’ operations. Registered laboratories are available for the practice of experiments where students can automatically follow the correction of activities through virtualization generated so that the student can understand their actions. With this system, students can monitor their learning in real time and receive evidence of successes and errors through gamification of their performance.

Much has been discussed recently about the implications and strategies adopted by higher education institutions concerning the realization of online practical classes due to COVID-19. Some teaching institutions used virtual meetings to reorganize the lesson plan to continue teaching and assessing the students due to the suspension of face-to-face classes due to this epidemic. Although the most adopted strategy among the courses has been direct communication between students and professors through e-mail, telephone, social networks, and message apps, education lives in a time when transitions seem to occur much faster than in recent decades. For example, online engineering courses, specifically robotics, face difficulty implementing practical activities online because the educational tools available remotely are scarce or very expensive, thus becoming an obstacle to circumvent. Given this, this article presents a methodological strategy for the teaching-learning practices of engineering through Learning-IoT. This is a methodology for teaching robotics using the Internet of Industrial Things (IIoT) concepts. It enables the connection of the student to a Physical Objects Framework proposed in practical online activities and by the project method responsible for stimulating critical thinking. Thus, the initial ergonomic interface test experiments and usability of the proposed methodology demonstrated in the experiments in online practical classes that memorization and decision-making fit new possibilities or functionalities.