https://repositorio.ufu.br/handle/123456789/5464 2026-07-16T03:19:25Z https://repositorio.ufu.br/handle/123456789/48815 Title: Criando gêmeos digitais no Renode: um estudo de caso com microcontrolador STM32L4 e rádio Zigbee AT86RF233 Abstract: The consolidation of cyber-physical systems and Internet of Things (IoT) applications has increased the demand for verification and validation methods that reduce cost, risk, and development time in resource-constrained devices. Although digital twins are often employed at higher abstraction layers (e.g., cloud integration, visualization, and system-level analytics), faithfully reproducing low-level hardware behavior remains a bottleneck for firmware validation on microcontrollers, especially for low-power families that are still underrepresented in simulation platforms. In this context, this work presents an extension of the Renode simulator, an open-source development framework designed to accelerate embedded system and IoT development by enabling emulation of physical hardware, by implementing a new firmware-level digital twin for the STM32L4 family. The proposed platform is instantiated and validated on the STM32L496 microcontroller and simulated with the IEEE 802.15.4 AT86RF233 transceiver. The proposed approach follows incre- mental, test-oriented development, building a complete new platform through .repl descriptions, C# peripheral models, and .resc automation scripts, covering func- tional integration across subsystems. In addition, this work introduces a newly developed power consumption estimator that accumulates energy statistics over the simulation virtual time, selecting typical current profiles according to frequency, vol- tage, and run/sleep operating modes based on device datasheet metrics. Experimental results demonstrate the execution, without modifications, of a production firmware binary (.elf), whose state machine simultaneously exerci- ses initialization, ADC acquisition with DMA transfers, timer-based timing, radio communication, RX/TX windows and acknowledgment handling, and transitions to low-power operation. Instrumentation via a packet hook enabled frame interception and UDP forwarding for external decoding of telemetry, increasing protocol obser- vability and supporting the validation of RX/TX flows in the simulated medium. Within the defined scope, the power estimation was applied only to the STM32L496 microcontroller and did not include the radio module consumption, serving as a proof of concept for reproducible energy analyses during simulation. 2026-03-06T00:00:00Z https://repositorio.ufu.br/handle/123456789/48700 Title: Deep learning-based detection of carotid artery atheromas in panoramic radiographs Abstract: Stroke is among the leading global causes of mortality and long-term disability resulting from neurological sequelae, with carotid atherosclerosis being one of its main etiological mechanisms. This condition often progresses silently, which limits the use of standard diagnostic examinations such as Doppler ultrasound and computed tomography angiography (CTA) for population screening. However, calcifications located at the carotid bifurcation may be incidentally visualized on panoramic dental radiographs, making this exam a potential opportunistic screening tool due to its wide availability, low cost, and coverage of the C3–C4 region, where atheromas usually manifest. In this context, the present study aimed to develop and evaluate a deep-learning-based method using the MobileNetV2 architecture for the automatic detection of carotid atheromas in panoramic radiographs, aiming to optimize the clinical value of this routine exam as a screening tool. A total of 378 publicly available and fully anonymized Regions of Interest (ROIs) were used, cropped in the carotid region (640×320 px) and divided into training, validation, and test sets (264/57/57). The images underwent normalization, grayscale channel replication, and real-time data augmentation, and class imbalance was addressed using weighted loss. The model was implemented using MobileNetV2 with pretrained weights and a two-stage training scheme, consisting of initial backbone freezing followed by partial fine-tuning (~70%), with batch normalization and dropout (0.3). On the independent test set, the model achieved robust and well-balanced performance, with 94.7% accuracy, 95.7% sensitivity, 94.1% specificity, and AUC and AUPRC values of 0.963 and 0.968, respectively, using a threshold optimized by Youden’s J index. These results demonstrate strong discriminative capability and reinforce the potential of panoramic radiography combined with deep learning as an opportunistic screening tool for carotid atheromas. The proposed method shows promising performance and supports the use of routine panoramic radiographs for opportunistic screening, enabling the early referral of patients for specialized vascular evaluation and strengthening the interface between oral and systemic health. 2026-02-27T00:00:00Z https://repositorio.ufu.br/handle/123456789/48387 Title: LoRa e IEEE 802.15.4 a 2,4 GHz: Análise Comparativa de Comunicação Sem Fio e Viabilidade Energética com Fontes Renováveis em Cenários Industriais Abstract: The Fourth Industrial Revolution, characterized by the interconnection of systems, relies on the Internet of Things (IoT) as a fundamental pillar, fostering the convergence between the physical and digital worlds. Within the Industrial IoT (IIoT) landscape, technologies such as ZigBee and LoRa have driven the adoption of monitoring solutions due to their flexibility and low power consumption. However, ensuring network reliability and energy autonomy remains a critical challenge for scalability. In this context, this study presents a comparative analysis of two wireless solutions operating in the 2.4 GHz ISM band: the IEEE 802.15.4 standard, implemented via the AT86RF233 transceiver, and the LoRa physical layer (PHY), utilizing the SX1280 transceiver. Specifically, the work evaluates the robustness of LoRa’s proprietary modulation against the standardized IEEE protocol, while also investigating the feasibility of applying photovoltaic energy harvesting. Experimental results indicate that IEEE 802.15.4 exhibited greater stability in packet delivery, maintaining success rates above 90% at distances up to 150 meters, albeit with significantly higher energy consumption due to the external amplification required to reach 12 dBm. In contrast, LoRa technology demonstrated superior energy efficiency, consuming approximately half the current in the tested configurations, while maintaining competitive transmission rates. Additionally, the technical feasibility of photovoltaic panels was confirmed, as they successfully supplied the full energy demand during continuous transmission under high solar irradiance conditions. 2026-02-06T00:00:00Z https://repositorio.ufu.br/handle/123456789/48306 Title: Uma arquitetura de ambiente virtual interativo suportada por imagens aéreas e esféricas para visualização de dados e de diagrama unifilar de subestações de energia elétrica Abstract: Electrical substations are critical infrastructures whose complexity and high voltage levels require solutions that ensure safety, agile navigation, and efficient maintenance. This research developed a navigable digital platform designed to integrate and visualize orthomosaic images generated by drones, CAD (Computer-Aided Design) plans, and spherical georeferenced images captured at different points in the substation yard. The interactive interface allows switching between layers and links each piece of equipment to its single-line diagram, enabling immediate access to technical specifications and safety guidelines. The methodology involves visual data capture, photogrammetric processing, georeferencing via control points, and integration into a database with each asset's attributes. The results demonstrate that virtual navigation reduces the time needed for inspection, accelerates the identification of critical equipment, and improves communication between teams, speeding up response times in preventive and corrective maintenance. The usage of drones, digital modeling, and spherical images in a unified environment represents a significant advance in modernizing the substation operation and monitoring practices, promoting greater efficiency, safety, and reliability in the management of these critical assets. 2025-12-19T00:00:00Z