CT - TCC - Engenharia de Computação

Permanent URI for this collectionhttps://repositorio.ufrn.br/handle/123456789/33188

Browse

Now showing 1 - 20 of 237

Navegação autônoma local: estudo empírico sobre o ajuste de modelos de linguagem-visão compactos para o uso de interfaces gráficas
(Universidade Federal do Rio Grande do Norte, 2026-07-02) Rodrigues, Thiago Fernandes Pinheiro de Medeiros; Oliveira, Luiz Affonso Henderson Guedes de; Prolo, Carlos Augusto; Santos, Breno Santana
The autonomous automation of graphical user interfaces (GUIs) by computer use agents (CUAs) is dominated by large models, whose dependence on large-scale infrastructure imposes cost, latency, and privacy barriers. These barriers motivate interest in compact agents executable on consumer hardware, but the appropriate training techniques for this scale range and their main bottlenecks remain poorly characterized. Given this, this work empirically investigates whether reinforcement learning fine-tuning improves the performance of a compact CUA, in the four-billion-parameter range, trained exclusively under commercial hardware constraints. To this end, a vision-language model (VLM) from the Qwen3-VL family is fine-tuned on the AgentNet dataset through reinforcement learning (RL), adopting the Group Relative Sequence Policy Optimization (GSPO) algorithm under an off-policy regime, utilizing single-step episodes and a multi-criteria reward function. The resulting model is evaluated on the ScreenSpot-Pro and OSWorld-Verified benchmarks in a controlled comparison with the equivalent base network. The results reveal a consistent gain in spatial grounding, with overall accuracy improving from 33.7% to 38.7%, primarily concentrated in the localization of textual elements. In contrast, a slight regression in long- horizon agentic performance is observed, dropping from 15.1% to 11.3%, predominantly attributed to an under-termination failure rather than a loss of perceptual or interface manipulation capabilities. An analysis of internal metrics associates this behavior with the scarce representation of the termination action and the short training duration, indications reinforced by a preliminary training phase that demonstrates the feasibility of a calibrated termination policy when this action receives sufficient representation. It is concluded that the reinforcement alignment strategy is structurally viable for compact architectures and highly amenable to optimization, providing a documented empirical baseline for future investigations into local CUAs.
Votação eletrônica descentralizada com blockchain e Proof of Authority
(Universidade Federal do Rio Grande do Norte, 2026-06-16) Silva, Mateus Cavalcanti Alves Teixeira; Falcão, Eduardo de Lucena; Souza, Samuel Xavier de; Viegas, Carlos Manuel Dias
Electronic voting systems require mechanisms capable of preserving integrity, vote uniqueness, availability, and auditability. In digital architectures, these requirements depend not only on the voting interface, but also on how the system records votes, validates transactions, and organizes trust among participants. This work proposes, implements, and evaluates a prototype of a decentralized electronic voting system based on peer-to-peer networking, blockchain, and Proof of Authority (PoA) consensus, focusing on permissioned networks composed of previously known validators. The solution was implemented in Go using the libp2p library, adopting a modular architecture, event-driven processing, and deterministic round-robin leader selection. Experimental validation was conducted in a controlled environment with three validator nodes through tests covering result integrity, double-voting prevention, and crash-stop fault tolerance. The results showed that the system correctly finalizes valid votes, preserves vote uniqueness in the final blockchain state, and maintains operation after the failure of one validator. These findings indicate that the proposed approach is technically feasible within the evaluated experimental scope and provides a basis for future improvements in smaller institutional scenarios.
Plataforma web para gerenciamento e execução de automações contábeis com aplicação prática em escritório parceiro
(Universidade Federal do Rio Grande do Norte, 2025-07-24) Silva, Elder Santos; Pedrosa, Diogo Pinheiro Fernandes; Balza, Micael; Costa, Samuel Alves da
Estudo comparativo de variantes de autoencoders no algoritmo de agrupamento profundo K-Deep Autoencoder
(Universidade Federal do Rio Grande do Norte, 2025-12-12) Freire, Magnus Brígido Paulo; Silva, Leonardo Enzo Brito da; https://orcid.org/0000-0003-3639-9175; http://lattes.cnpq.br/8092398102384904; http://lattes.cnpq.br/7371306960224864; Martins, Daniel Lopes; http://lattes.cnpq.br/2947003499946183; Silva Junior, Silvan Ferreira da; https://orcid.org/0000-0001-6269-494X; http://lattes.cnpq.br/0669123874358910
Proposta de implementação totalmente paralela do algoritmo k-Nearest Neighbors em FPGA
(Universidade Federal do Rio Grande do Norte, 2025-12-11) Damasceno, Augusto Matheus Pinheiro; Fernandes, Marcelo Augusto Costa; https://orcid.org/0000-0001-7536-2506; http://lattes.cnpq.br/3475337353676349; https://orcid.org/0000-0001-8457-3911; https://lattes.cnpq.br/7303360052203792; Goldbarg, Mateus Arnaud Santos de Sousa; https://orcid.org/0000-0001-7536-2506; http://lattes.cnpq.br/5393775937989569; Alves, Pedro Victor Andrade; https://orcid.org/0009-0003-2176-793X
The growing demand for Artificial Intelligence applications based on Large Language Models (LLMs) has driven the development of techniques such as Retrieval-Augmented Generation (RAG), which combines information retrieval with text generation to reduce hallucinations and improve response accuracy. However, the vector search stage in RAG presents a significant computational bottleneck, especially in systems requiring low latency and energy efficiency. This work proposes a fully parallel implementation of the k-Nearest Neighbors (k-NN) algorithm on FPGA (Field-Programmable Gate Array) for accelerating searches in vector databases. The developed architecture uses fixed-point arithmetic instead of floating-point to significantly reduce logic resource consumption and system latency. The implementation was carried out using Xilinx System Generator integrated with MATLAB/Simulink, targeting the Virtex-6 XC6VLX240T-1FFG1156 FPGA present in the ML605 kit. The system comprises specialized modules for quantization, inner product similarity score calculation, and k-nearest neighbor selection, exploiting massive hardware parallelism. The results demonstrate that the system occupied low area, enabling the algorithm to be scaled to process a larger number of vectors. The total system latency was 18,6 ns, making the solution viable for real-time inference applications of RAG systems.
Representação matemática e agrupamento de assovios de golfinhos a partir de contornos tempo-frequência
(Universidade Federal do Rio Grande do Norte, 2025-12-08) Vilar Neto, Pedro Rêgo; Oliveira, Luiz Affonso Henderson Guedes de; Gendriz, Ignacio Sánchez; http://lattes.cnpq.br/6338710569530857; http://lattes.cnpq.br/7987212907837941; https://lattes.cnpq.br/9762811944048548; Tiago Tavares Leite Barros; http://lattes.cnpq.br/1321568048490353; Ferreira, Maria Luane Stamatto; http://lattes.cnpq.br/8906977726202357
This work presents a computational pipeline for the automatic characterization of dolphin whistles from spectrograms collected using autonomous underwater acoustic recorder. The vocalizations were manually segmented and annotated to generate time–frequency masks, which were then converted into tonal contours. Each contour was smoothed and fitted using degree-9 Chebyshev polynomials, producing a compact set of features that describe its frequency modulation. Based on these representations, dimensionality reduction was performed using UMAP, followed by density-based clustering with HDBSCAN, resulting in 22 main whistle groups. Analysis of the mean curves revealed distinct tonal patterns, including ascending, descending, convex, and concave profiles, consistent with structural tendencies observed in real spectrograms. While some clusters showed strong correspondence with traditional manual labels, others displayed greater internal heterogeneity, indicating transitional regions between different modulation shapes. The results demonstrate the effectiveness of the proposed approach in organizing whistle structures and identifying emerging acoustic patterns, providing a promising quantitative basis for future research in automatic classification and behavioral analysis in bioacoustics.
Mulheres em TI e Computação: diagnóstico da desigualdade de gênero na UFRN
(Universidade Federal do Rio Grande do Norte, 2025-12-09) Barbosa, Jaysa Keylla Siqueira; Pereira, Mônica Magalhães; Alsina, Pablo Javier; Silva , Lyrene Fernandes da
The underrepresentation of women in Information Technology and Computer Science is a widely documented phenomenon in the literature, characterized by the concept of “leaky pipeline”, which describes the progressive loss of women throughout their academic and professional trajectory. This study investigated female participation in Information Technology and Computer Science courses and departments at the Federal University of Rio Grande do Norte (UFRN), specifically at the Instituto Metrópole Digital (IMD), the Departamento de Engenharia de Computação e Automação (DCA), and the Departamento de Informática e Matemática Aplicada (DIMAP). The adopted methodology used a quantitative approach, with analysis of secondary data obtained from the UFRN Open Data Portal and institutional records provided by the Information Technology Superintendence. The analysis included 16,671 student records from 2010 to 2025 and 117 active faculty members, in addition to historical data on career progression and management positions. The results revealed that women represent only 13,9% of the student body and 20.5% of the faculty, confirming the leaky pipeline phenomenon. Temporal analysis showed a negative trend in female participation over time, with a reduction of 11.9 percentage points between 2010 and 2025. Regarding career progression, the study differentiated vertical progressions (class changes) and horizontal progressions (level advancements), revealing distinct patterns of equity: vertical progressions showed a significant difference between genders (p=0.0495), while horizontal progressions demonstrated exemplary equity (p=0.6523), with an average interval of approximately 2 years for both genders. Analysis of management positions revealed the existence of a “glass ceiling”: while women have proportional participation in course coordinatorships, their presence significantly decreases in department head and directorship positions. The study concludes that gender barriers in IT and Computer Science areas at UFRN manifest both in access to the field and in advancement to leadership positions, and that more regulated mechanisms tend to produce greater gender equity, which demands institutional policies that promote greater transparency and standardization in progression processes.
Assistência na geração de documentação com base no processamento de linguagem natural e inteligência artificial
(Universidade Federal do Rio Grande do Norte, 2025-11-24) Madruga, João Victor Moura Monteiro; Doria Neto, Adriao Duarte; Martins, Daniel Lopes; Santos, Keylly Eyglys Araújo dos
Technical documentation plays a crucial role in software engineering, ensuring system maintainability and evolution; however, its manual production is often costly and neglected, resulting in technical debt. This work analyzes the application of Artificial Intelligence, with an emphasis on Natural Language Processing (NLP) and the Transformer (GPT) architecture, as an assistance tool for the generation and maintenance of technical documentation. The methodology is characterized as a literature review composed of articles and undergraduate theses, using search platforms such as Google Scholar and Portuguese- and English-language databases from Scielo, published between 2019 and 2025. The results indicate that the adoption of AI-based assistants promotes a reduction in the time required to produce documents, in addition to increasing standardization and quality management. It is concluded that the integration of generative models is viable and effective for mitigating documentation obsolescence, provided that human supervision (human-in-the-loop) is maintained to validate technical accuracy and mitigate AI hallucinations.
Aplicação de estêncils compactos e reordenamento de dados na modelagem de ondas acústicas
(Universidade Federal do Rio Grande do Norte, 2025-12-08) Freire, Lucas Morais; Souza, Samuel Xavier de; Schneider, Kleiton Andre; Barros, Tiago Tavares Leite; Tiago Tavares Leite Barros
This work presents the motivations, implementation details and tests of two acoustic wave modeling algorithms: the Reordered Acoustic Modeling and the Compact Reordered Acoustic Modeling. Seismic wave modeling is an utmost important process in several areas of geophysics and seismic imaging and processing. Due to temporal and spatial cache locality matters, these algorithms were developed to try to improve the performance of wave modeling in both CPUs and GPUs alike. They were implemented using Mamute as a base, which is a seismic imaging and processing software, and both showed performance improvements for larger problem when compared to other algorithms, justified by the reduction of the cache-miss rate.
Plataforma Sapparchi: integração de comunicação assíncrona e roteamento inteligente em arquiteturas edge-fog-cloud continuum
(Universidade Federal do Rio Grande do Norte, 2025-12-08) Maia, Rafael Pinheiro Carlos; Cacho, Nélio Alessandro Azevedo; Lopes, Frederico Araujo da Silva; Souza, Arthur Emanoel Cassio da Silva e
This work presents the development and integration of an asynchronous messaging system based on RabbitMQ for the Sapparchi platform, a distributed task orchestration solution for Edge-Fog-Cloud environments. The proposed solution implements asynchronous communication with support for multiple protocols, including AMQP for Fog and Cloud nodes and MQTT for resource-constrained edge devices, with HTTP as a fallback mechanism. An adaptive routing algorithm was developed to automatically select the most suitable protocol according to the computational layer of the target node. The work also includes the implementation of a node health monitoring mechanism with circuit breaker, capable of detecting failures and preventing requests from being sent to unavailable workers. Finally, automatic registration and re-registration mechanisms were created to ensure system recovery after Manager or Worker failures. Qualitative results indicate increased system resilience and better suitability for intermittent connectivity scenarios typical of edge devices.
Desenvolvimento de um sistema administrativo dinâmico e acessível para o site do LAPPS
(Universidade Federal do Rio Grande do Norte, 2025-12-03) Barros, Alex Pereira; Barros, Tiago Tavares Leite; Viegas, Carlos Manuel Dias; Falcão, Eduardo de Lucena
The Laboratory of Parallel Architectures for Signal Processing (LAPPS), at UFRN, previously maintained its institutional website in a predominantly static way, with content hardcoded into the source code. This approach created a strong dependency on developers for simple updates, leading to errors, outdated information and limited compliance with digital accessibility guidelines, especially WCAG 2.1 and Brazilian regulations. This work aimed to develop a dynamic and accessible Academic Administrative System for the LAPPS website, enabling autonomous management of publications, people, projects, research applications and news. The solution was implemented in PHP with PDO and MySQL, using a modular architecture based on the Page Controller pattern, with CRUD modules, authentication with session control, and security best practices (parameterized queries, data validation and sanitization and secure password hashes). WCAG 2.1 level AA guidelines were incorporated through semantic HTML5, keyboard navigation, visible focus and adequate color contrast. Evaluation with functional tests, tools such as Google Lighthouse and WAVE, and manual inspection with screen readers showed that the defined requirements were met, with accessibility scores ≥ 90 on main screens. The system significantly reduced the effort required to update the website, decentralized content management and made the portal more accessible and aligned with digital inclusion practices, representing a solution that can be replicated by other research groups.
Implementação de uma interface gráfica em python/tkinter para o algoritmo de estereotomografia
(Universidade Federal do Rio Grande do Norte, 2025-12-04) Lavor, Gabriel Cruz de; Barros, Tiago Tavares Leite
Adaptação da StyleGAN2 para geração bidirecional de faces individuais via transfer learning
(Universidade Federal do Rio Grande do Norte, 2025-11-24) Silva, Jonas Peixoto da; Dória Neto, Adrião Duarte; Santos, Keylly Eyglys Araújo dos; Martins, Daniel Lopes; Oliveira, Thiago Henrique Freire de
This work presents an approach to optimize the generation of individual faces using transfer learning based on the StyleGAN2 network. The objective is to reduce training time and avoid the need to start each facial generation from scratch while maintaining high quality in the generated images. An architecture was adapted and coupled to the pretrained generator, aiming to promote specialization, as well as the selective freezing of parts of the original generator and the discriminator to enable processing optimization. Subsequently, the approach was combined with an independent encoder responsible for mapping real images to the latent space, allowing the reconstruction of faces from their latent vectors. The obtained results demonstrated consistent generations and reconstructions, evidencing the potential of the methodology for personalized facial synthesis applications.
Arquitetura de monitoramento e observabilidade de aplicações em Kubernetes
(Universidade Federal do Rio Grande do Norte, 2025-08-01) Lima, Cauan Marcelo; Viegas, Carlos Manuel Dias; 0009-0008-5097-9444; http://lattes.cnpq.br/7003477479031891; Falcão, Eduardo de Lucena; Silva, Ivanovitch Medeiros Dantas da
In a scenario of increasing digital transformation, high availability, resilience, and service scalability have become market requirements as they directly impact user experience. Concurrently, Kubernetes has emerged as a hosting platform that offers numerous benefits, yet by itself it does not provide a complete set of monitoring and observability features. This study presents an alternative based on open-source tools capable of delivering an environment that enables diagnostics and supports the improvement of hosted services. The proposed architecture provides an efficient, accessible, and robust model for monitoring and observing applications running in containers orchestrated by Kubernetes. By integrating components such as Argo CD, HashiCorp Vault for secure secrets management, and an observability toolset composed of OpenTelemetry, Prometheus, Grafana Tempo, Coroot, and Netdata, the approach was experimentally validated through scenarios involving service execution with databases, CPU, memory, and disk stress tests, and the simulation of systems with failures. The results demonstrate the architecture’s ability to identify bottlenecks, detect incidents, monitor service health, and record anomalies within the cluster and hosted systems. It is concluded that it is fully feasible to build a modern and reliable observability environment using technologies that ensure security, traceability, and diagnostics, establishing a solid foundation for Kubernetes environments aligned with contemporary operational demands.
Construção e aprimoramento de sistema para gestão de processos em um setor de produção de indústria têxtil
(Universidade Federal do Rio Grande do Norte, 2025-12-03) Queiroz Neto, Fernando Manoel de; Silva, Ivanovitch Medeiros Dantas da; http://lattes.cnpq.br/3608440944832201; http://lattes.cnpq.br/6676378921167560; Medeiros, Morsinaldo de Azavedo; http://lattes.cnpq.br/2910614991697297; Medeiros, Thaís de Araújo de; http://lattes.cnpq.br/4746141228253120
The growing need for operational excellence and competitive advantage requires large industrial players to strategically adopt data-based management tools. In this context, this paper details the improvement and professionalization of a Virtual Visual Management (Electronic Kanban) system applied to the knitwear sector of a large textile industry. Since the system, initially conceived as a functional prototype, had serious engineering shortcomings, such as dependence on local infrastructure, low availability, and security vulnerabilities, the overall objective was to transform it into a robust and secure production environment. To this end, the methodology involved a complete re-engineering of the architecture, including the migration of infrastructure to a centralized and monitored Virtual Machine (VM), the implementation of information security via an encrypted vault (CyberArk) and strict access control, as well as the functional expansion of the scope to new critical processes, including the Review, Routing and Preparation modules. Contributions in Computer Engineering ensured data reliability and traceability, enabling the identification of root causes of defects, which directly resulted in a reduction in the average number of daily setups from 12.5 to 11 and a 3% decrease in the overall reprocessing of the mesh produced, thus demonstrating the value of the rigorous application of infrastructure and system security standards in industrial environments.
Desenvolvimento de um sistema web de controle financeiro para a secretaria de carnaúba dos dantas com aplicação de boas práticas de engenharia de software e acessibilidade digital
(Universidade Federal do Rio Grande do Norte, 2025-12-05) Roque, Ivamar Dantas; Dantas, Rummenigge Rudson; http://lattes.cnpq.br/1868960602254610; http://lattes.cnpq.br/6952603524612306; Burlamaqui, Aquiles Medeiros Filgueira; http://lattes.cnpq.br/8670475877813913; Silva, Bruno Marques Ferreira da; http://lattes.cnpq.br/7878437620254155
Efficient cost management is a central pillar in public administration, yet many orga- nizations still rely on traditional methods, such as manual records or spreadsheets, which are error-prone and hinder managerial analysis. Against this backdrop, this work propo- sed the Development of a Web Financial Control System for the Secretariat of Carnaúba dos Dantas. The general objective was to create a robust solution (ConSec) by articula- ting good Software Engineering practices with Digital Accessibility (WCAG) principles, ensuring usability and inclusion for all employee profiles. The methodology adopted the Client-Server architectural model, with the Backend implemented as a RESTful API in .NET Core (C#), responsible for business logic and data persistence via Entity Fra- mework Core (ORM), and the Frontend as an Angular (TypeScript) Single Page Appli- cation (SPA), focused on user interaction. Accessibility was treated as a primary Non- Functional Requirement (NFR), applying the four WCAG principles (Perceivable, Ope- rable, Understandable, and Robust - POUR) from the interface design phase. The results show the development of a functional and accessible web application, featuring operatio- nal modules (cost registration and auditing) and a Managerial Dashboard that transforms raw data into visual insights for decision-making. The system successfully passed ac- cessibility audits using tools like Lighthouse and manual tests (keyboard navigation and NVDA), confirming adherence to the WCAG standard. Finally, ConSec was validated by fully processing and analyzing the Secretariat’s real financial data for the 2024 fiscal year, covering 18 cost centers (e.g., Fleet, Infrastructure, School Meals), confirming its effectiveness in providing an unprecedented capacity for auditing and financial control for strategic planning
Fortalecendo a Segurança de Software: Estudo de Caso de uma Pipeline com SBOM, SAST e DAST no POP-ERP
(Universidade Federal do Rio Grande do Norte, 2025-12-03) Clemente, Pedro Vitor Bezerra; Falcão, Eduardo de Lucena; http://lattes.cnpq.br/8335184809043358; http://lattes.cnpq.br/2791740260745622; Viegas, Carlos Manuel Dias; http://lattes.cnpq.br/3134700668982522; Rodrigues, Taniro Chacon; http://lattes.cnpq.br/6438516851578298
This work presents the development and implementation of an automated security pi- peline integrated into the Continuous Integration and Continuous Delivery (CI/CD) pro- cess of the PoP-ERP system, aiming to strengthen secure development and mitigate vul- nerabilities throughout the software lifecycle. The solution adopts DevSecOps principles and combines three complementary layers of verification: static application security tes- ting (SAST), software supply chain analysis, and dynamic application security testing (DAST). The methodology includes a systematic literature review, architectural mode- ling, incremental implementation, controlled experimentation, and critical evaluation of the obtained results. The experiments demonstrated the effectiveness of the pipeline in identifying structural vulnerabilities, dependency flaws, and insecure runtime behaviors, highlighting the importance of automation and continuous security integration within the development process. As a contribution, the study provides a reproducible approach ap- plicable to other CI/CD environments, strengthens the security posture of PoP-ERP, and promotes the consolidation of DevSecOps practices within the institution.
MarIA - Um chatbot inteligente para atendimento de vítimas de violência contra a mulher.
(Universidade Federal do Rio Grande do Norte, 2025-12-01) Félix, Tereza Stephanny de Brito; Viegas, Carlos Manuel Dias; http://lattes.cnpq.br/3134700668982522; http://lattes.cnpq.br/3520300459156656; Brito Junior, Agostinho de Medeiros; http://lattes.cnpq.br/0958617290020120; Santos, Clausan Liano Dantas
A violência contra a mulher é um problema estrutural que demanda soluções tecnológicas para facilitar o acesso das vítimas aos serviços de emergência. Este trabalho apresenta o desenvolvimento de um chatbot para o Telegram, utilizando Modelos de Linguagem de Grande Escala (LLM - Large Language Model) e técnicas de Inteligência Artificial (IA), para classificação e atendimento automatizado de ocorrências de violência contra a mulher. A solução foi projetada para permitir que vítimas entrem em contato de maneira discreta e segura, utilizando mensagens de texto, áudio, imagens e localização. O chatbot processa as informações recebidas, classifica a gravidade da ocorrência podendo direcionar os dados para os órgãos competentes, para agilizar o atendimento. Além da implementação do sistema, este trabalho discute os desafios técnicos, éticos e jurídicos envolvidos na aplicação da IA para segurança pública, destacando os impactos sociais da tecnologia no combate à violência de gênero.
Transparência verificável em modelos de IA por meio da auditoria de datasets
(Universidade Federal do Rio Grande do Norte, 2025-12-03) Chagas Filho, Cristian Soares de Souza; Falcão, Eduardo Lucena; Oliveira, Luiz Affonso Henderson Guedes de; Barros, Tiago Tavares Leite
The increasing use of Artificial Intelligence (AI) models in critical applications has highlighted the need for robust mechanisms to ensure the transparency and integrity of the data used in their training. Currently, the lack of verifiability in the machine learning supply chain exposes models to security risks, such as data poisoning, and hinders external auditing. This work presents the development and validation of the hf-merkle tool, a library that integrates Merkle Tree cryptographic structures into the Hugging Face Hub ecosystem. The solution allows for the generation of immutable inclusion proofs for datasets, the automatic publication of these proofs in model repositories, and independent verification via a Command Line Interface (CLI). The experiments conducted demonstrated the technical feasibility of the approach: although the tree generation time and proof file size exhibit linear growth relative to the number of files, the latency for integrity verification proved to be extremely low. Thus, the tool contributes to algorithmic governance by providing a practical and decentralized means to ensure data provenance in AI pipelines.
Um sistema de sensoriamento de uso de sistemas de refrigeração ambiente usando georreferenciamento
(Universidade Federal do Rio Grande do Norte, 2025-12-01) Cabral Filho, Lyndon Jonhson; Brito Junior, Agostinho de Medeiros; Araújo, Fabio Meneghetti Ugulino de; Maitelli, Carla Wilza Souza de Paula
This work presents the development of a management and control system for ambient cooling systems with georeferencing support, aimed at optimizing the use of these devices in large areas and buildings. The problem addressed is the high energy consumption resulting from the unnecessary operation of the equipment, often caused by the lack of real-time monitoring and efficient remote control. The main objective was to implement a web application, consisting of a back-end and front-end, capable of receiving, processing, and displaying data from ambient cooling devices on an interactive map, also enabling remote control. The system was developed in Java, using the Spring Boot framework for the back-end and Thymeleaf for the presentation layer. The interface allows users to view the location and status of each device (on/off, presence or absence of people) through colored markers. The tests carried out demonstrated that the application provides accurate information and enables agile interventions, establishing itself as an effective tool for energy management and contributing to smarter and more sustainable buildings.

SIGAA

Browse

collection.page.browse.recent.head