Rafael Kunst

Mental health issues significantly affect workers and industries by reducing productivity, generating economic losses, and contributing to psycho-physical conditions. Technological advancements are transforming workplaces, emphasizing mental health care through digitalization and automation. This study introduces Kaire, an intelligent model designed to identify and predict work-related stressors in industrial settings. Kaire utilizes an ontology-based framework combined with semantic rules and machine learning algorithms, including Support Vector Machine (SVM) and Random Forest (RF), to classify stressors and forecast events. Synthetic data simulating industrial routines evaluated Kaire’s performance, with SVM achieving 99.36% accuracy, 99.55% precision, 98.84% recall, and 99.19% F1 score. The Group Stress Index (GSI) analysis showed 88.39% similarity between calculated results and simulator outputs. While event prediction scored lower across all evaluated configurations, stressor identification and GSI estimation showed consistent results in the simulated scenario. Kaire contributes to stress identification and monitoring, supporting industries in promoting workers’ mental health and well-being.

Although cognitive neuroscience and neuro-software engineering have advanced, there remains a lack of publicly accessible neural datasets that capture brain activity during complex tasks, such as the merging of refactored source code. This limitation undermines reproducibility and hinders the development of cognitively aware tools for high-cognitive-load scenarios.We introduce BrainMergeEEG, the first open EEG microstate dataset collected during refactored code-merging tasks, a demanding activity in collaborative software development. Our dataset includes recordings from 42 participants who completed resting-state sessions and ran controlled source merge tasks, totaling about 8.0 GB of EEG data. Data are organized by task condition and behavioral performance. To promote transparency and reuse, we provide preprocessing scripts and a fully reproducible experimental pipeline. BrainMergeEEG establishes a neurocognitive benchmark for investigating developers’ behavior during source code merge tasks, supporting future research on brain-aware software engineering. The dataset and related materials are available at https://zenodo.org/records/ 17386294.

Implementing artificial intelligence models on edge devices (EdgeAI) has gained significant popularity due to its potential to enable real-time applications, achieve low latency, and conserve bandwidth. Additionally, reducing dependence on internet connections or cloud infrastructure provides a more secure and reliable execution environment. However, the resource-limited nature of edge devices poses challenges in communication, storage, and computing. Addressing these challenges requires a comprehensive understanding of existing domain optimisation strategies. This survey reviews the current state of the art in EdgeAI optimisation, focusing on communication protocols, storage solutions, and computing architectures that enhance performance and energy efficiency. The contributions of this review are twofold: (i) We highlight key trends, identify gaps in existing research, and propose promising directions for future research to improve the deployment and performance of EdgeAI systems further. (ii) We develop a structured taxonomy that categorises optimisation strategies into computing, storage, communication, and cross-cutting optimisations, offering a clear framework to understand their interrelated approaches and serving as a comparative framework to identify gaps that single-domain surveys often overlook. This survey is a valuable resource for researchers and practitioners seeking to navigate the complex landscape of EdgeAI optimisation and understand the impact of various optimisation pillars and their interactions.

To evaluate the accuracy of neural networks and Naïve Bayes models in diagnosing ectopic pregnancy, using clinical data, hCG levels, and transvaginal ultrasound findings from a real dataset. This was a retrospective cohort study based on a public dataset of 2,495 first-trimester pregnant women with confirmed pregnancy under 13 weeks, documented transvaginal ultrasound reports, and follow-up on pregnancy outcome. The cohort presented a natural imbalance (8.5% ectopic, 91.5% intrauterine pregnancies), reflecting real-world clinical prevalence. Data on risk factors, clinical symptoms, ultrasound findings, and serial hCG levels were included. The dataset was preprocessed and split into training (80%) and testing (20%) sets using stratified sampling based on pregnancy outcome to preserve the proportion of ectopic cases in both sets. The main outcome measures were accuracy, sensitivity, specificity, and F1 score. The neural network model achieved an accuracy of 99.4%, sensitivity of 94.6%, specificity of 97.2%, and an F1 score of 95.9%. The Naïve Bayes model showed an accuracy of 96.5%, sensitivity of 98.1%, specificity of 71.2%, and an F1 score of 82.5%. Both models were validated without evidence of overfitting. The neural network model demonstrated statistically significant superior accuracy and reliability in diagnosing ectopic pregnancy compared to the Naïve Bayes model (McNemar's test, p < 0.001), suggesting the potential of machine learning models, particularly deep learning, to enhance early diagnosis and clinical decision-making.

This Systematic Literature Review (SLR) synthesizes the rapidly evolving state of Agentic AI, analyzing peer-reviewed studies published between 2021 and 2025. We organize the current body of knowledge into a taxonomy comprising five distinct research fields: Memory Cognition, Networking Systems, Trust Safety, Evaluation Limits, and Applications Use-Cases. Our intersection analysis reveals a research landscape currently in an "exploitation" phase, heavily skewed toward Applications Use-Cases (31.9%) and the practical implementation of Autonomy. While the community is successfully operationalizing existing autonomous capabilities, we identify a critical "research desert" regarding Adaptability and New Models. The findings indicate a scarcity of architectural frameworks designed to support long-term resilience and dynamic behavior modification in unforeseen environments. To advance Agentic AI from brittle prototypes to robust real-world systems, we conclude that future research must pivot from solely demonstrating autonomy toward engineering adaptability and developing novel modeling paradigms that prioritize responsiveness to change.

Pests that attack crops are one of the leading causes of low production and economic losses. Identifying and diagnosingforecasts is one way to reduce losses and maintain quality in production. Defoliation of crops harms production andproduction quality. Using computational technologies such as Artificial Intelligence andMachine Learning has enabledthe identification of pests faster and earlier. In this paper, we propose a Deep Learning-based architecture aimed atidentifying and diagnosing insects by analyzing defoliation in crop images. Different Convolutional Neural Networkapproacheswere considered to evaluate the proposed architecture. The training and testing of themodelswere performedusing images collected by drone in a natural environment. The approach that presented the best performance in ourscenario was the VGG16. The average accuracy in the validation phase was 0.95, while in the test set, we obtained 0.86.

Este artigo realizou uma revisão bibliográfica sobre o uso de aprendizado de máquina (ML) no diagnóstico de gravidez ectópica (EP). Foram analisados estudos publicados entre 2020 e 2024, utilizando critérios rigorosos de inclusão e exclusão. Os resultados mostram que técnicas avançadas, como redes neurais convolucionais, algoritmos bayesianos e máquinas de vetores de suporte, apresentam alta precisão diagnóstica, especialmente ao integrar múltiplas fontes de dados, como informações clínicas, laboratoriais e ultrassonográficas. Esses modelos demonstraram potencial para reduzir erros clínicos e melhorar a eficiência diagnóstica, superando limitações dos métodos tradicionais. No entanto, desafios como a falta de validação clínica robusta, a escassez de dados representativos e a complexidade dos modelos, que muitas vezes operam como "caixas pretas", ainda impedem sua implementação prática. Esta revisão contribui para um entendimento mais claro do estado da arte, destacando as abordagens mais eficazes, os principais desafios e as oportunidades para futuras pesquisas. A adoção responsável dessas tecnologias tem potencial para transformar o diagnóstico da EP, melhorando a segurança e os resultados clínicos das pacientes.

Stress is a mental disorder responsible for impacting the industry through psychosomatic illnesses, loss of productivity, and accidents caused by stressful workplaces. Conversely, the literature indicates that fostering mental well-being among workers can boost motivation and performance while alleviating symptoms of stress. The fourth industrial revolution incorporated technologies into work that allowed the automation of processes and control of environments. The fifth revolution introduced the application of research and innovation aimed at a human-centered consciousness, enabling the advancement of mental health through sensors and wearables. Despite advancements in stress classification technologies, there remain opportunities for further research into identifying stress motivators within industrial work environments. In this sense, this paper proposes an ontology to identify stressors considering personal and environmental data, allowing knowledge generation related to work stressors for mitigating the problem. The methodology utilized in this ontology development consisted of seven stages and two evaluation phases. The findings addressed four key questions related to competency as outlined in the model. The results revealed potential stressful scenarios, including the timing of occurrence, shared locations, environmental factors, and identifying groups experiencing moments of stress. This study presents as a scientific contribution the first ontology to address the identification of work-related stressors in the industrial environment.