Sarita Mazzini Bruschi
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Sarita Mazzini Bruschi has a total of 83 co-authors. Here are the top co-authors by number of shared publications:
- Julio Cezar Estrella
Associate Professor at Universidade de São Paulo, São Carlos, Brazil
6 shared publications - Marcos José Santana
University of Sao Paolo
5 shared publications - Regina Helena Carlucci Santana
University of Sao Paolo
5 shared publications Marcos J. Santana
Universidade de São Paulo, Instituto de Ciências Matemáticas e de Computação, Sao Carlos - SP, Brasil
4 shared publicationsRegina Helena Carlucci Santana
Institute of Mathematical and Computer Sciences, University of São Paulo, São Carlos, Brazil
4 shared publications- Daniel Cordeiro
Universidade de São Paulo
3 shared publications
Latest Publications
- 25 May 2026
- Published by Sociedade Brasileira de Computação
- p. 43-56
Avaliação Multicritério de Detectores Leves de Pessoas para Edge Computing em SBCs: Desempenho, Latência e Eficiência Energética
Este trabalho apresenta uma avaliação comparativa de detectores leves de pessoas para computação de borda nas plataformas Jetson Nano e Raspberry Pi 4B. Foram avaliados os modelos YOLOv3-tiny, YOLOv5n, YOLOv7tiny, YOLOv8n e SSDLite320 MobileNetV3-Large sob um protocolo padronizado, utilizando ONNX Runtime em CPU. A análise considerou desempenho preditivo, latência, FPS, uso de memória, temperatura e consumo energético. Os resultados dos experimentos indicam que os modelos da família YOLO obtiveram melhor desempenho preditivo geral, enquanto o YOLOv5n apresentou o melhor equilíbrio entre qualidade de detecção, latência e custo energético. Além disso, a Raspberry Pi 4B mostrou maior eficiência energética em todos os modelos avaliados.
- 25 May 2026
- Published by Sociedade Brasileira de Computação
- p. 43-56
Avaliação Multicritério de Detectores Leves de Pessoas para Edge Computing em SBCs: Desempenho, Latência e Eficiência Energética
Este trabalho apresenta uma avaliação comparativa de detectores leves de pessoas para computação de borda nas plataformas Jetson Nano e Raspberry Pi 4B. Foram avaliados os modelos YOLOv3-tiny, YOLOv5n, YOLOv7tiny, YOLOv8n e SSDLite320 MobileNetV3-Large sob um protocolo padronizado, utilizando ONNX Runtime em CPU. A análise considerou desempenho preditivo, latência, FPS, uso de memória, temperatura e consumo energético. Os resultados dos experimentos indicam que os modelos da família YOLO obtiveram melhor desempenho preditivo geral, enquanto o YOLOv5n apresentou o melhor equilíbrio entre qualidade de detecção, latência e custo energético. Além disso, a Raspberry Pi 4B mostrou maior eficiência energética em todos os modelos avaliados.
- 25 May 2026
- Published by Sociedade Brasileira de Computação
- p. 155-168
Modelagem e Simulação Comparativa entre Semáforo Inteligente e Tradicional para Detecção de Veículos de Urgência em Cruzamento Urbano
Veículos de urgência enfrentam dificuldades na prestação de socorro em vias urbanas. Esta complexidade, exige soluções que possam melhorar e otimizar o tráfego para travessias de cruzamentos durante o resgate. Os recursos dos Sistemas de Transporte Inteligente integrados com tecnologias da Internet das Coisas, embarcados aos semáforos, estabelecem um Semáforo Inteligente (Smart Traffic Light STL). Soluções STL contribuem em monitorar e gerenciar cruzamentos, liberando a passagem de veículos de urgência. Dez minutos de atraso no atendimento pré-hospitalar aumentam a chance de óbito em 9%. No Brasil em 44.822 ocorrências do SAMU-192, apenas 11% dos atendimentos ocorreram em menos de cinco minutos, enquanto 62% ultrapassaram 10 minutos. Diante desse problema, este trabalho realizou uma modelagem analítica e análise comparativa por simulação (SUMO) entre semáforo tradicional e semáforo inteligente para priorização da travessia de veículos de urgência em cruzamento urbano realístico. Também foi implementada uma maquete experimental, para reproduzir fisicamente o cenário investigado. Os resultados demonstraram ganhos consistentes para os veículos de urgência, reduzindo o tempo total no atendimento para 5,7 seg e com mínimos impactos (trade-offs) para os veículos comuns do cruzamento.
- 25 May 2026
- Published by Sociedade Brasileira de Computação
- p. 155-168
Modelagem e Simulação Comparativa entre Semáforo Inteligente e Tradicional para Detecção de Veículos de Urgência em Cruzamento Urbano
Veículos de urgência enfrentam dificuldades na prestação de socorro em vias urbanas. Esta complexidade, exige soluções que possam melhorar e otimizar o tráfego para travessias de cruzamentos durante o resgate. Os recursos dos Sistemas de Transporte Inteligente integrados com tecnologias da Internet das Coisas, embarcados aos semáforos, estabelecem um Semáforo Inteligente (Smart Traffic Light STL). Soluções STL contribuem em monitorar e gerenciar cruzamentos, liberando a passagem de veículos de urgência. Dez minutos de atraso no atendimento pré-hospitalar aumentam a chance de óbito em 9%. No Brasil em 44.822 ocorrências do SAMU-192, apenas 11% dos atendimentos ocorreram em menos de cinco minutos, enquanto 62% ultrapassaram 10 minutos. Diante desse problema, este trabalho realizou uma modelagem analítica e análise comparativa por simulação (SUMO) entre semáforo tradicional e semáforo inteligente para priorização da travessia de veículos de urgência em cruzamento urbano realístico. Também foi implementada uma maquete experimental, para reproduzir fisicamente o cenário investigado. Os resultados demonstraram ganhos consistentes para os veículos de urgência, reduzindo o tempo total no atendimento para 5,7 seg e com mínimos impactos (trade-offs) para os veículos comuns do cruzamento.
- 26 November 2025
- Published by Sociedade Brasileira de Computação
- p. 21-24
BTRFS vs Ext4: Filesystems Impact on Energy Consumption of eMMC-Based Embedded Computers
Advancements in microelectronics have expanded embedded systems’ capacities, with a significant number now incorporating processing units supporting Linux-based distributions. In energy-constrained environments, selecting an energyefficient filesystem for embedded devices is imperative, yet many performance evaluations focus solely on their time efficiency. To overcome these limitations, this paper proposes a performance evaluation on eMMC-based embedded computers to understand the impact of different filesystems on their energy consumption. We generated a basic Linux image and copied it to the eMMC storage twice. For each copy, we formatted the root filesystem partition with different filesystems (EXT4 and BTRFS) and conducted file operation benchmarks. Measurements of mean power during each benchmark execution were performed by an external microcontrolled device. From the results, it can be inferred that EXT4 is the best overall option for power saving when executing file operations.
- 26 November 2025
- Published by Sociedade Brasileira de Computação
- p. 21-24
BTRFS vs Ext4: Filesystems Impact on Energy Consumption of eMMC-Based Embedded Computers
Advancements in microelectronics have expanded embedded systems’ capacities, with a significant number now incorporating processing units supporting Linux-based distributions. In energy-constrained environments, selecting an energyefficient filesystem for embedded devices is imperative, yet many performance evaluations focus solely on their time efficiency. To overcome these limitations, this paper proposes a performance evaluation on eMMC-based embedded computers to understand the impact of different filesystems on their energy consumption. We generated a basic Linux image and copied it to the eMMC storage twice. For each copy, we formatted the root filesystem partition with different filesystems (EXT4 and BTRFS) and conducted file operation benchmarks. Measurements of mean power during each benchmark execution were performed by an external microcontrolled device. From the results, it can be inferred that EXT4 is the best overall option for power saving when executing file operations.
- Open Access
- 10 November 2025
- Published by Sociedade Brasileira de Computação
- p. 189-194
Dev.Learn - Relato de experiência de um curso de extensão para ensinar a desenvolver sites
This paper reports on a community outreach course focused on web development, conducted by the USPCodelab Sanca group. With a mission to foster technological innovation at USP, this initiative aimed to empower high school students from local schools by promoting their digital inclusion. The course featured weekly theoretical and practical classes with one-on-one mentorship, using essential technologies such as HTML, CSS, and introductory JavaScript. This experience demonstrated the potential of university outreach to create a significant social impact, aligning directly with the UN Sustainable Development Goal 4 (Quality Education) and Goal 8 (Promote sustained economic growth). The course not only increased participants’ interest in technology, but also effectively transitioned them from passive technology consumers to active creators, achieving meaningful digital inclusion.
- 10 November 2025
- Published by Sociedade Brasileira de Computação
- p. 189-194
Open AccessDev.Learn - Relato de experiência de um curso de extensão para ensinar a desenvolver sites
This paper reports on a community outreach course focused on web development, conducted by the USPCodelab Sanca group. With a mission to foster technological innovation at USP, this initiative aimed to empower high school students from local schools by promoting their digital inclusion. The course featured weekly theoretical and practical classes with one-on-one mentorship, using essential technologies such as HTML, CSS, and introductory JavaScript. This experience demonstrated the potential of university outreach to create a significant social impact, aligning directly with the UN Sustainable Development Goal 4 (Quality Education) and Goal 8 (Promote sustained economic growth). The course not only increased participants’ interest in technology, but also effectively transitioned them from passive technology consumers to active creators, achieving meaningful digital inclusion.
- Open Access
- 28 October 2025
- Published by Sociedade Brasileira de Computação
- p. 109-120
Estimating CO2 emissions of distributed applications and platforms with SimGrid_Batsim
This work presents a carbon footprint plugin designed to extend the capabilities of the Batsim simulator by allowing the calculation of CO2 emissions during simulation runs. The goal is to assess the environmental impact associated with task and resource management strategies in simulated environments. The plugin is developed within SimGrid—the underlying simulation framework of Batsim—and computes carbon emissions based on the simulated platform’s energy consumption and carbon intensity factor of the simulated machines. Once implemented, it is integrated into Batsim, ensuring compatibility with existing simulation workflows and enabling researchers to assess the carbon efficiency of their scheduling strategies.
- 28 October 2025
- Published by Sociedade Brasileira de Computação
- p. 109-120
Open AccessEstimating CO2 emissions of distributed applications and platforms with SimGrid_Batsim
This work presents a carbon footprint plugin designed to extend the capabilities of the Batsim simulator by allowing the calculation of CO2 emissions during simulation runs. The goal is to assess the environmental impact associated with task and resource management strategies in simulated environments. The plugin is developed within SimGrid—the underlying simulation framework of Batsim—and computes carbon emissions based on the simulated platform’s energy consumption and carbon intensity factor of the simulated machines. Once implemented, it is integrated into Batsim, ensuring compatibility with existing simulation workflows and enabling researchers to assess the carbon efficiency of their scheduling strategies.
- Open Access
Estimating CO$_2$ emissions of distributed applications and platforms with SimGrid/Batsim
This work presents a carbon footprint plugin designed to extend the capabilities of the Batsim simulator by allowing the calculation of CO$_2$ emissions during simulation runs. The goal is to comprehensively assess the environmental impact associated with task and resource management strategies in data centers. The plugin is developed within SimGrid -- the underlying simulation framework of Batsim -- and computes carbon emissions based on the simulated platform's energy consumption and carbon intensity factor of the simulated machines. Once implemented, it is integrated into Batsim, ensuring compatibility with existing simulation workflows and enabling researchers to assess the carbon efficiency of their scheduling strategies.
Open AccessEstimating CO$_2$ emissions of distributed applications and platforms with SimGrid/Batsim
This work presents a carbon footprint plugin designed to extend the capabilities of the Batsim simulator by allowing the calculation of CO$_2$ emissions during simulation runs. The goal is to comprehensively assess the environmental impact associated with task and resource management strategies in data centers. The plugin is developed within SimGrid -- the underlying simulation framework of Batsim -- and computes carbon emissions based on the simulated platform's energy consumption and carbon intensity factor of the simulated machines. Once implemented, it is integrated into Batsim, ensuring compatibility with existing simulation workflows and enabling researchers to assess the carbon efficiency of their scheduling strategies.
- Open Access
- 28 May 2025
- Published by Sociedade Brasileira de Computação
- p. 74-77
Simuladores para Computação em Nuvem: um estudo de disponibilidade
O uso da computação em nuvem tem crescido continuamente entre as empresas, impulsionado por seus inúmeros benefícios. Para validar pesquisas nessa área, os simuladores de computação em nuvem desempenham um papel fundamental. Ao longo dos anos, o avanço das pesquisas levou ao desenvolvimento de diversos simuladores, resultando em uma ampla gama de opções e milhares de artigos científicos sobre o tema. No entanto, muitos desses simuladores já não estão mais disponíveis. Neste trabalho, foi realizado um levantamento dos simuladores existentes e uma análise de sua disponibilidade, com o objetivo de fornecer informações úteis para pesquisadores da área.
- 28 May 2025
- Published by Sociedade Brasileira de Computação
- p. 74-77
Open AccessSimuladores para Computação em Nuvem: um estudo de disponibilidade
O uso da computação em nuvem tem crescido continuamente entre as empresas, impulsionado por seus inúmeros benefícios. Para validar pesquisas nessa área, os simuladores de computação em nuvem desempenham um papel fundamental. Ao longo dos anos, o avanço das pesquisas levou ao desenvolvimento de diversos simuladores, resultando em uma ampla gama de opções e milhares de artigos científicos sobre o tema. No entanto, muitos desses simuladores já não estão mais disponíveis. Neste trabalho, foi realizado um levantamento dos simuladores existentes e uma análise de sua disponibilidade, com o objetivo de fornecer informações úteis para pesquisadores da área.
- Open Access
- 29 April 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 13, 77550-77564
Anomaly Detection and Root Cause Analysis in Cloud-Native Environments Using Large Language Models and Bayesian Networks
Cloud computing technologies offer significant advantages in scalability and performance, enabling rapid deployment of applications. The adoption of microservices-oriented architectures has introduced an ecosystem characterized by an increased number of applications, frameworks, abstraction layers, orchestrators, and hypervisors, all operating within distributed systems. This complexity results in the generation of vast quantities of logs from diverse sources, making the analysis of these events an inherently challenging task, particularly in the absence of automation. To address this issue, Machine Learning techniques leveraging Large Language Models (LLMs) offer a promising approach for dynamically identifying patterns within these events. In this study, we propose a novel anomaly detection framework utilizing a microservices architecture deployed on Kubernetes and Istio, enhanced by an LLM model. The model was trained on various error scenarios, with Chaos Mesh employed as an error injection tool to simulate faults of different natures, and Locust used as a load generator to create workload stress conditions. After an anomaly is detected by the LLM model, we employ a dynamic Bayesian network to provide probabilistic inferences about the incident, proving the relationships between components and assessing the degree of impact among them. Additionally, a ChatBot powered by the same LLM model allows users to interact with the AI, ask questions about the detected incident, and gain deeper insights. The experimental results demonstrated the model’s effectiveness, reliably identifying all error events across various test scenarios. While it successfully avoided missing any anomalies, it did produce some false positives, which remain within acceptable limits.
- 29 April 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 13, 77550-77564
Open AccessAnomaly Detection and Root Cause Analysis in Cloud-Native Environments Using Large Language Models and Bayesian Networks
Cloud computing technologies offer significant advantages in scalability and performance, enabling rapid deployment of applications. The adoption of microservices-oriented architectures has introduced an ecosystem characterized by an increased number of applications, frameworks, abstraction layers, orchestrators, and hypervisors, all operating within distributed systems. This complexity results in the generation of vast quantities of logs from diverse sources, making the analysis of these events an inherently challenging task, particularly in the absence of automation. To address this issue, Machine Learning techniques leveraging Large Language Models (LLMs) offer a promising approach for dynamically identifying patterns within these events. In this study, we propose a novel anomaly detection framework utilizing a microservices architecture deployed on Kubernetes and Istio, enhanced by an LLM model. The model was trained on various error scenarios, with Chaos Mesh employed as an error injection tool to simulate faults of different natures, and Locust used as a load generator to create workload stress conditions. After an anomaly is detected by the LLM model, we employ a dynamic Bayesian network to provide probabilistic inferences about the incident, proving the relationships between components and assessing the degree of impact among them. Additionally, a ChatBot powered by the same LLM model allows users to interact with the AI, ask questions about the detected incident, and gain deeper insights. The experimental results demonstrated the model’s effectiveness, reliably identifying all error events across various test scenarios. While it successfully avoided missing any anomalies, it did produce some false positives, which remain within acceptable limits.
- 1 March 2025
- Published by World Scientific Pub Co Pte Ltd in Journal of Hyperbolic Differential Equations
- Vol. 22 (01) , 111-139
Dynamical properties of a second-order in time partial integrodifferential equation: Existence and regularity of pullback attractors
In this paper, we consider a second-order in time partial integro-differential equation under suitable integrability conditions. We show existence of pullback attractors in functions spaces similar to fractional Sobolev spaces with an evolution process approach.
- 1 March 2025
- Published by World Scientific Pub Co Pte Ltd in Journal of Hyperbolic Differential Equations
- Vol. 22 (01) , 111-139
Dynamical properties of a second-order in time partial integrodifferential equation: Existence and regularity of pullback attractors
In this paper, we consider a second-order in time partial integro-differential equation under suitable integrability conditions. We show existence of pullback attractors in functions spaces similar to fractional Sobolev spaces with an evolution process approach.
- Open Access
Estimating CO2 emissions of distributed applications and platforms with SimGrid/Batsim
This work presents a carbon footprint plugin designed to extend the capabilities of the Batsim simulator by allowing the calculation of CO2 emissions during simulation runs. The goal is to comprehensively assess the environmental impact associated with task and resource management strategies in data centers. The plugin is developed within SimGrid---the underlying simulation framework of Batsim---and computes carbon emissions based on the simulated platform's energy consumption and carbon intensity factor of the simulated machines. Once implemented, it is integrated into Batsim, ensuring compatibility with existing simulation workflows and enabling researchers to assess the carbon efficiency of their scheduling strategies.
Open AccessEstimating CO2 emissions of distributed applications and platforms with SimGrid/Batsim
This work presents a carbon footprint plugin designed to extend the capabilities of the Batsim simulator by allowing the calculation of CO2 emissions during simulation runs. The goal is to comprehensively assess the environmental impact associated with task and resource management strategies in data centers. The plugin is developed within SimGrid---the underlying simulation framework of Batsim---and computes carbon emissions based on the simulated platform's energy consumption and carbon intensity factor of the simulated machines. Once implemented, it is integrated into Batsim, ensuring compatibility with existing simulation workflows and enabling researchers to assess the carbon efficiency of their scheduling strategies.