• Total de projetos de pesquisa (7)
  
  1. 2024-Atual. Assessing the multistate flow networks reliability and some applications
     Descrição: Multistate flow networks (MFN) provide a powerful framework for understanding and modeling complex systems with multiple states and transitions. MFNs are used to model a wide rangeof real-world problems in various fields, including epidemiology (disease progression modeling),engineering (reliability analysis), finance (credit risk modeling), business (customer behavioranalysis), and more. The problem of reliability evaluation of MFNs is essential for ensuring thesafety, availability, and efficiency of complex systems, making this problem increasingly attractive in recent decades. It identifies critical components and states within a network, allowingorganizations to implement cost-effective maintenance and repair strategies. Reliable MFNs underpin the continuous operation of essential services and industries such as healthcare, finance,and telecommunications that rely on consistent network performance to provide uninterruptedservices, thereby maintaining economic stability. Several researchers worldwide have been working in this field of research over the last decades. However, the investigation continues as therelated problems are NP-hard. This project aims to introduce exact and approximation techniques and approaches for addressing the network reliability problems outlined in the proposal.It achieves this by employing graph theory and network flow properties as fundamental tools inproblem-solving. In addition to its primary objectives, this research proposal seeks to leveragethe findings derived from the project to enhance existing solutions for evaluating the reliability ofreal-world systems, including power transmission and distribution systems and communicationnetworks. Furthermore, our novel methodologies will be implemented in high-level programminglanguages such as C and Java. These implementations will facilitate comparative analyses withexisting approaches documented in the literature. Additionally, this work desires to lay the basefor the development of potential simulators for evaluating the reliability of MFNs.. Situação: Em andamento; Natureza: Pesquisa. Integrantes: Emilio de Camargo Francesquini - Integrante / FORGHANI-ELAHABAD, MAJID - Coordenador. Financiador(es): Fundação de Amparo à Pesquisa do Estado de São Paulo - Auxílio financeiro.
     Membro: Emilio de Camargo Francesquini.
  2. 2024-Atual. EcoSustain - Data and Computer Science for the Environment
     Descrição: Never before has it been more urgent to restore damaged ecosystems than it is now. The UN Decadeof Ecosystem Restoration aims to prevent, halt, and reverse the degradation of ecosystems on everycontinent and ocean. Such actions can help eradicate poverty, combat climate change, and prevent massextinction. Our current environmental challenges are global and relate to many problems in several areas.They include illegal deforestation, drastic reduction of biodiversity in several biomes, water, groundwaterand air pollution, global warming, high energy footprint, improper waste treatment, use of fossil fuels,abusive use of harmful fertilizers and pesticides, epidemic diseases, and even deaths or total loss ofproperty due to natural disasters like floods, tropical cyclones, hurricanes, typhoons, natural fires, andmany others.Due to the extensive advances in Information and Communication Technology (ICT), Data Science,and Artificial Intelligence in the last two decades, there are multiple opportunities to apply thesetechnologies and knowledge to benefit the environment more broadly. For example, wireless sensornetworks, Internet of Things, flying networks of drones, high-capacity sensors, data filtering andinterpretation, and data analytics can be used to collect and process data from natural resources, animals,or flora to monitor their condition or health, and contribute to automate protection and restorationprocesses. In addition, we can use Data Science and Machine Learning to classify events as regular andabnormal and better predict the result of ecosystem restoration or improvement measures.The EcoSustain project aggregates a group of highly qualified, interdisciplinary researchers fromsome of the best Brazilian universities, building upon their experience with Computer Science, SoftwareEngineering, Simulation, Data Analytics, Machine Learning, IoT, and Environmental Sciences. The aimis to investigate, model, and devise technological solutions for creating software systems, communicationprotocols, networked services, machine learning models, and other artifacts to monitor and analyzeecosystems and natural resources in real-time, as well as to ensure the effective prevention, forecasting,and mitigation of environmental degradation processes caused by humans and their lifestyles.. Situação: Em andamento; Natureza: Pesquisa. Integrantes: Emilio de Camargo Francesquini - Coordenador / Markus Endler - Integrante.
     Membro: Emilio de Camargo Francesquini.
  3. 2024-Atual. CCD-Cidades Carbono Neutro
     Descrição: The CCD - Carbon Neutral Cities was conceived with the ambition of assisting cities in combating climate change, with the vision of proposing solutions aimed at increasing the climate resilience of infrastructure and people, incorporating nature-based technologies into cities, and bringing solutions to the market that operates in cities and contributes to emissions in cities. The CCD Carbon Neutral Cities proposes a model of operation from sectoral perspectives that impact cities and generate carbon emissions, especially carbon dioxide and methane, bringing lines of action through proofs of concepts and pilots that demonstrate viability and envision positive aspects on the path to a low-carbon society. The development of this Center proposal stemmed from the challenges presented by the Secretary of Science, Technology, and Innovation (SCTI) in Smart, Resilient, and Sustainable Cities, resulting in three strategic lines of action, namely: 1. Decarbonization of urban environments, technologies for reducing greenhouse gas emissions (GHG), and climate resilience. 2. Digital transformation and monitoring of services and operations for GHG emissions mitigation and climate resilience in cities. 3. Public policies, training, technology transfer, and enabling networks. These lines of action interact with the sectoral paths identified in the proposal, such as sustainable urban development, road infrastructure and mobility, buildings and construction, renewable energy, and basic sanitation. These sectoral paths were defined through interaction with the market, initially carried out through a public call, inviting companies, as well as governments and non-governmental institutions, to become members of CCD-Carbon Neutral Cities with the purpose of assisting cities in transitioning to carbon neutrality. This proposal has received expressions of interest from nine (09) companies, six (06) non-governmental entities, and seven (07) government institutions from the State of São Paulo; three (03) at the municipal level and two (02) at the federal level, totaling a financial counterpart of R$ 6,200,000.00 over five years, in addition to the economic counterpart. To address the scientific and technological challenges for carbon neutrality in cities, several research and technology institutions (ICTs) in the state of São Paulo are partners, including USP, UNIVESP, UFABC, UNIFESP, USJT, and UNINOVE. Other federal entities include CEMADEN and UFPE. These actors make up the technical and government structure of Carbon Neutral Cities, with an international committee initially chaired, if the proposal is accepted, by the Instituto Superior Técnico de Lisboa - IST, whose professors Paulo Ferrão and Manuel Heitor assisted in the construction of this proposal.. Situação: Em andamento; Natureza: Pesquisa. Integrantes: Emilio de Camargo Francesquini - Coordenador / Liedi Légi Bariani Bernucci - Integrante. Financiador(es): Fundação de Amparo à Pesquisa do Estado de São Paulo - Auxílio financeiro.
     Membro: Emilio de Camargo Francesquini.
  4. 2022-2023. Aplicações de teoria do escalonamento para otimizar o uso de energia verde em plataformas de nuvens computacionais
     Descrição: Cloud computing platforms support most services and applications that we use every day, such associal networks, e-mail, video games, and video streaming, and is a key element for the development ofsmart cities. However, cloud computing platforms consume a massive amount of electricity: the datacenters that host these platforms consume 1 of the of the power generated globally. In order to reducethe costs and environmental impact resulting from this energy consumption, data centers are installingrenewable energy sources such as solar farms in their facilities. The availability of solar energy is notconstant, resulting in challenges in scheduling tasks to reduce nonrenewable energy consumption.Some data centers have batteries that can store renewable energy, but they self-discharge and losetheir capacity as time goes on, which characterizes another challenge: deciding when to store or usebattery power. In this project, we will study how to use scheduling theory to reduce brown energyconsumption and costs for data center operators. In order to tackle the scientific and technologicalchallenges to solve this problem, we will develop a multi-objective scheduling algorithm for virtualmachines that are submitted to geographically distributed cloud computing platforms, consideringthat they have batteries and variable (intermittent) renewable energy supply such as energy fromwind and solar farms. We will also propose a method to estimate the dimensions of the renewablepower source and batteries, given the size of data centers and their expected workloads.. Situação: Concluído; Natureza: Pesquisa. Alunos envolvidos: Graduação: (2) / Mestrado acadêmico: (2) / Doutorado: (1) . Integrantes: Emilio de Camargo Francesquini - Integrante / Daniel de Angelis Cordeiro - Coordenador. Financiador(es): Fundação de Amparo à Pesquisa do Estado de São Paulo - Outra.
     Membro: Emilio de Camargo Francesquini.
  5. 2021-Atual. Tendências em Computação de Alto Desempenho, do Gerenciamento de Recursos a Novas Arquiteturas de Computadores
     Descrição: Nesse projeto vamos juntar esforços e competências de diversos pesquisadores da ciência da computação na área de processamento de alto desempenho (HPC). Nosso principal objetivo é promover pesquisa de alto impcato em duas áreas de HPC: Gerenciamento de Recursos e Novas Arquiteturas de Computadores. Cada uma dessas áreas está dividida em diversos tópicos. Cada tópico será coordenado por um pesquisador experiente e será conduzido com a ajuda de colaboradores e de estudantes. Com esse projeto, queremos fazer não apenas pesquisa, mas também aumentar a sinergia dos pesquisadores da área no estado de São Paulo. Situação: Em andamento; Natureza: Pesquisa. Integrantes: Emilio de Camargo Francesquini - Integrante / Alfredo Goldman vel Lejbman - Coordenador. Financiador(es): Fundação de Amparo à Pesquisa do Estado de São Paulo - Outra.
     Membro: Emilio de Camargo Francesquini.
  6. 2019-2024. INCT da Internet do Futuro para Cidades Inteligentes
     Descrição: Desenvolvimento de modelos, técnicas, métodos e sistemas de software visando à melhoria da vida da população urbana por meio da utilização da tecnologia da informação e da comunicação. Essa pesquisa envolve técnicas avançadas de Computação em Nuvem, Internet das Coisas, Computação de Alto Desempenho, Engenharia de Software, Big Data e Aprendizado de Máquina.. Situação: Concluído; Natureza: Pesquisa. Integrantes: Emilio de Camargo Francesquini - Integrante / KON, FABIO - Coordenador.
     Membro: Emilio de Camargo Francesquini.
  7. 2019-2023. Paralelização de Problemas de Geofísica em Clusters Heterogêneos CPU-CPU na Nuvem
     Descrição: Imageamento s#305;#769;smico e#769; reconhecidamente um problema de Geof#305;#769;sica que e#769; central a#768; etapade explorac#807;a#771;o de petro#769;leo. A soluc#807;a#771;o deste problema se baseia em te#769;cnicas de ReverseTime Migration (RTM) como Common-Reflection-Surface (CRS) e Full Waveform In-version (FWI) que realizam computac#807;o#771;es complexas em volumes gigantescos de dados.Te#769;cnicas como FWI, por exemplo, oferecem soluc#807;o#771;es precisas para o problema de image-amento s#305;#769;smico, a um custo computational muito elevado. Por outro lado, a disponibi-lizac#807;a#771;o pela indu#769;stria de nuvens com no#769;s CPU-GPU tem indicado que problemas comoFWI podem se beneficiar de um nu#769;mero significativo de no#769;s CPU-GPU com elevada ca-pacidade computacional. A meta deste projeto e#769; desenvolver uma plataforma, chamadaBR-CLOUD, baseada em padro#771;es da industria, que permita uma soluc#807;a#771;o eficiente na nu-vem para problemas complexos de imageamento s#305;#769;smico. Para viabilizar esta plataformapretende-se integrar e ampliar modelos de execuc#807;a#771;o paralela ja#769; usados pela indu#769;striacomo OpenMP, MPI e CUDA. O modelo resultante deve ser totalmente compat#305;#769;vel comestes padro#771;es, ser independente de verso#771;es de ferramentas, e permitir o descarregamentode programas cient#305;#769;ficos complexos em nuvens heteroge#770;neas de maneira transparente,balanceada e tolerante a#768; falhas. Descarregamento de computac#807;a#771;o (i.e. computation of-floading) e#769; um modelo de programac#807;a#771;o paralela em que os fragmentos do programa (porexemplo, lac#807;os quentes) sa#771;o anotados para que sua execuc#807;a#771;o seja realizada em hardwarededicado ou dispositivos aceleradores como GPUs. A equipe deste projeto possui umagrande experie#770;ncia em paralelizac#807;a#771;o de co#769;digo, e propo#770;s o primeiro modelo que esten-deu o padra#771;o OpenMP para permitir descarregamento de computac#807;a#771;o transparente nanuvem. Este modelo, chamado OmpCloud, segue o padra#771;o OpenMP, e foi testado comuma aplicac#807;a#771;o cient#305;#769;fica complexa nas nuvens da Amazon AWS e Microsoft Azure, pro-duzindo bons resultados. Este projeto tem dois objetivos. Primeiro pretende-se estenderOmpCloud para viabilizar paralelismo de tarefas de OpenMP 4.X na nuvem, integrandoo modelo resultante de maneira transparente com drivers MPI. Isto permitira#769; a execuc#807;a#771;ona#771;o bloqueante e pipelined das tarefas de fomatac#807;a#771;o e distribuic#807;a#771;o de dados com a com-putac#807;a#771;o em GPU dos kernels dos programas. Espera-se com isto superpor late#770;ncias deexecuc#807;a#771;o temporal e espacialmente de modo a melhorar o desempenho final da aplicac#807;a#771;o.O segundo objetivo deste projeto e#769; desenvolver te#769;cnicas de otimizac#807;a#771;o de co#769;digo paraGPUs que melhorem a localidade de refere#770;ncia e a late#770;ncia de descarregamento de dados.. Situação: Concluído; Natureza: Pesquisa. Integrantes: Emilio de Camargo Francesquini - Integrante / Guido Costa Souza de Araújo - Coordenador.
     Membro: Emilio de Camargo Francesquini.
  
  

Prêmios e títulos