A carregar...
Publicação
Sistema de Organização do Conhecimento para Suporte à Circularidade
| Nome: | Descrição: | Tamanho: | Formato: | |
|---|---|---|---|---|
| 6.06 MB | Adobe PDF |
Autores
Orientador(es)
Resumo(s)
A transição para a Economia Circular (EC) requer uma integração mais eficiente, transparente
e interoperavel dos dados industriais ao longo de toda a cadeia de valor. A crescente
Exigência de rastreabilidade e de transparência sobre a origem das matérias-primas,
os métodos de fabrico e o impacto ambiental tem impulsionado o desenvolvimento de infraestruturas
digitais capazes de representar, gerir e partilhar informa¸cão de forma segura,
normalizada e sustentável.
Neste contexto, a presente dissertação propõe e valida uma arquitetura digital baseada no
Asset Administration Shell (AAS), concebida para estruturar, integrar e operacionalizar
informa¸cão relevante para a circularidade industrial. O trabalho foi desenvolvido segundo
a metodologia Design Science Research (DSR) e contempla dois artefactos tecnológicos
complementares. O primeiro e um modelo semântico em LinkML, que permite representar
de forma uniforme ativos, produtos e processos industriais, assegurando coerência estrutural,
interoperabilidade e extensibilidade futura. O segundo e um pipeline de integração
e partilha de dados, implementado sobre o Eclipse BaSyx e o Eclipse Dataspace Connector
(EDC) Extension for AAS, que possibilita a atualização dinâmica dos submodelos do
AAS em tempo real e a partilha soberana de dados através de DataSpaces.
A arquitetura foi validada em ambiente experimental, demonstrando a capacidade de
integrar fluxos de dados heterogéneos, alimentar submodelos como CircularityFeatures,
ProductData e LifeCycleInventory, e calcular métricas de circularidade como o Material
Circularity Indicator (MCI). O AAS revelou-se um elemento estruturante para a cria¸c˜ao
de um Digital Product Passport (DPP), permitindo rastrear e consolidar informa¸c˜ao sobre
o ciclo de vida e a circularidade dos produtos.
Os resultados confirmam o potencial da abordagem proposta para promover a interoperabilidade,
a soberania e a inteligência dos dados industriais, contribuindo para o desenvolvimento
de sistemas cognitivos alinhados com os princípios da Indústria 5.0 (I5.0).
O conceito de Cognitive Digital Twin (CoDT) emerge como uma evolução natural do
AAS, permitindo o enriquecimento e a interpreta¸c˜ao inteligente dos dados para suportar
Decisões mais sustentáveis, humanas e circulares.
São ainda discutidas as limitações associadas a maturidade dos componentes utilizados
(BaSyx DataBridge e EDC) e apontadas direções futuras, como a automatiza¸c˜ao
de políticas de partilha e o fortalecimento da camada cognitiva do AAS.
The transition towards a Circular Economy (CE) requires a more efficient, transparent, and interoperable integration of industrial data across the entire value chain. The growing demand for traceability and transparency regarding the origin of raw materials, manufacturing processes, and environmental impacts has driven the development of digital infrastructures capable of securely representing, managing, and sharing data in a standardized and sustainable way. In this context, this dissertation proposes and validates a digital architecture based on the AAS, designed to structure, integrate, and operationalize information relevant to industrial circularity. The research follows the DSR methodology and includes two complementary technological artefacts. The first is a semantic model developed in LinkML, which enables the uniform representation of assets, products, and industrial processes, ensuring structural consistency, interoperability, and future extensibility. The second is a data integration and sharing pipeline, implemented using Eclipse BaSyx and the EDC Extension for AAS, which allows the real-time updating of AAS submodels and the sovereign sharing of data through DataSpaces. The proposed architecture was validated in an experimental environment, demonstrating its ability to integrate heterogeneous data flows, populate submodels such as Circularity- Features, ProductData, and LifeCycleInventory, and compute circularity metrics such as the MCI. The AAS proved to be a foundational component for the creation of a DPP, enabling the traceability and consolidation of product lifecycle and circularity information. The results confirm the potential of the proposed approach to enhance interoperability, sovereignty, and intelligence in industrial data management, contributing to the development of cognitive and sustainable systems aligned with the principles of I5.0. The CoDT emerges as a natural evolution of the AAS, enabling intelligent data enrichment and interpretation to support more sustainable, human-centered, and circular decision-making. The dissertation also discusses the current limitations of the implemented components (BaSyx DataBridge and EDC), and outlines future research directions, including the automation of data-sharing policies and the strengthening of the AAS’s cognitive layer.
The transition towards a Circular Economy (CE) requires a more efficient, transparent, and interoperable integration of industrial data across the entire value chain. The growing demand for traceability and transparency regarding the origin of raw materials, manufacturing processes, and environmental impacts has driven the development of digital infrastructures capable of securely representing, managing, and sharing data in a standardized and sustainable way. In this context, this dissertation proposes and validates a digital architecture based on the AAS, designed to structure, integrate, and operationalize information relevant to industrial circularity. The research follows the DSR methodology and includes two complementary technological artefacts. The first is a semantic model developed in LinkML, which enables the uniform representation of assets, products, and industrial processes, ensuring structural consistency, interoperability, and future extensibility. The second is a data integration and sharing pipeline, implemented using Eclipse BaSyx and the EDC Extension for AAS, which allows the real-time updating of AAS submodels and the sovereign sharing of data through DataSpaces. The proposed architecture was validated in an experimental environment, demonstrating its ability to integrate heterogeneous data flows, populate submodels such as Circularity- Features, ProductData, and LifeCycleInventory, and compute circularity metrics such as the MCI. The AAS proved to be a foundational component for the creation of a DPP, enabling the traceability and consolidation of product lifecycle and circularity information. The results confirm the potential of the proposed approach to enhance interoperability, sovereignty, and intelligence in industrial data management, contributing to the development of cognitive and sustainable systems aligned with the principles of I5.0. The CoDT emerges as a natural evolution of the AAS, enabling intelligent data enrichment and interpretation to support more sustainable, human-centered, and circular decision-making. The dissertation also discusses the current limitations of the implemented components (BaSyx DataBridge and EDC), and outlines future research directions, including the automation of data-sharing policies and the strengthening of the AAS’s cognitive layer.
Descrição
Palavras-chave
Circularidade Passaporte Digital to Produto Asset Administration Shell Cogntive Digital Twin Industria 5.0