Resumo:
The demand for Edge Computing solutions in Industry 4.0 is increasing, especially in
production environments that require real-time decision-making. However, implementing
Edge Computing in industrial networks based on the Internet of Things (IoT) poses
significant challenges, such as interoperability, low latency, and scalability. To support
software architects in defining architectures for this context, this work proposes a Reference
Architecture (RA) named ED-operator RA. The development of ED-operator RA was
based on a Systematic Literature Review (SLR), through which relevant artifacts were
identified, including non-functional requirements, software solutions, middleware, and other
reference architectures previously proposed in the literature. The proposed architecture
was modeled using multiple architectural views, including context view, container and
module view, deployment view, control flow view, and data flow view. To ensure the
validity of the architecture, this work presents an architectural evaluation based on the
SAAM (Scenario-Based Architecture Analysis Method), supported by an experiment
conducted using a prototype of the ED-operator RA. As a result, the ED-operator RA
consistently demonstrated compliance with the defined requirements, reconciling scalability,
interoperability, security, privacy, and low latency in IoT environments based on Edge
Computing. Furthermore, this study provides a structured and reproducible methodology
for the development and evaluation of reference architectures.
