Cloud Computing is revolutionizing smart manufacturing by offering on-demand and scalable computer systems that facilitate plant data analysis and operational efficiency optimization. DevOps is a methodology, widely used for developing Cloud Computing systems, that streamlines software development by improving its integration, delivery, and deployment. Although cloud application designers within a DevOps team are assumed to have development and operational knowledge, this does not fall within the skills of experts that design analytics applications of plant data. The deployment environment is also relevant since, as such applications are often hosted in the Fog, the proliferation of application components may hinder their composition and validation. This work is aimed at embracing the Platform Engineering approach to provide a tailored toolkit that guides the design and development of OpenFog compliant applications for the experts in the Smart Manufacturing domain. The platform uses Model Driven Engineering techniques and a flow-based visual editor to allow application designers to graphically compose applications from components previously delivered by component developers, abstracting them from the underlying technologies. As a result, containerized applications, ready to be deployed and run by a container orchestrator, are obtained. The feasibility of the proposal is proved through an industrial case study.

The paradigm of the fourth industrial revolution poses the combined use of technologies that provide companies with adaptability to face uncertain market scenarios. Different institutions have presented their reference architectures to apply principles of the fourth industrial revolution in a standardized way, being the German proposal Industry 4.0 (I4.0) the most widespread. Nowadays most companies are still in the computerization and digitalization phases prior to being able to connect to I4.0. This situation highlights the need for solutions for the integration of physical assets in I4.0. However, most of the work in this area focuses on ad-hoc solutions, or at least technology-dependent ones. This article presents a multilayer approach for integrating physical assets in I4.0. The goal is to separate the different concerns related to the integration of physical assets into abstract layers. This approach allows addressing these concerns in a generic way, regardless of the type of asset and/or the technology used, as well as efficient, since it addresses each concern separately. This approach is complemented by a methodology that provides guidelines for integrating physical assets according to the multilayer approach. The applicability of the multilayer approach and the integration methodology is put to the test by integrating a roboticarm.

Cyber-Physical Systems (CPS) are devoted to be the main participants in Industry 4.0 (I4.0) solutions. In recent years, many authors have focused their efforts on making proposals for the design and implementation of CPS based on different digital technologies. However, the comparative evaluation of these I4.0 solutions is complex, since there is no uniform criterion when it comes to defining the test scenarios and the metrics to assess them. This paper presents a technology-independent CPS demonstrator for benchmarking I4.0 solutions. To that end, a set of testing scenarios, Key Performance Indicators and services were defined considering the available automation cells setup. The proposed demonstrator has been used to test an I4.0 solution based on a Multi-agent Systems (MAS) approach.

The I4.0 Component is a key concept of the Platform Industrie 4.0 initiative. The I4.0 Component is made up of an asset performing services and an Asset Administration Shell (AAS) representing the asset in the system. I4.0 Components must meet the requirements of interoperability, identification, representation, information management, integration, and asset management. Industrial agents are a suitable approach to implement I4.0 Components since they meet these requirements and provide additional features such as distributed decision-making capabilities. Nevertheless, the development of generic approaches for the integration of AASs and assets is not straightforward due to the diversity of physical assets in the factories. This paper proposes a four-layered architecture for the implementation of I4.0 Components based on industrial agents, two in the agent-based AAS and the other two in the physical asset. The division into layers enhances the separation of concerns, encapsulating the different integration aspects at different abstraction levels. This approach is complementary to the standard IEEE 2660.1-2020 on recommended practices for industrial agents, which focuses on the deployment of the interface between the agent and the physical asset. As a proof of concept, the proposed architecture will be applied for the integration of a software agent and a robot from an assembly cell, aiming to create an I4.0 Component.

With cloud computing gaining momentum in industrial environments, the next step seems to move the computing infrastructure closer to the devices that collect the data at plant level. Fog computing is a paradigm that takes advantage of the fast response times of working relatively close to the plant, and the storage, processing and availability features of the cloud. Fog computing can be used to improve the controllability of automation processes by introducing a higher-level control loop: Fog-in-the-loop (FIL). FIL allows capturing data from the plant, processing it to extract information and feedback actions to the plant based on the processing results. Therefore, FIL applications are context-aware applications that require the deployment of distributed components and dynamic reconfiguration. In this context, we are involved in a project that aims at integrating a model-based, multipurpose Multi-Agent System (MAS) platform with a containerized application orchestrator to meet the requirements of Fog-in-the-loop applications. Specifically, this paper describes a custom scheduler for Kubernetes orchestrator that distributes the scheduling task among the processing nodes by means of the MAS. This new scheduling approach proved to be faster than the centralized scheduling approach used by the default scheduler of K8s.

Adopting the Fog computing paradigm in traditional Cloud infrastructures reduces latency and optimizes network consumption between data providers and data consumers because the system processes the data closer to its source. However, interoperability issues remain as heterogeneous and distributed data sources with different information models must be accessed for performing Cloud-level processing and analysis. This paper describes an integration architecture based on the OPC UA Aggregation Server paradigm, which introduces Fog-level data processing in industrial automation environments. The proposed architecture allows the creation of a configurable aggregated address space at Fog-level where information from data sources at Edge-level is integrated and served to consumers at Cloud-level. The architecture was tested both to verify that the functional requirements are met, and to analyse the time costs involved in the aggregation of Edge-level servers.