Nowadays, emerging technologies such as Big Data, Internet of Things, smart sensor networks, virtualization and shared pools of configurable computer resources, among others, are starting to be adopted and adapted in industrial environments. This fact is leading to an exponential growth of data that can be transformed into valuable information to be used to impact the performance, quality and controllability of automation processes. The digitization of industry is the main goal of the industry 4.0 concept. It deals with the introduction of new methodologies and technologies that allow to factories to become Smart Factories by means of advanced capabilities, based on a digital infrastructure that provides a distributed intelligence to manage distributed resources and maintain a certain quality of service. In this context, recent research in the manufacturing domain as well as in other domains such as e-Health, surveillance and natural disaster prevention, recent research has successfully integrated Multi-Agent Systems (MAS) and Model-Driven Engineering (MDE) to develop a distributed intelligence that allows industries and institutions to make tactical decisions in real time and achieve a certain level of adaptation. However, MAS and MDE alone do not provide the necessary foundation for building a digital infrastructure that allows capturing data from the plant, processing massive data to extract information and feedback actions to the plant to achieve the requirements of the factory of the future. Recent research trends point to the integration of the aforementioned technologies (MAS and MDE) with blockchain technology (BCT) and Fog-computing, with the expectation of providing distributed storage and cloud-level features, respectively, in order to decentralize the decision making when a soft change occurs (those that cause delays but do not need global restructuring)  or to allow a bigger involvement from a decision level above the production level (i.e., factory level).

In this sense, this project proposal consists of improving efficiency, quality and traceability of production systems by means of introducing a high layer of control through the fog. The objective is to create a generic platform (an architecture, a methodology and support tools) for the definition, development and execution of user-defined Fog-in-the-loop applications.

The general goal of the project is to improve the efficiency, quality and traceability of production systems by means of introducing a high layer of control through the fog. In particular, the project will focus on the design and development of a complete architecture able to execute user-defined applications in terms of plant data acquisition, filtering, processing, establishing methods and developing tools for enabling fog-in-the-loop applications.

To achieve the general goal, the project targets a set of specific goals:

• Distributed intelligence towards Product Oriented Manufacturing. The goal here is to extend and generalize the MAS architecture result of the previous project and establish a methodology for customizing it at different levels (such as field, plant and fog). In this specific goal the generic architecture will be validated for the case of flexible POM (Product Oriented Manufacturing). The target is to achieve product-centric manufacturing by means of an agile production reconfiguration in response to demand changes, lot size-one demand or sudden resource failures.
• Massive plant data acquisition, storage, filtering and processing. In this specific goal container technology will be explored to support industrial communications, ubiquitous processing and technology independence. And distributed technologies, such as blockchain will be explored for supporting distributed storage. The target is to give support to applications functionality.
• Definition of process state aware applications (acquire-filter-process-feedback-to-plant). The use of model-based techniques will be explored to achieve different goals: separation of concerns for defining applications from the user point of view and with enough expressive power to specify reactive context-aware applications, allowing the feedback to the plant.
• Management of pro-active context-aware applications. This specific goal is devoted to analyze existing cloud container management and adapting them to meet the requirements of fog-in-the-loop applications taking into account aspects such as availability, data integrity, agile dynamic deployment, and low latencies. Customization of the generic MAS platform will be integrated with the selected cloud orchestrator to meet the requirements.
• Design of customizable platform and development tools. This specific goal integrates the previous 2 to obtain a model-based intelligent architecture and a tool suite supporting application definition, fog application generation and management.

• To introduce distributed intelligence in product-oriented manufacturing (POM) systems
• To introduce the fog as a higher layer of control giving birth to what we called “Fog-in-the-loop” applications.

Spanish Ministry of Science, Innovation and Universities; European Regional Development Fund (FEDER), European Union