Model-Driven Development Towards Distributed Intelligent Systems

Abstract

A Distributed Intelligent System (DIS) encompasses a set of intelligent subsystems and components that collaborate to solve problems, reason, understand language and ideas, plan, and learn. In this context, digital twins represent a specific type of DIS, integrating intelligent components to create dynamic virtual replicas of physical entities and provide advanced services. However, despite their benefits, the development of DISs presents significant challenges, primarily due to their inherent complexity, the need for interoperability among heterogeneous elements, and the complexities of process orchestration. To address these issues, this paper proposes a Model-Driven Development (MDD) approach that includes a Domain Specific Language (DSL) and Model-to-Text (M2T) transformations for the design, validation, generation, and deployment of DISs. With the proposed DSL, developers can focus on system design from a high-level abstraction perspective. Then, M2T transformations allow all system code and deployment infrastructure to be automatically generated. This approach significantly reduces the need for labor-intensive manual tasks, reduces the likelihood of human error, and speeds up the development process. The applicability of the proposal is validated using an agricultural digital twin that incorporates a number of intelligent components.