Towards Modular and Plug-and-Produce Manufacturing Apps

摘要

Industry 4.0 redefines manufacturing systems as smart and connected systems where software solutions provide additional capabilities to the manufacturing equipment. However, the connection of manufacturing equipment with software solutions is challenging due to poor interoperability between different original equipment manufacturers (OEMs), making it difficult to integrate into the manufacturing system. Hence, there is a need for a methodology to develop modular "plug-and-produce" applications in the manufacturing domain to meet the requirements of Industry 4.0. This work investigates the "appification" of manufacturing processes where the goal is to sub-divide the process into independent, re-configurable digital manufacturing applications. In this context, "appification" means separating the digital implementation from the physical implementation of the system by making the former modular and independent so that digital implementations can be re-used without depending on the physical parts of the system. In this paper a framework for the development of such manufacturing "apps" is presented. This framework consists of four main elements: a modular plug-and-produce architecture, a manufacturing apps development kit, a communication protocol, and a construction methodology. The modular plug-and-produce architecture is developed using the recent advances in microservices, containerization, and communication technologies. The manufacturing apps development kit (MAPPDK) has been developed to facilitate the implementation of manufacturing apps using high-level programming languages. MAPPDK allows to control manufacturing equipment from external computational devices. The methodology for developing different modules for different types of manufacturing processes is also provided. The proof of concept is shown experimentally by the "appification" of a sorting process using an industrial robot arm, a gripping end-effector, a third-party vision camera, and an intelligent vision module.