Referenzarchitektur für die Implementierung einer interoperablen Spritzgießfertigung

RWTH Publications (RWTH Aachen)

Abstract

Global markets subject companies in high-wage countries such as Germany to immense cost pressure. One strategy for maintaining competitiveness is the adoption of Industry 4.0, leveraging digital technologies to establish intelligent, interconnected Cyber-Physical Production Systems (CPPS) that operate autonomously and support data-driven decision making. This contribution lays the groundwork for implementing interoperable CPPS in the injection molding domain with a focus on Production Planning and Control (PPC). First, a modular reference architecture is introduced that comprises four distinct layers: The decision layer, which addresses typical production queries; the Digital Shadow layer, which identifies and processes models, services, and data relevant to those queries; the asset layer, where both physical and virtual production resources are organized; and the data layer, which semantically describes the raw data from those assets. An integrated knowledge base ensures seamless data flow across these layers. The central role in decision support is played by the Digital Shadow, derived as a metamodel that issues action recommendations based on relevant models, services, and data. To enable interoperability of models, services, and data, the PPCinIM ontology is designed as a core component of the knowledge base, which formally describes the assets and their relationships in the injection molding domain with a focus on PPC. The PPCinIM ontology is built on established standards and adhering to FAIR principles. In addition, this contribution introduces the concept of a model catalog, whose central element is the model signature. It uniquely characterizes a model by specifying its purpose or required input data. Leveraging the PPCinIM ontology and the model catalog concept, extensive submodels for the Asset Administration Shell are developed for an injection molding machine, an injection mold, a temperature control unit, and a hot runner device. These submodels have been standardized and their elements and metadata integrated into the ECLASS data catalog. Furthermore, submodels that capture the model signature and source systems were developed, enabling interoperable exchange with other entities. The architecture’s implementation is demonstrated through two PPC use cases from injection molding production, showing that standardized Asset Administration Shells and model definitions facilitate efficient decision support via Digital Shadows.

Authors 1

1. Patrick Sapel corresponding Aachen

   RWTH Aachen University

   Affiliation as printed

   RWTH Aachen