A novel method to build knowledge graph models for the configuration and operation design of smart and connected industrial products

Abstract Smart and connected industrial products (SCIPs), characterized by their capabilities of self-monitoring, environment awareness, machine–machine/machine–human communication and collaboration, intelligent decision-making, etc., have become the fundamental elements for cyber-physical systems, digital twin, industrial internet of things, etc. Configuring the components in SCIPs and modeling their interaction and operation mechanisms are important during SCIPs design. However, existing product design methods were originally developed for none smart and connected products. This could limit the accuracy of SCIP modeling during the design stage and consequently, it may cause more reworks during the implementation stage of the designed SCIPs. In this regard, a SCIP configuration and operation design method is established, including (i) meta knowledge graph (KG)-based configuration of the components in the physical system and status monitoring system of a required SCIP, (ii) event-state swimlane flowchart-based analysis of the dynamic interaction, operation, and data monitoring mechanisms among the components, and (iii) event-state KG based modeling of the overall workflow, monitoring data self-updating and intelligent operation mechanisms of the SCIP. Compared with existing SCIP design methods, the work provides a specific method for not only the configuration of the static components in customized SCIPs, but also the dynamic interaction, data acquisition/storing/transmitting, and intelligent function implementation mechanisms of the configured SCIP using a kind of event-state KG. The event-state KG is both human-readable and computer-programmable, and it can self-update according to predefined reasoning algorithms during the operation of the SCIP. The configuration and operation design modeling of a robot-based grinding processing line is used as a case study.