IIC Journal of Innovation 12th Edition | Page 33

Digital Twin Development for Serial Manipulators: Data Driven Optimized Planning and Sequencing of Tasks Finally, the authors will draw some conclusions and suggest future directions for the research topic. values. Sensor data will then be combined with data from enterprise systems such as Enterprise Resource Planning (ERP), PLM and Manufacturing Execution System (MES). The combined data will then be aggregated accordingly depending on functions of the DT. The bridge between the physical object and its DT is the integration that includes edge, communication interfaces and security. After data is collected, aggregated and transferred to the DT through the integration, analytics techniques such as AI, machine learning and simulation, are executed to monitor, simulate and predict behaviors of the physical object. Subsequently, an application is required to combine all the components and realize the business values. The application can visualize the variations of the physical object, multi-layer data and business related KPIs in real-time. It can also reflect the deviation of the current performance to the expected performance. In addition, it can derive insightful advice to improve the performance. Finally, the application can suggest actions by the analytics to control or interfere the physical object when it is necessary through an actuator mechanism, which is the final element of the DT. D IGITAL T WIN C ONCEPT Before the use case of the DT model of this article is presented in detail, it is worthy to discuss background knowledge about the DT concept. Enders and Hoßbach 15 summarized the essential features of a DT as follows: First, a DT is defined as a virtual model of a physical object, which can be a product, a machine, a process, a factory, a supply chain, etc. Second, the physical object is connected with its twin. Third, thanks to the connection, the DT can reflect current or historical behaviors of the object, simulate and predict its future states, and control it. Finally, these features must create observable business values. In order to explain how a DT can provision enterprises with business values, its fundamental components should be well understood. According to Parrott and Warshaw, a DT is constituted by the following components: sensors, data, integration, analytics, DT application and actuator. These components are depicted in Figure 1. Sensors are installed properly to collect data of the physical object that is required to realize the pre-defined business 15 In the next section, the DT model developed by ARTC and its business values will be presented and explained in detail. Enders, M. R., & Hoßbach, N. (2019). Dimensions of Digital Twin Applications-A Literature Review. - 28 - November 2019