ABSTRACT
Due to fierce competition in the market, supply chains are becoming increasingly complex. The latest technologies are continuously being evaluated and used by companies to bring value to the end-user, the customer. Efforts to convert factories to smart factories and the digitization of factory and supply chains are currently underway. With the fourth industrial revolution, the focus is on using the Internet of Things (IoT), cloud computing and analytics to take the supply chains to the next level. One such application involves creating a digital twin of the supply chain, which can be used to predict the near-future state of the supply chain.
Once we are able to predict the future state, then any disruptions or risks can be anticipated, and corrective actions can be taken. It is also possible to make the digital twin model suggest the best course of action, by using prescriptive analytics. There are two prerequisites for creating a digital twin of the supply chain. The first is to have a simulation model of the entire supply chain at a sufficient level of detail so as to be useful for the purpose. The second pre-requisite is for the simulation model to be connected to the physical supply chain and to obtain a snapshot of the current system. This can be done through enterprise resource planning, IoT, sensors etc. Once the digital twin is ready and updated with the latest information/state of the physical supply chain, we can run multiple replications of the simulation model to project and see how our current planning decisions will perform. Continuous feedback from the real system and learning from the gap between the simulation and the physical system would add greatly to the precision of the model. The digital twin will also incorporate all the variations and uncertainties present in the supply chain and all the external risks, such as weather, political unrest, etc.
