 Dear participants, welcome to the course on supply chain digitization. It is jointly being taught by Professor Priyanka Burma, Professor Sushmita Narayana and Professor Devabrata Das from IM Mumbai. So, in today's session, we are going to talk about supply chain digital twin. We will give the introduction of supply chain digital twin, what it is and how does it work and then in subsequent lectures, we will also try to develop a digital twin and show how it works. So, what is digital twin? So, a digital twin is a virtual representation of a physical object system or process. It is a digital counter part that mimics the real world entity in term of its attributes, behavior and interaction. So, one of the important part here is a digital counter part that mimics the real world entity. Digital twins are typically created using data from sensors, IoT devices and other sources along with advanced modeling and simulation technique. So, towards the end of this lecture, we will see how advanced AI ML models are being used to develop digital twin. We will also see how we get data from sensors and IoT devices and how this data can be integrated into a digital twin. So, in this course, specifically we will focus on digital twin from supply chain point of view, specifically supply chain and operations management point of view. So, before we go ahead, we need to understand what are the different types of digital twins exist in supply chain and operations management. So, in the first one which is product level digital twin, for example, the digital twin of aircraft engine helps us to monitor the performance of aircraft engines in real time. Then we have process level digital twins. For example, the digital twin of refining process in oil and gas industry helps the decision maker to optimize process parameters, predict equipment failure and improve safety. Then we have company level digital twin. You can also say enterprise level digital twin. For example, the digital twin of a manufacturing company helps the manager to optimize production schedules, monitor equipment health and improve overall efficiency of the organization. Then we have supply chain network level digital twin. For example, the digital twin of a food and beverage supply chain will help the decision makers to optimize sourcing, manufacturing and distribution process to ensure product quality, traceability and compliance with regulatory requirement. Then we have network of networks level digital twin. So, what is network of network? It represents an interconnected system of multiple supply chain, each with its own set of suppliers, manufacturers, distributors and customers. So, we have seen that there are various types of digital twins which exist in supply chain operation management, product, process, company, supply chain network and network of networks. So, we have developed a pyramid structure. There is a logic behind it. So, if you see the complexity level of the digital twin, at product level it is the lowest. And then once we move from product to process to company to supply chain network, supply chain network to network of networks, the complexity of the digital twin increases. And we will have highest level of complexity if we develop a digital twin of network of networks. So, therefore, this slide gives you the summary of various types of digital twin exist in supply chain operation management and what is the complexity level of each type. Now, in the digital twin data is an important part. So, without data we cannot develop a digital twin. So, we need to get this data from the physical system and put this into a digital system. So, to get those data we need different IT systems. In the product level information we get from CAD and CAM system. CAD is computer added design, CAM is computer added manufacturing. Then we get process level data. So, let us take the example of manufacturing process. So, we get manufacturing level data from MES that is manufacturing execution system. Then we have company level digital twin for that we need data from an organization from an enterprise. So, for that we use enterprise resource planning that is ERP. Then we need data to design supply chain network and to develop a digital twin for that. So, there are some special software to do that. For example, any logistic. So, it is one such software which can be coupled with external data sources such as logistics service providers, weather data, financial market data and we can develop a digital twin for supply chain network. So, these are various sources of data and few examples of different IT systems which can be used to generate and process data and connect physical system with the digital system. Now, there are few digital technologies because in digital twin we have to connect the physical system with the digital system and digital system is a replica of the physical system. So, I need to get the systems connected. So, IoT sensors is one such tool which helps us to connect physical entity with the digital entity. So, it will get data from the physical entity and these sensors will pass the information to the digital entity. Then we have blockchain which is one such technology which helps us to connect and trace different parts, component and products across the supply chain. Then we have AI and ML algorithm. This is very interesting because in digital twin it has a decision support system. So, I need to develop artificial intelligence, machine learning, optimization algorithms to optimize the overall system so that performance is improved. Then we have cloud. Since we will be having lots of data sources, we need to store them, we need to process them and also we have to apply AI, ML and advance optimization algorithm to do such kind of computing. We need a good cloud facility. So, let us see a sample of a supply chain digital twin how it looks like. So, you can see this diagram. I have two factories located in Orangabad and Nasik. These two factories are sending finished goods to V1D and VAP distribution centers then from these two distribution centers the products are being transported to Mumbai, Pune, Surat and Ahmedabad to satisfy the customer's demand. So, this is a physical supply chain. I can see the movement of products from factory to DC. I can see the movement of product from DC to various customer's locations. So, now if I create a digital visualization of this physical supply chain then it may look like this. So, what we are doing? We are putting each of these points that is factories, distribution centers and customers on the map. So, if you see the right hand side I have a factory located at Orangabad. I have another factory located in Nasik. So, what I am doing it over here? I am putting the exact latitude and longitude of the Orangabad factory and putting it on the map. So, I know where the factory is located. I can visualize it in a better way. Similarly, I can visualize where my Nasik factory is located exactly what is the latitude and longitude of the Nasik factory. So, it gives us a better visibility in terms of their physical location on the map. Then I can also locate the distribution centers. The first distribution center located is in Bapi. Second one is located in V1D. So, I know where exactly they are located on the map. Then also know the location of my customers. One is located in Ahmedabad. Second one is located in Surat. Third one is in Mumbai. And the fourth one is in Pune. So, the one of the advantages of digital visualization of physical supply chain is it gives me the complete feasibility of the physical supply chain. I know where exactly my factory is located. I know where exactly my distribution centers are located. I know where exactly my customers are located. So, it gives me much more visibility and clarity in terms of their location. Now, this is one. The second thing is I also need to know the movement of products from one factory to another DC from DC to customer locations and so on. So, therefore, if you look into this visualization, I can see from factory Orangabad, my finished goods will move through this route to DC Bapi. From factory Orangabad, my finished goods will move to this V1D distribution center using this route. Similarly, I can see from Nasik factory, the finished goods can be transported to VP using this route. From Nasik factory, the finished goods can be transported to V1D using this route. Then from these two distribution centers Vapi and V1D, I can also see how it can be connected to the customers. You can also see how it can be connected to the customers. So, I am not only able to see their exact location on the map, I am also able to see through which route the movement of products will happen. So, this is exact map of the physical supply chain and I can also see their routes. So, this is one of the advantages of digital visualization of a physical supply chain. Now, the question is what is supply chain digital twin? It is just a visualization like of course, it is giving a better visualization in terms of the location, in terms of the movement of products from one location to another location, which specific route they are taking that also it is giving, but is it the digital twin or something else. So, of course, it is much beyond that the visualization. So, what is supply chain digital twin? It is a virtual system comprised of the following. The first one is digital visualization which we saw in the last slide, digital visualization of a physical supply chain and its element. Then we have digital technologies, this provides information about physical supply chain and its element. We have discussed briefly, there are some technologies, let us say IOT sensors and then block chains, then ERP cloud. So, this helps us to connect physical supply chain with the digital supply chain and it generates data from the physical supply chain and connects it with the digital supply chain. So, these technologies are enabler for my digital twin. Then in the right hand side, you can see that digital twin is not only digital visualization and digital technology, it is much beyond that it has decision making support also. So, it is not only giving you the better visualization, it is not only connecting physical system with a digital system, it also gives you better decision making support. How it gives? It has descriptive analytics capability as well as predictive and also prescriptive analytics capability. So, let us talk about descriptive analytics. That scripted analytics means what has happened in the past. So, using digital twin, I can find out what has happened in my overall supply chain. If I have let us say three products A, B and C, I can find out what was the sale of product A in the last two, three years. I can also find out at which factory, how much was my utilization and so on. So, let us say this is my sales data of product A, B and C in the last three years. So, I can see that product C, the sales was high in the last three years. So, I am curious to know whether the sales of product C will remain high in upcoming years or not. So, for that I can use digital twin specifically the predictive analytics concept. So, it will tell me what will happen in the future. So, one of the example is demand forecasting. Suppose using descriptive analytics I found out that product C's demand was high in the last three, four years. I want to find out whether it will continue to remain high or not. So, for that I can use demand forecasting techniques. I can use specifically the predictive analytics model to find out whether for product C it will keep on going up or not or it will start going down. So, if it goes up then I have to find out whether I have enough capacity to produce that many products. So, for that I need the help of prescriptive analytics. It will tell me what should be done in a specific scenario. If I get to know using predictive analytics that my demand of product C will keep on going up in the next two, three years then I need to find out where do I get the products from. Do I have enough capacity or not? If I do not have enough capacity in my factory should I increase the capacity of the factory or should I have another factory in another location? If I decide to open another factory in a different location what should be the location? So, that my overall cost is minimized both fixed cost as well as available cost. So, similarly let us say I found out that demand of product C will increase in the next 3, 4 years. Can I satisfy the demand using the existing distribution centers? I have two distribution centers let us say one is VAPI another one is VVANDI is these two distribution centers enough for me? Do I need more DC? If I decide to open another DC then where should I open the DC? What should be the capacity of the DC? That DC would be supplying products to which customers? All of this can be answered using prescriptive analytics module. So, now let us see like what are the advantages of supply chain digital twin? First of all it gives end to end visibility to the managers. So, visibility is very important and we have seen that digital twin is a replica of a physical supply chain. So, therefore it will give me lot of feasibility. I know where my factory is exactly located, I know where my DCs are exactly located, I know where my customers are exactly located. I also know how the movement of products are happening from factory to DC from DC to customers and so on. So, all of this visibility I will get if I have a digital twin. Then manager can easily trace the products, parts, component, raw material along the supply chain. So, traceability is an important component of digital twin. So, for example, suppose you found out that I have a product which is faulty, faulty product. So, I need to investigate why that product has become faulty. So, to do that first I can do diagnostic analysis and found out that one of the part was faulty. Let us say part A was faulty. Then we further investigated and found out that in the part A there was a component B. Which was faulty. So, if I have a digital twin, I can trace which part was faulty, I can trace which component was faulty. I can further go ahead and trace from which supplier this component was procured from, where that supplier is located at which plant they manufacture this component. Not only that I can also trace back and find out from which raw material suppliers this component was produced. So, therefore, I can trace the products across the supply chain. So, to do that I may need the help of blockchain technology, but if I have a digital twin which is blockchain supported, then I can trace not only the part, I can trace the component. I can also trace the raw material suppliers and specifically from which plant, which location the component was produced, which location the part was produced and which products this parts were used. So, all of this I can trace it easily if I have a digital twin which is blockchain enabled. So, therefore, it gives me better traceability. Then decision making process becomes more transparent, more transparent and accurate. Since it captures real time data and gives solution which are analytically driven. So, let us see few examples and then it will become much more clear to us. The first example is production planning and control. So, if I have a digital twin, then it helps me to do better production planning and control. I can find out at which factory, how much quantity of product A, how much quantity of product B, how much quantity of product C should I produce. So, that supply and demand gap is reduced. So, whatever is demand, I can match that by producing that much quantity in different factories. So, it helps me for doing better production planning and control. It also helps me for inventory control. So, let us take an example that you have 2 DC and 3 products. I have to find out at which DC, how much quantity of product A, how much quantity of product B, how much quantity of product C should I store. So, that my overall cost is minimized. So, what is the cost? One cost is cost of overstocking, another cost is cost of understocking as far as inventory control is concerned. So, therefore, if I have a digital twin with me, it helps me to decide my inventory level at factory, it helps me to decide my inventory level at DC, it helps me to decide the inventory level at customer location. So, that overall cost of overstocking as well as cost of understocking is minimized in each of this location. Then we have shipment control. So, I can find out using digital twin, where my track is currently located. I can also find out using digital twin from which factory to which DC, I should send the products. From which DC to which customers, how much quantity should I send to. So, it helps me to control the shipment. Then I also can do quality control. So, there is another example of quality control. Suppose you are managing a cold supply chain. So, therefore, I have to make sure that temperature is controlled as well as humidity is controlled. So, therefore, if I am moving the products in a truck, I have to make sure that I am continuously monitoring the temperature, I am continuously monitoring the humidity of the products when it is moving in the truck. So, if I have good quality control system in place. So, while the truck is on the move, while the truck is on the move, I will be able to trace it, I will be able to monitor it, I will be able to track it. What is the temperature at this current moment? What is the humidity at this current moment in the truck? So, that the products are safe and food or any other products which I am transporting can be monitored properly and there is no issue in terms of quality. Then there is a resiliency management. So, digital doing can help us to do manage resiliency in a better way. So, it will tell me alternate recovery plan after a disruption. Suppose, one supply location got disrupted. Then what should be my alternate recovery plan? So, if I have a digital twin with me, it will tell me which supplier at what location should I contact and how much quantity should I procure from them. So, that my overall impact of disruption is minimized. Then it has also capability to do sustainability management, specifically environmental impact due to redesigning a supply chain. Suppose, you have done some supply chain network optimization and I need to find out if I go with this specific supply chain network, then how much overall CO2 emission I will be incurring. So, what is the environmental impact of my new supply chain? That means, if I located a factory in a different location, how much extra CO2 I will be emitting? Or how to be my overall environmental impact? If I set up another DC in a remote location, then how much environmental impact I will have? So, I can do the environmental impact study if I have a digital twin with me for my overall supply chain. Then it helps managers to collaborate smoothly with other partner. So, collaborate as well as smoothly. This is very important. Everybody would like to collaborate with their partners, but if you have a digital twin, you can smoothly collaborate with your other partners because you have good visibility with you. You can trace each and every products, components as well as raw materials. The system becomes more transparent and accurate. So, therefore, collaboration will be much easier and what are the advantages of collaboration? I can do collaborative demand planning and forecasting. So, therefore, my forecast will be much more accurate and if the forecast is accurate. So, obviously, I will have less understocking, I will have less overstocking and the supply demand gap will be reduced. So, therefore, it is beneficial for the overall supply chain. I can also do collaborative sourcing planning. That means, I can collaborate with my other supply chain partners and decide where to procure the raw material firm. Who should be my suppliers? Who will give me the component? So, that overall cost is minimized and environmental impact is minimized and the whole process becomes sustainable both from economy as well as environment as well as social point of view. So, therefore, if I summarize the whole discussion, digital twin is a digital companion for a decision makers. So, it helps you to take decision in a better way. That means, what do you mean by better way? The decision should be analytically driven. So, therefore, I have descriptive analytics, predictive analytics and prescriptive analytics, tools and techniques which can help me to take decision in a systematic way in an analytical way. So, this is overall summary of supply chain digital twin. I am sure you got some idea about this supply chain digital twin. So, in the next class, we will see like how to form a digital twin, how to develop a digital twin and we will see some examples of supply chain digital twin, what are their exact benefit we will discuss in the next class. So, thank you so much for listening carefully. See you in the next class. Thank you.