In recent years you may have heard the term “Digital Twins”. This concept was born in 2002 by Dr Michael Grieves as part of his research programme at the University of Michigan, where he wanted to recreate highly detailed models of complex real-world objects and processes in virtual environments. However, it is only now that the deployment of Digital Twin capabilities has accelerated due to new technologies such as Cloud architectures or intelligent sensors and devices (IoT), which allow simulation models to have accurate data in real-time.
Based on this idea, the Digital Twin consists of creating a virtual replica of a product, process or service using real data, and simulating its behaviour in the face of different changes or stimuli, making it possible to analyse its performance and results to improve its efficiency. It is based on data science and applies advanced analytics techniques, Machine Learning, AI, optimisation algorithms, constraint management and the latest demand forecasting methods, to make it a very powerful tool. In this way, it is possible to compare what happened in a given scenario with the predictions and analyse deviations, updating and improving the models over time.
According to the consulting firm Gartner, by 2021, half of all large industrial companies will be using Digital Twins, which will translate into a 10% efficiency improvement. Boston Consulting claims that companies that have implemented them have achieved sustainable inventory reductions of up to 5% through better resource planning. And Capgemini says in a recent study that supply chains with this technology have shown greater resilience to the pandemic.
The market for Digital Twins is expected to grow by more than 38% each year, according to recent research by MarketsandMarkets.
Types of Digital Twins
There are arguably two different types of Digital Twins, models that simulate different scenarios about a physical product and object, and models that replicate complex processes.
For example, if we were talking about a manufacturing company we could simulate how a physical product would respond to certain changing environmental conditions or usage patterns, looking at wear and tear, the behaviour of materials under extreme conditions or how the item might break in a real usage scenario. And on the other hand, the production process could be simulated, what machinery to use, how to build the product, with what processes and in what order, etc. Although, as we will see later, this Digital Twin can be extended along the entire supply chain, with simulations in logistics processes, for example.
Companies such as our technology partner Delmia Quintiq take a unique approach by creating an integration flow of both twins, bringing simulation capabilities together on a single platform so that their use and optimisation generate greater value.
Uses of the Digital Twin
As explained in the previous point, one of the most common uses of Digital Twins is in the manufacturing industry. It can be said that simulation is the essence of what we call Digital Manufacturing. Thanks to Digital Twins, it is possible to design and manufacture a product in the virtual world and put it to “work” to see how it performs its functions. Years of use and operation can be simulated in hours or days, maintenance and repairs can be tested to see their effect, and operating costs, etc. can be determined. This makes the planning and management of operations more accurate and efficient, ensuring that deadlines and the level of service expected by customers are met. This technology has become a cornerstone for example in the automotive industry, bringing great value to international companies such as Honda, Tesla and Kreisel Electric.
We can see the implementation of Digital Twins all along the supply chain, from the factory to the shop. At the point of sale, for example, this technology enables real-time exchange of consumer information from the shop and the virtual environment for analysis. In this way, a more efficient arrangement of products or employees can be suggested to optimise sales.
In healthcare, cardiovascular researchers are creating highly accurate Digital Twins of the human heart for clinical diagnostics and medical education. And in the energy sector, massive amounts of data are captured and analysed in oil wells, which are used to build digital models to guide drilling activities in real-time.
Digital twins in the supply chain
As our friends at Supply Nexus explain in one of their latest published papers, Digital Twins can be used to replicate an entire supply chain or a part of it that is fed in real-time with transactional data from the producing business.
In addition, Machine Learning and Artificial Intelligence capabilities, advanced analytics techniques, optimisation algorithms, constraint management and the latest demand forecasting methods can be applied to the Digital Twin to make it a very powerful tool in Supply Chain management.
The following table shows the different points where this technology can be applied in the Supply Chain:
In the Supply Chain, this technology can be applied in all areas. Supply Nexus highlights some examples:
- Planning: The Digital Twin enables scenario planning so that the business can make decisions based on commercial criteria rather than solving problems as they are detected. For example, balancing service levels with inventory needs and costs.
- Warehouse management: It can simulate changes in demand, layout modifications, or even simulate the robotisation of a traditional warehouse. The system could provide information on the availability of each product, make predictions and autonomously make decisions on stock or deliveries.
- Re-planning of the logistics network: It allows the simulation of scenario changes, for example, in the case of wanting to prepare e-commerce in physical shops, it could simulate which centres are the optimal ones to use, taking into account numerous variables: geographical location, number of customers, expected online sales, integration with the transport network, number of employees, climate, etc. Another activity that can be carried out is the monitoring of shipments to better understand and manage the physical assets available.
- Optimisation of the transport network: Simulations of the different transport routes, personnel or fleet planning, pollutant emissions, etc. can be carried out.
- Incidents in the Supply Chain: On the one hand, it allows the simulation of different scenarios that involve an interruption in the Supply Chain. In this way, we can see the effect that a pandemic such as COVID-19 can have on our Supply Chain or the problems that the Filomena storm can cause, allowing us to draw up contingency plans for these situations. On the other hand, once the disruption has occurred, different recovery scenarios can be simulated so that we can return to normal operations in the shortest possible time.
Supply Chain performance can be simulated by identifying volatile points and uncertainties, as well as optimisation points.
Decisions or analyses that previously required days or weeks of work can now be made in a moment, or even automatically and in real-time.
The Digital Twin is a virtual environment that is a replica of a real one, which obtains and integrates data in real-time through sensors installed in the machines or Big Data technology. In addition, it analyses data and draws conclusions using AI and ML, and troubleshoots proposes optimisations and predicts operations.
The Digital Twin model offers unparalleled capabilities to track, monitor and diagnose assets. It is changing Supply Chains, bringing a wide range of options to facilitate collaborative environments and data-driven decision making, simplifying business processes and creating new business models. It is a tool that helps to improve our Supply Chain making it much more resilient.
The future of Digital Twins is focused on improving in the areas of accuracy and data usability/accessibility. The more realistic and accurate the simulation, the better the results.
Do you want to know how we can help you at decide4AI?