In the age of digital transformation, many industrial companies are looking at technologies that will make them fit for the future. In addition to buzzwords such as Industry 4.0, Internet of Things (IoT) and machine learning, the term “digital twin” is often mentioned in this context. This virtual image of a product or manufacturing plant can play a decisive role in making production and development processes more efficient, in reducing costs and in increasing availability.
“The virtual image of a real object that represents its properties and behavior” – admittedly, the definition of a digital twin does not sound particularly exciting at first glance. However, if one realizes what actually lies behind this initially rather sober description, quite promising perspectives open up.
A digital twin is the virtual image of a real object. A digital double, so to speak. The real object can be a final product to be manufactured, a machine, a network of machines or an object that does not yet exist but is in the planning stage. In the digital world it is described by models with algorithms, data and states. The second part of the definition talks about the digital twin representing properties and behavior. In other words, it represents the sum of the characteristics and functions of the real object, to which it behaves exactly the same. Whereby “exactly” is to be understood here rather theoretically. This is because the lack of accuracy of the underlying data often leads to certain limitations. However, these usually do not have any negative consequences for the function to be mapped.
The question still remains as to what the digital twin has to do with Industry 4.0, Internet of Things and Machine Learning. In view of the above definition, it would seem reasonable to assume that the digital twin is the connecting element that gives form and benefit to the aforementioned technologies. In principle, Industry 4.0, IoT and Machine Learning also work on their own. However, those who want to derive the greatest possible benefit from the technologies fully exploit the existing potential with a digital twin.
The greatest advantage of a digital twin is that a real object is no longer needed to study the behavior and properties of this real object. This is done by a simulation based on the virtual image. Departments in charge of product development no longer need to manufacture an object to check whether it meets the requirements placed on it, whether its behavior provides the desired functions, and which aspects might be critical in terms of wear and susceptibility to failure. Thus, the use of digital twins not only speeds up the development of a product, but also increases the efficiency of the manufacturing process and provides important insights in terms of maintenance windows and fail-safety of the installed components.
The benefits of the digital twin can best be illustrated using typical use cases from the real world. During product development, prototypes are essential to determine whether the features and functions of the planned product meet the requirements. However, a typical dilemma in development projects is that prototypes are always available too late and in too small numbers. With the help of the digital twin, business departments can use virtual prototypes early in the development process to validate requirements, correct errors and implement optimizations in various areas. This approach not only minimizes risks, but also ensures a better end product.
Similarly, complex networks such as manufacturing plants can be mapped as a digital twin. This visualizes production-relevant data and prepares it in such a way that companies can optimize both the individual machines in the plant and the entire production process.
In addition, sensors at relevant points on the machines can be used to identify signs of wear before a breakdown occurs. This enables maintenance windows to be planned in advance. This concept of predictive maintenance helps to intervene at an early stage and minimize unplanned downtime in production.
How exactly the digital twin based on Industry 4.0 and IoT generates benefits is illustrated by the example of a production network from the manufacturing industry. The production network consists of a sensor-equipped, networked and fully automated press and an identically equipped punching machine. A digital twin of this production network is to be built. The following steps describe what the procedure might typically look like.
The basis for everything else are digital descriptions of the machines used. They can include CAD data, parts lists, maintenance plans and many other data. This creates the first model of the digital twin, which is networked with the real machines to exchange production and status-relevant data. The digital twin has the identical status as the real machines. SAP Business Technology Platform provides a solid basis for the exchange. The functions and modules available there help to realize the virtual image of a machine. Based on the collected data, it is possible to optimize production processes by simulating them on the digital twin, both within an individual machine and across the entire network. In addition, the knowledge gained can be used to improve the final product to be manufactured as well.
In a further step, the machines transmit sensor data in order to detect wear parts with the help of machine learning and thus optimize the design of the machines or schedule maintenance windows in production before there is a longer-term breakdown of machines due to a defect. In addition, Machine Learning offers the possibility to refine the models of the digital twin. SAP Internet of Things provides important functions for this, for example, the connection of sensor technology to the cloud, the evaluation of data using machine learning, and the graphical visualization of the insights gained.
The solutions of the SAP Intelligent Asset Management suite (IAM) provide support here. On the one hand, SAP Asset Intelligence Network (AIN) platform connects the various stakeholders such as manufacturers, operators and service partners with each other in order to obtain a common view of an asset. This keeps all parties involved on the same level with regard to machine data, documentation, maintenance information, replacement components and other relevant information.
On the other hand, SAP Asset Performance Management (APM), which emerged from SAP Asset Strategy and Performance Management (ASPM) and SAP Predictive Asset Insights (PAI) solutions, optimizes maintenance. In combination with an SAP S/4HANA system, or more precisely with the Enterprise Asset Management module, an end-to-end maintenance process can be mapped, which enables the development of suitable maintenance strategies through to the monitoring of equipment using machine learning. This creates the basis for predictive maintenance.
The technological prerequisites for making production more efficient and reliable with the help of digital twins are available. In the age of Industry 4.0, the smart factory can already be implemented in many ways today. SAP provides the appropriate tools for this. Now it is up to the companies to take advantage of the available opportunities and make their manufacturing processes fit for the future.
The scenarios outlined represent only a small selection of examples from the digital production world. With the help of the right tools, they are well suited as an introduction to Industry 4.0. The digital twin provides companies with many advantages. They benefit from shorter development and production times, more fault-tolerant products and higher availability. The autonomous factory, in which intralogistics is networked with manufacturing processes and the product and which controls and optimizes itself, is not far away.