top of page

Digital Manufacturing and Smart Factories: Optimizing in real time with the Digital Twin

Machines that autonomously coordinate production processes, service robots that cooperate with humans in assembly and driverless transportation systems that autonomously carry out logistics orders – this is the future vision of many companies. A factory where all components are linked and communicate with each other in real time via the Internet: machines, people, tools, and resources. The Smart Factory – the intelligent factory that organizes itself.

But what are the (technical) requirements for a Smart Factory? Which advantages does it offer? And what does a Smart Factory have to do with Digital Twins and Digital Manufacturing?

Digital Manufacturing is the Basis for the Digital Twin

A Digital Twin is the digital image of a real product or process: From an individual machine to a vehicle and its driving characteristics to an entire factory, everything can be reproduced virtually. The digital models are based on real-time data from sensors of physical systems. The Digital Factory provides the necessary data basis for the Smart Factory.

As a reminder, the Digital Factory provides methods and tools to virtually plan factories, production facilities and logistics. It maps the processes of a real factory in digital 3D models, simulates them and checks if the processes are feasible.

Dynamic data acquisition in real time in combination with the visualization of the data creates virtual models of industrial plants, so-called Digital Twins. These digital models are adaptive and can therefore adjust to reality. The condition is a two-way synchronization: The Digital Twin takes information from the operating data and the plants and compares it with its status. In case of deviation from reality the Digital Twin updates the virtual model. The physical object and its twin are permanently linked so that both develop identically. 

The goal: The operation and optimization of the factory in real time. Based on the recorded data and algorithms the physical network of machines can act more autonomously and flexibly than before – thanks to digital planning and simulation.

This requires that the machines are interconnected via the Internet of Things (IoT). By means of IoT sensors, the operational, status and process data are recorded in real time and sent to a cloud. The large amounts of data generated by the machines – also known as Big Data – converge on a digital platform. However, the data itself is useless. First, it must be evaluated or analysed in real time. In the smart factory so-called machine learning and AI algorithms conduct this task. In this way, a dynamic, real-time, and automatic self-organization of the production systems is to be achieved: the Smart Factory.

Potentials of the Smart Factory: Operation and Optimization in real time

The potential of a Smart Factory is diverse: First and foremost, the Smart Factory promises improving quality and efficiency in the (manufacturing) company. With the Digital Twin, development times and risks can be reduced – thanks to the highly efficient planning, optimization and adaptation of complex systems and processes. Predictive maintenance and monitoring become possible, downtimes can be avoided, and energy consumption is reduced. Business models can be expanded or redesigned and the quality of the products is improved. For this purpose, the large amount of data from the networking of the existing systems and machines is essential just as the visualized evaluation of the same: This creates the necessary transparency about the company's own processes, which in turn forms the basis for any optimization. The crucial question is: Where can processes be optimized or automated with technology? With agile decision-making companies can thus better decide how the next steps should look like. Consistently thought through to the end, this development leads to self-organizing production with the help of intelligent systems.

In practice, a process in the factory of the future could look like this: The customer orders online – and thus starts the production process. The ordered products autonomously reserve the corresponding process steps, book machines or required materials and control production. The production facilities organize the sequence of the orders and take care of maintenance and repair work if necessary.

Challenges on the Way to the Smart Factory

So much for theory – most companies are still at the beginning on the way to a Smart Factory with automated production: The 2020 study by the Fraunhofer Institute for Production Systems and Design Technology IPK in cooperation with msg systems ag on the Digital Twin in manufacturing industry showed that 85 percent of the surveyed companies have already developed concepts for the Digital Twin – but only 54 percent have an end-to-end strategy for Digital Twins. Only eight percent of companies are already using Digital Twins to their full extent. So far, they have mostly served as data-providing systems or are used for validation and error analysis. The design as autonomous or adaptive systems has so far only been considered in a few concepts. The diverse potential of the Digital Twin is so far largely untapped. 35 percent intend to change their business model with the Digital Twin and want to accelerate existing processes in the company in this way. For the future, 35 percent of the companies surveyed expect their Digital Twin to be able to map the entire system with its environment by 2040.

A major obstacle on the way to the Smart Factory is networking and data consistency – keyword company-wide collaboration and networking of the Digital Twins with each other. To take this step, it is necessary to standardize the platforms and communication interfaces by dissolving existing system boundaries and uniting application functionalities via a common and consistent database. The information relevant to the Digital Twin must also be provided consistently throughout the entire product life cycle. Another challenge is to integrate the simulation models of individual Digital Twins into a simulation of the entire system – a central requirement for the Smart Factory. If all systems can automatically exchange information and speak a common "language", only then will the vision of a self-managing factory become reality.


bottom of page