Digital twins in Manufacturing

Digital Twin, as a promising framework to address the integration between the cyber and physical in manufacturing, is gaining increasing attention from both academics and industry. Through high-fidelity models and real-time data from sensors, Digital Twins can project physical assets or processes into the digital world to reflect the whole lifecycle process of the corresponding counterpart while maintaining convergence, co-evolution, and symbiosis of physical and digital assets.

This talk will highlight recent advances developed in our laboratory to address Digital Twins for Smart Manufacturing. Applications in Product Engineering, Additive Manufacturing and Robotics will be presented to illustrate our research findings.

Simulation based digital twins are becoming increasingly important for a flexible, fast and safe integration of cobots. A digital twin can create insights into the operational behavior and make optimizations to a cobot system quickly and safely. While maintaining a digital twin throughout the life cycle of cobot systems can also be useful for fast and trustable validation of any future modifications, thus making them truly adaptable. Human interface with digital twins will also be presented during the talk.

SDU Robotics have developed a Digital Twin (DT) methodology based on the perspective that DTs need to be the central interfaces between robotic technologies and any digital methods for planning and control. The DT methodology started with addressing the gap between collaborative robots (“safe to use/low payloads”) and conventional industrial robots (“high payloads/unsafe to use”). In the work with industrial partners such as LEGO and KUKA on DT-based solutions which allow constant monitoring and analyze robotic systems in order to establish safe human-robot interactions – even with industrial robots. In other examples, simplified DTs have been used to analyze the potential of deploying collaborative robots for SMEs and then quickly scale the analysis to controlling actual robot installations. The advantages of the DT methodology are particularly visible when leaving the highly controlled environments of factory automation. By extending the DT concept to represent also production environments, DTs have enabled entering of novel robotic applications on construction sites and on shipyards in several ongoing projects.

Aarhus University Centre for Digital Twins are currently involved in 12 research projects which are externally funded from many different agencies including Horizon 2020, Innovation Foundation Denmark, the Independent Research Fund Denmark and public authorities. The full budget for partners in total is close to one billion Danish Crowns and the AU budget is more than 80 million Danish crowns. This goes all the way from basic research over applied research to innovation together with different companies from all over Europe (more than 120 companies in total).

The presentation will provide an overview of how digital twins can be used in many different application domains (including agriculture, civil engineering and manufacturing). This presentation will also illustrate the limitations and possibilities with digital twins at different levels of ambition.

Industrial robots are used extensively in today’s society. In many applications, it is important to have some kind of real-time sanity checking during operation.

For instance, if the robot performs some task, do the currents of its electric actuators correspond to what we expect? If not, did we hit a human coworker? Alternatively, did the friction of the gear transmission slowly change from the expected value? If a change in friction did take place, was it due to temperature changes affecting the lubricant of the transmission or is it due to the gear being worn out? To be able to answer these questions, we must have a digital twin that accurately resembles the industrial robot.

Thus, in this presentation, Emil Madsen will explain how to obtain a high-fidelity digital twin of an industrial robot.

Unique and innovative possibilities are emerging when combining simulations and real time data in digital twins. During this presentation, FORCE Technology will give an introduction to the concept of simulation and physics driven digital twins. Here you can find out how to leverage insight and information from high fidelity simulations, e.g. developed in the design phase, on assets in operation. The approach opens among others for running what-if-scenarios based on physics and simulations and doing predictive maintenance.

The Danish Technological Institute has developed a true Digital Twin (DT) of a slaughter line with real-time communication between the physical and the virtual world. How can a 3D model of a slaughter line with virtual objects of both products and machines help to obtain a more sustainable food production? What are the potential benefits of having a DT model to run in parallel with the actual slaughter line? The next step we are looking into is how to visualize different types of data from different data domains. What is the value of continuous monitoring and diagnosing data using artificial intelligence to visualize errors in the DT?

The Digital Twin (DT) model will drive 21st century manufacturing by dramatically improving both efficiency and effectiveness. DT information will replace wasting resources and make Lean Manufacturing a reality. As DTs advance, manufacturing will move from unplanned failures to failure prognostics that will remove problems before they occur. This manufacturing evolution not only requires technology advances, but will also requires cultural changes. As the originator of the DT concept, Dr. Grieves is uniquely positioned to give his perspective on 21st century manufacturing.