According to the estimates of the analytical company Gartner, two thirds of all enterprises that have implemented the internet of things, will use at least one digital twin in time of their operation. Therefore, in order to upgrade digital twins, one small step is enough to do so.

5G network development

At this year’s Hannover fair, the companies Ericsson and Comau introduced a digital twin equipped with 5G connectivity. This demonstration introduced a digital form of assembly line in a car factory. This device was, using 5G connectivity, gathering data from sensors monitoring machines and also processes. The monitored parameters like vibration, temperature and pressure were displayed on the digital dashboard on an ordinary tablet. These data are usable for identification of operation interruption that could possibly slow down the entire production process.

Thanks to 5G connectivity, it was possible to gather the data steadily, continuously and in great volumes in real time which is a necessary requirement for operation of automated systems. Minimum rate of delay enables displaying data, such as visual outputs and data analytics is able to predict errors and failures or identify components that should be fixed.

Artificial intelligence introduction

Artificial intelligence adds repeatability to digital twins. For example, in cooperation with Bosch, NVIDIA has developed a system of predictive maintenance for boilers used in households. This system sends a message to the owner in case the pressure has dropped or the temperature has risen beyond the set limit. For inclusion of hundreds of boilers out of the company’s portfolio, a digital twin using the same algorithm and neural network enabling it to self-learn has been created. This way, the digital twin could be replicated within the full range of the boilers type offer.

As Andrew Cresci, Industrial Business Development Manager of NVIDIA, states: “The beauty of a digital twin is that once you have got your mind around one type of machine, it’s relatively easy to retrain and adapt and reuse; you can take AI and repeatedly reuse it instantly making data scientists more productive,” Artificial intelligence monitors data and identifies anomalies. Therefore, it is constantly improving in learning and identifying errors. The system specialist evaluates and describes individual errors making them usable in the next phase of learning. 

“In my opinion, this is one of the biggest benefits of AI – an ability to be repeatable and learn in a new environment very quickly,” Cresci said. As the demonstration of the digital form of assembly line showed at the Hannover fair, the data delivered by the sensors reached a much higher rate of relevancy, using the connectivity.

Information can be gathered not only for evaluation of non-standard situations. But also for understanding why the deviation from the set course occurs.