Interview with Elena Hidalgo, Digital Industry Area Manager at CETIM
A digital twin is a term we hear more and more, but can you explain exactly how this model works and why is it key for GIGAGREEN?
The concept of Industry 4.0 has been used a lot in the last few years. This is a way to express that the Fourth Industrial Revolution is here and that it involves the so-called Key Enabling Technologies. Advanced Manufacturing is one of those and, here, the Digital Twin (DT) is supposed to be a turning point to be able to get more competitive processes.
According to IBM’s definition, a DT is a virtual model designed to accurately reflect a physical object. The principal aim of a DT is, thus, that it allows the simulation of a real process where it gets information beforehand about how the real model will work, without the costs assumed using physical models or prototypes. In this sense, different types of DT could be created as Components, Assets, Systems or Process Twin. Likewise, depending on the DT type, different technologies could be used. For example, Artificial Intelligence (AI) can be applied to create predictive models, classify data, extract patterns, and detect anomalies to optimise the process. It can also be combined with sensors or Artificial Vision for objecting and defecting, code reading, positioning, or even for predictive maintenance.
The reasoning behind the use of DTs is, on the one hand, that they allow for predicting possible consequences in different scenarios without any implementation in real conditions. Thanks to this, it is also possible to deploy corrective measures in the process before the problem appears, avoiding thus, critical situations from happening. Moreover, with this technology, it is possible to prepare the process to face possible future problems and ensure it will be able to address them. On the other hand, a DT is an excellent tool to identify bottlenecks, staff needs, or other improvements needed. Finally, as we mentioned above, the possibility to integrate AI algorithms as well allows data processing to detect patrons or deploy machine learning or neuronal networks.
In conclusion, by using DT, we guarantee that we optimise and prepare our products, processes, or services without unnecessary investments in material or time spent in building physical prototypes. However, their complexity implies counting on specific knowledge from an expert to build an efficient simulation model. For GIGAGREEN, CETIM placed at the disposal of the consortium its expertise in deploying different types of DTs.
In this case, CETIM is designing a DT based on a simulation model made with AI to make the scale-up of the batteries manufacturing process possible, thanks to a digital model able to improve the total waste of time, money, and energy. For CETIM, this is the hardest part of the work, because it implies understanding deeply, not only the whole process but also the project parameters and their different relations among them. As a result of the first analysis, CETIM is adjusting a set of variables to start building the dataset, which is the basis of the AI model.
