GreyCat: A Framework to Develop Digital Twins at Large Scale

Digital Twins (DTs) have become a pivotal technology for enhancing the understanding, monitoring, and ultimately autonomous piloting of systems across various domains, including large-scale critical infrastructures such as smart electricity networks. The development of DTs necessitates developing diverse services that utilize different models and a digital shadow, which encompasses both real-time and historical data from the physical counterpart. The extensive scale of large infrastructures presents significant challenges, including managing numerous parameters, heterogeneous data, and the complex computations required, particularly with the increased use of AI algorithms. Current technologies, built by stacking multiple databases and using general-purpose languages, are inadequate for efficiently implementing digital twin services that need runtime reactivity. This tool demonstration paper introduces GreyCat, a framework designed for the development of digital twins over large-scale digital shadows. GreyCat combines imperative object-oriented programming, database persistent indexes, and scalable memory management to facilitate the creation of comprehensive and efficient digital twins. We demonstrate the ease use of GreyCat through simple examples and showcase its effectiveness in constructing the national digital twin of Luxembourg's electricity grid, which is currently operational and managing billions of data points. Reflecting on the development of GreyCat over the past years, we discuss the main lessons learned and identify open questions for future digital twin development frameworks.