A digital twin is more than just a 3D model or a data stream; it's a complex combination of elements capable of tracking a system throughout its existence. After analyzing numerous use cases, two researchers from Télécom SudParis have identified a gap in the description of the architectures needed to build these digital doubles.
Digital twins are not mere virtual copies. Their value lies in their ability to make data collected from a system interact with models that describe and explain its behavior and evolution. However, the organization of these various components to support a product or service from conception to operation remains a complex challenge.
Noël Crespi and Roberto Minerva, researchers at Télécom SudParis, addressed this issue by coordinating a first collective work in 2023, titled The Digital Twin, with Adam T. Drobot. This book, featuring over 80 specialists, describes available technologies and their applications in sectors such as industry, health, energy, and cities.
Despite its over 1,200 pages, only a twenty-page chapter directly addresses the software architecture of digital twins. While applications and technological components are numerous, their organization within a complete system is still poorly described, with large companies closely guarding their know-how.
« We realized that an architectural reflection on digital twins was truly lacking. Outside of proprietary solutions, there isn't yet a clear roadmap, » notes Noël Crespi. Teams wishing to develop their own digital twin must find answers to questions such as: Where to start? How to choose and connect different modules? How to make predictions?
To fill this gap, the two researchers have produced their new book, The Foundation for Digital Twins: An Architectural, Technical, and Software Perspective. They present it as an open set of functions, rules, and constraints designed to guide the construction and operation of digital twins.
