Reference Architecture to Exploit Data as a Product Through the Cloud-Edge Continuum

The Internet of Things (IoT) computational power and storage requirements are constantly increasing, implying a subsequent rise in the amount of data created by devices at the edge of the network. While cloud computing has been a well-established method of purchasing services for many applications, it may not be enough to handle the amount of data from IoT devices while simultaneously addressing the heterogeneous application needs. The Cloud-Edge-IoT (CEI) continuum, where sensor networks and data coexist with cloud and edge resources, appears then to be a suitable solution to address future data-intensive applications. CEI continuum employs edge computing, cloud computing, and the IoT to boost the Industrial IoT (IIoT) development. However, the distributed nature requirements of digital manufacturing present difficulties for typical cloud-centric techniques: to address some of these issues, an analysis of current Reference Architectures (RAs) has been performed. Consequently, the opportunity for a new RA has been discussed, leading to include capabilities supposed to mitigate some CEI continuum issues, among which the concept of “Data as a Product”, supported by specific Data Container. In addition, it has been noticed how Federated Learning services can improve privacy and security issues, as well as computing effort of the CEI.