An Edge-Cloud based Reference Architecture to support cognitive solutions in the Process Industry
Antonio Salis,Angelo Marguglio,Gabriele De Luca,Sergio Gusmeroli,Silvia Razzetti
DOI: https://doi.org/10.48550/arXiv.2202.06622
2022-02-14
Distributed, Parallel, and Cluster Computing
Abstract:Process Industry (PI e.g. Steel, Metals, Chemicals, Cement, Asphalt, Ceramics) is one of the leading sectors of the world economy, characterized however by intense environmental impact, and very high energy consumption. In spite of a traditional low innovation pace in PI, in the recent years a strong push at worldwide level towards the dual objective of improving the efficiency of plants and the quality of products, significantly reducing the consumption of electricity and CO2 emissions has taken momentum. Digital Technologies (namely Smart Embedded Systems, IoT, Data, AI and Edge-to-Cloud Technologies) are enabling drivers for a Twin Digital-Green Transition, as well as foundations for human centric, safe, comfortable and inclusive work places. Currently, digital sensors in plants produce a large amount of data which in most cases constitutes just a potential and not a real value for Process Industry, often locked-in in close proprietary systems and seldomly exploited. Digital technologies, with process modelling-simulation via digital twins, can build a bridge between the physical and the virtual worlds, bringing innovation with great efficiency and drastic reduction of waste. In accordance with the guidelines of Industrie 4.0, the H2020 funded CAPRI project aims to innovate the process industry, with a modular and scalable Reference Architecture, based on open source software, which can be implemented both in brownfield and greenfield scenarios. The ability to distribute processing between the edge, where the data is created, and the cloud, where the greatest computational resources are available, facilitates the development of integrated digital solutions with cognitive capabilities. The reference architecture is being validated in the asphalt, steel & pharma pilot plants, allowing the development of integrated planning solutions, with scheduling and control of the plants.