Digital twins of food process operations: the next step for food process models?. (October 2020)

• Record Type: Journal Article
• Authors: Verboven, Pieter; Defraeye, Thijs; Datta, Ashim K; Nicolai, Bart
• Is Part Of: Current opinion in food science. Volume 35(2020)
• Page Start: 79
• Page End: 87
• Publication Date: 2020-10
• Subjects: Food -- Microbiology -- Periodicals; Food -- Analysis -- Periodicals; Food -- Biotechnology -- Periodicals; Food -- Periodicals; Food -- Research -- Periodicals; Nutrition -- Periodicals; Food -- Biotechnology; Food -- Microbiology; Food -- Analysis; Food; Food -- Research; Nutrition; Food Technology -- Periodicals; Food Analysis -- Periodicals; Food Microbiology -- Periodicals; Periodicals; Electronic journals; 641.3
• DOI: 10.1016/j.cofs.2020.03.002
• Languages: English
• ISSNs: 2214-7993

Abstract

Highlights: Digital twins have emerged recently as digital replicas of industrial processes. Physics-based food processinduce comprehensive mechanistic twins. Digital twins will give augmented insight and more versatile operational management. First applications of digital twins in agrofood sector have been proposed. Abstract : Food process modeling has matured with the development of multiscale, multiphase and multi-physics approaches. More comprehensive numerical tools and software platforms for improving insights and optimizing designs and processes have emerged. In the context of industrial digitalization and the advent of the Internet of Things, the concept of the digital twin has recently emerged as a means for more versatile process operational management. The digital twin is defined as a virtual replica of the real process operation, which is connected to the real world by sensor data and advanced big data analytical tools. While all elements are available for implementing digital twins, with the different types of models playing a central role, it will require a multidisciplinary approach for successful implementation and operation. The first agrofood applications still need to be demonstrated. This paper mainly focusses on the role more physics-based models can play, in addition to data-driven and hybrid models.