• Published: May 2022
• Author: Byong-Woo Jeon
• Read from 2025.11.24 to 2026.01.28

Backgrounds (Motivation)

• Looking at just the term "digital twin", it gives the impression of a digitally created twin, which might be dismissed as merely 3D modelling of a product. However, it actually refers to a system where all data from when a product is planned until it's disposed of remains in virtual space, affecting and interacting with actual products and services.
• When I was working at Superbin, we had a vision to operate the Flake factory by introducing the digital twin concept. However, with the capabilities at the time, we only managed to effectively monitor processes from the control room by introducing SCADA (I'm not sure whether the factory infrastructure was actually built based on SCADA).
• That's how I became interested in the concept of digital twins. It's a book published in May 2022, but I'm only reading it now, three and a half years later.

Book review

• Just as taking an introductory psychology class doesn't make you a psychological counsellor, reading this book in its entirety doesn't give you the design capability for a digital twin factory.
• This book comprehensively covers the knowledge required to apply 'digital twins' without limiting it to a specific industry. Of course, it uses examples from parts of manufacturing businesses, but the content in the book will still feel quite abstract if you're trying to follow along and apply it step by step.
• This book covers an extremely broad range of knowledge needed for implementing digital twins—the A-to-Z, if you will—starting from hardware facilities, components and materials, through to artificial intelligence and optimisation algorithms, databases, cloud systems, sensor-to-server communication, and even software-to-software communication. The areas I'd been working in, I could get through relatively easily as a kind of revision, but topics I was encountering for the first time could be difficult to understand. I was impressed by the author's incredibly broad and comprehensive knowledge.
• To implement the digital twin described in this book, you'd first need to master the relevant standards (ISO #####) down to the component level for production facilities, and down to individual module level for the software that will operate the system, maximising modularisation. Then you'd need to apply all the necessary sensing technology to extract data required to copy not just the production targets but also the production infrastructure into the digital world. And it doesn't stop at simple monitoring—you need to build a simulation environment that can provide insights genuinely useful to the actual factory.
• You'd obviously need groups of highly skilled software developers in each field—embedded software, firmware, fullstack software, and AI. Plus you'd need facility and tool experts for each production line to be applied in the physical environment, process design experts, and a coordinator with broad knowledge like this book's author who can orchestrate all these people.
• I suppose ignorance is bliss, as they say. When I heard about the plan to apply digital twins whilst building a smart factory for flake recycling at SuperBin, I didn't realise that's what it meant. In the end, I don't know the details of how the factory was actually implemented, but I suppose when you're chasing big dreams, you do learn a lot along the way.

Part 1: The Concept of Digital Twin

Part 2: Dev. of Digital Twin