Overview

Understanding the levels of a digital twin

Digital twin technology evolves through several stages, each building upon the previous one:

• Level 1 – Mirroring: connecting real-time plant data to a virtual model, ensuring the digital twin reflects the actual plant state.
• Level 2 – Simulation: enabling “what-if” analysis and validation of different operating scenarios.
• Level 3 – Prediction: forecasting plant behaviour and detecting potential issues before they occur.
• Level 4 – Optimisation: recommending optimal operating conditions to maximize efficiency and reliability.
• Level 5 – Autonomous Operation: achieving intelligent, self-adjusting plant operations with minimal human intervention.

Among these, mirroring forms the critical foundation—not only to simulate, but to generate metadata that continuously reflects the plant state. However, effective mirroring cannot be achieved with a steady-state model alone. The Mirror Model (M1) must be a dynamic model, continuously capturing real-time data and transient events.

Online mirroring process digital twin: why two-model approach (M1 + M2)

A single model cannot capture the full spectrum of plant behaviour, such as:

• Inventory tanks where level and mass balance evolve dynamically.
• Transition phases such as start-up, shutdown, or feed changeover.
• Batch or semi-batch operations, which inherently require a time dimension.
• Asset performance degradation over time.

That’s why the two-model approach is essential

• M1 (mirror model): a dynamic model, always running to track continuous data and transient events.
• M2 (calibration model): a steady-state model, executed periodically to recalibrate parameters (fouling factors, efficiencies, catalyst activity) and keep the twin consistent in the long term.

Together, this structure ensures that mirroring supports both continuous real-time tracking and sustained alignment with the physical plant.

Another advantage of the two-model approach is that it naturally extends into a prediction digital twin. Once the mirroring loop is established, it becomes straightforward to connect with:

• M3 (prediction model): used for “what-if” studies, scenario analysis, and on-demand optimisation.

In other words, the mirroring structure (M1 + M2) provides the solid foundation for predictive capabilities. Without accurate and continuously calibrated mirroring, predictions would quickly drift from reality.

Speakers

Jay Yun, CEO, SIMACRO 

Jay is the founder and CEO of SIMACRO and an experienced professional in process modelling, simulation, and digital transformation. He began his career at AspenTech, where he developed expertise in advanced simulation and optimisation software widely used across the chemical and energy industries.

He went on to work at Technip and SABIC as a process engineer and subject matter expert in modelling and simulation. In these roles, he applied advanced tools to solve plant challenges, optimise operations, and support large-scale project execution.

With over 28 years of experience spanning software, EPC, and operating companies, Jay has gained a rare combination of perspectives across the entire value chain. This diverse background became the foundation for his entrepreneurial journey in establishing SIMACRO, where he now leverages his global expertise to deliver innovative digital process solutions.

Wonseok Lee, Executive VP, SIMACRO

Wonseok is an Executive VP Operations for SIMACRO. Prior to this he has worked at AspenTech plus Hyprotech as e.g. senior manager area sales and various consultant and engineer roles. He has an MS in chemical engineering with over 25 years of experience in process design, modelling and simulation.

He has developed deep field expertise in business consulting and solution sales through his key roles at AspenTech and Hyprotech along with a proven ability to deliver innovative solutions and drive value for clients in the process industries.

The material presented has not been peer-reviewed. Any opinions are the presenter’s own and do not necessarily represent those of IChemE or the Process Management and Control Special Interest Group . The information is given in good faith but without any liability on the part of IChemE.

Time

10:00–11:00 GMT.

Software

The presentation will be delivered via Microsoft Teams. We recommend downloading the app from the Microsoft website, rather than using the web portal.