Virtual Commissioning Evolves into a Model-Driven Digital Twin

Virtual commissioning needs to go beyond emulating controls logic for emulation hardware. Model-driven digital twins make virtual commissioning more accessible and adds the power of simulation technology and capability to the overall automation process.

Dick Slansky, senior analyst, ARC Advisory Group
Dick Slansky, senior analyst, ARC Advisory Group

Virtual commissioning (VC) technology allows manufacturing and controls engineers to virtually simulate manufacturing production systems and validate that the physical packaging machines, conveyance systems, automotive production systems, robotic work cells, and controls will all physically function as designed.

VC uses a virtual model that represents an accurate 3D simulation of mechanical, electrical, and controls systems to validate the physical functions of a production system prior to implementation. The inherent complexity of integrating the different engineering disciplines previously necessitated a labor-intensive commissioning process. VC technology and applications were developed to greatly reduce or eliminate the physical process, shortening the time to launch, and ultimately producing significant cost savings.

The initial VC applications emerged as part of the overall digital manufacturing portfolios offered by product lifecycle management suppliers. Here, 3D CAD models of machines, robotic work cells, and production systems could be used to virtually simulate motion and production functions. The other part of VC involved creating software that would emulate the control systems to virtually test the physical system. Today, we are seeing the convergence of traditional VC with the more recent emergence of the concept and implementation of the digital twin.

Evolution of VC
The automation industry has long acknowledged the benefits of using virtual models to simulate the performance of physical systems, enabling integration issues to be identified before entering the time-consuming and expensive process of physical commissioning. To successfully implement VC, however, the virtual factory model must be accurate. While these types of simulation models were used with some success in the aerospace and automotive industries, this was not the case in the overall automation market.

Controls engineers and automation researchers have organized four categories of general controls development:

  • Physical commissioning, which involved testing the physical systems (factory production systems) against the hardware without the assistance of virtual modeling tools.
  • Model-in-the Loop (MiL) where the application creates a logic model of the programmable logic controllers (PLCs), human-machine interfaces (HMIs), and electrical and mechanical systems. The application connects the logic model to a simulation model of the production system.
  • Software-in-the-Loop (SiL) is the software code that runs the logical model.
  • Hardware-in-the-Loop (HiL) testing, which uses a virtual production systems model to test the hardware controllers. This is some-times referred to as controls emulation.

The actual VC process is usually an iterative approach using MiL, SiL, and HiL concurrently. Once the MiL is complete, controls engineers use SiL testing to verify that the logic in the model is consistent once it has been compiled into machine code. If no errors are found at this stage, final HiL testing is conducted by compiling the software onto the physical PLC or HMI. Today, suppliers of robust VC development and simulation platforms typically provide a range of simulation and VC applications that meet this approach.

VC becomes part of the digital twin
Today, we’re seeing the convergence of established VC technology with the more recent emergence and implementation of digital twin technology across industry and business. While VC represents the simulation and modeling of machines and production systems to virtually validate the system and the controls that automate it, the digital twin is broader in scope and involves capturing sensor data from physical machines and systems in operation and using that data to create simulations in real time. Because of its real-time characteristics, a digital twin can simulate a system while it is operational, allowing manufacturers to monitor the system, create models for adjustments, and make changes to the system. 

Model-driven digital twin advances VC
For VC to become a practical technology across manufacturing and automation, automation generalists need to be able to create and use the virtual models for complex simulation applications. The development of advanced, model-driven design methods have taken the form of the digital twin. Additionally, the continuous advancement of simulation modeling applied to today’s production systems offers a much more robust and accurate virtual representation than the earlier and simpler modeling tools for VC. Moreover, the software development standards for model rendering and connectivity have been improved significantly. Taken together, these make VC more practical for the automation industry.

Model-driven digital twins
With systems design modeling tools, the creation of a model-driven digital twin can begin concurrent to the design process. Model-driven digital twins make VC more accessible and add the power of advanced simulation technology and capability to the overall automation process.

The primary goal of the commissioning process, whether physical or virtual, is to bring a completely integrated, assembled, and validated mechanical, electrical, and controls software production system into operation. The challenge for successful VC implementation goes beyond just virtually emulating controls logic for the automation hardware. It involves integrating all the engineering disciplines of mechanical, electrical, and software logic design into a systems design approach that normalizes the constraints that each system places on the other.

Companies in this article
More in Plant Maintenance