Digital Process Twins

Offer Greater Insights Through Simulation on Manufacturing Processes

From dashboard enhancements to predictive maintenance schedules, digital process twins empower engineers with a more complete picture of operations

Digital twins have become increasingly popular in the age of Industry 4.0. Created using advanced computer models and simulations, these digital representations allow engineers to mimic the behavior and performance of their real-world counterparts in real time. They are used to improve efficiency, reduce costs, and optimize performance.

For the process manufacturing industry, a digital twin is a representation of a physical asset, process, or product. There are many benefits to using digital process twins. They allow engineers to monitor and optimize the performance of physical assets and systems in real-time. They also can be used to predict the behavior of the same assets and systems under different conditions. Furthermore, digital twins can help determine when physical equipment is likely to fail. Operational experts can use that information to find the right time to perform predictive maintenance.

To develop the digital twin of a manufacturing process, engineers first must be able to access and analyze its operational data. Engineers can use this information to simulate the real-time streaming of data, which can be used to help predict process behavior.

In this article, we dive deep into the capabilities required to create, manage, and leverage operational digital twins into four capability areas:

  • Digital Twin Modelling & Governance: Modelling of all elements of a physical system (e.g., equipment, processes, sites, etc.) and their relationships,
  • Data Contextualization: Bringing all data into a coherent digital representation of the reality,
  • Analytics & Simulation/Modelling: Leveraging the (unified) data to derive insights and allow for fact-based decision making that results in actionable recommendations, and
  • Digital Twin Visualization: Visualization of the assets and systems that are enriched with data and insights.

Modelling & Governance

Core to the implementation of an operational digital twin is a software-based model that mirrors the real thing. That “thing” could include connected devices, equipment, production facilities, or even cities.

The most basic version of a digital twin is an asset hierarchy or asset structure. Engineers can use it to create asset definitions; configure and visualize the relationship between assets; and adjust data-access permissions.

Putting the Digital Process Twin in Context

An engineer can enrich a digital twin by using contextual data from other business systems. Contextual information can include production records, maintenance activities, engineering documentation, and even weather reports. When operational data is contextualized, engineers get a more complete picture of their process, as shown in Figure 1. For example, they can ignore shutdowns when contextual records show that the process was taken offline for maintenance or another known reason.

TrendMiner infogrpahic

Figure 1: Contextual data can include information from a variety of sources, as shown above. When used in conjunction with time-series data, contextual data gives engineers a more complete picture of their process.

Contextual information historically has been stored in separate systems maintained by different departments. As a result, the data becomes siloed. Obtaining this siloed data to contextualize a digital twin and keep it up to date requires a lot of manual labor and continued effort. However, digital twin vendors that are focused on contextualization use algorithms and machine learning (and/or artificial intelligence) to drive efficient data unification. The unified data source saves time and labor.

Furthermore, unified data becomes democratized to everyone in the company who needs to have access. Contextual items can include comments for discussion with internal experts to facilitate global collaboration and knowledge sharing. This means operational experts are empowered to do more with their process data by providing them with powerful yet robust and intuitive analytics solutions so that they can apply their knowledge for the benefit of their organization.

Applying Analytics

A digital twin with or without fully unified data, as powerful as it is, is merely the enabler. To be useful for root cause analysis or predictive maintenance in a process manufacturing setting, it requires details from the information hidden inside operational data.

Traditionally, analyzing operational data has required the help of a statistician with a strong mathematical background. Today, engineers are empowered to analyze operational data themselves with the help of advanced industrial analytics software. A second approach uses self-service advanced industrial analytics software. In this case, the business users and domain experts (typically process engineers) are empowered to derive insights themselves by using fit-for-purpose tools.

Both approaches are valid and the net value for both can be equally high. As such, it depends on the type of approach a company wants to take and their vision for the future in terms of empowerment and upskilling of their workforce. Some companies even take it a step further by combining the two approaches with collaboration between the central data teams and local teams of operational experts.

There are three main differences between creating statistical models for analysis and using advanced industrial analytics software. They are:

  1. Dynamic simulation of any process is more intensive in terms of complexity and effort compared to steady state simulation. Conversely, advanced industrial analytics software requires less effort (in terms of cost and time) mainly because the problems are not as complex.
  2. It is relatively expensive, in terms of manpower, to maintain real-time online (open loop) systems for simulation. Open loop systems also have a history of poor sustainability. Advanced industrial analytics software, by comparison, is cheaper to implement. Because domain experts develop new skills by evaluating data for themselves, advanced industrial analytics solutions also have greater sustainability.
  3. Simulation/modelling techniques are seen as black box solutions that are not adaptable to various needs. They therefore have lower adoption rates among the users. Conversely, empowering operational experts with advanced industrial analytics software drives acceptance of data-driven insights.

Digital Process Twin Visualization

Digital twin visualization covers a broad spectrum of expectations. These include:

  • An asset hierarchy/tree browser or 2D visualization of a digital twin graph model. This is used to easily explore relationships between objects and source systems.
  • A 2D digital visualization of P&IDs, possibly enhanced with 2D symbols representing equipment, up to 3D rendered CAD models that resemble the actual object’s size, shape, and location. In this type of visualization, engineers can easily explore a physical facility in a digital reality (these also are known as immersive walkthroughs).
  • Dashboards or cockpits with analytics-driven information about key process metrics, health status of the equipment (physical, economical, or sustainable excellence), and/or important process events possibly complemented with prescriptive advice to react to (upcoming) sub-optimal or abnormal process situations. Operational experts can use these to easily get insight into the actual status of real process and drive analytics-driven decision making.

Concluding Thoughts

The idea of a digital twin results in certain expectations. Today’s historical data already can be leveraged in a digital process twin for a better understanding of future performance. Creating a process twin helps engineers identify assets and their parameters quickly. They can use advanced industrial analytics software, such as the Digital Twin Manager in TrendMiner, to build asset hierarchies using a wizard-style drag-and-drop feature. Data from historians, cloud sources, dashboards, tags, knowledgebase links, and machine learning models created in MLHub become available for further analysis and deeper insights.

Using advanced industrial analytics software, companies can start small with their digital process twins and scale fast to meet their needs.

From our Blog

oil refinery

Saving Sulfur: Using Trends to Optimize Sulfur Recovery Units

Self-service data analytics places the optimization of sulfur recovery units in the hands of oil and gas process experts.
Energy revolution image
plant header

Oil & Gas and The Race to Net-Zero

, ,
Digitalization is pretty key in helping the Oil & Gas industry meet the energy needs of the not-so distant future.