Technical Article

Digital Twinning and its Use in SCADA Systems

June 28, 2020 by Anish Devasia

Learn all about Digital Twinning, and how this new technique can change the way we do a number of industrial operations!

Digital twinning refers to dynamic virtual copies of a live physical system. Unlike simulations that run a hypothetical scenario, digital twins are fed with real-time data of the physical entity they replicate. 

Digital twins can be used to radically improve industrial operations by offering avenues to reduce cost, monitor assets, improve maintenance, reduce downtime, run simulations, and predict future conditions using machine learning techniques. Though digital twinning can be used in various scenarios, it is predominantly used within IIoT applications related to engineering and manufacturing. 


Digital Twinning Implementation

Digital twins can be implemented at various levels,  including part twin, product twin, process twin, or system twins. As the names indicate, the digital twin can be just a part or component of a complete machine or product, a process including all the machines and parts involved in the process, or the whole system or plant.

Digital twins use 3D models, physical rules, mathematical calculations, historical data, and direct and  indirect data to analyze present conditions. They can then apply machine learning techniques to predict future events and detect anomalies. A use case of digital twinning would provide a better understanding of the concept.

Digital Twinning in the Air and the Cloud

Consider an aircraft designed by Airbus, Boeing, or Embraer with the same Rolls-Royce engine operating across multiple geographies. 


Figure 1. Jet engines may be the same model, but those models may have different characteristics which lead to different operation and maintenance.


As a manufacturer of jet engines, Rolls-Royce gives a schedule for engine maintenance without taking into consideration the varying running conditions the engines are operating in. But the same engine model carries different fuselage. From Airbus, Boeing, or Embraer, engines operate in different geographical conditions, have a varying frequency of operations, and work with drastically different operational loads.

Now consider this: Rolls-Royce has real-time data of all its engines operating in the world. A software model of every one of the engines running virtually on a cloud computing platform with the actual data from the physical engines in operation. Rolls-Royce would be able to track and analyze data for the same engine with different operating conditions at the same time. 

With this data, Rolls-Royce can determine maintenance schedules for each of its engines considering the individual operating conditions each engine experiences. Time the aircraft is in maintenance is time the aircraft is not in the sky. With predictive maintenance, aircraft can experience less downtime.

Digital twinning also helps with product improvement. Let's say that with the data they collect, Rolls Royce observes more wear with engines connected to the Airbus fuselage. This information can be passed to Airbus, who can help in a fuselage design better for engine performance. 

Another case could be regular wear at a specific area of shaft or turbines for engines mounted on aircraft operating in the Arctic region. This can be used for the design of improved materials in subsequent engines manufactured and provide remedies for engines currently operating in the Arctic. 

Not only does digital twinning improve the design of engines, it can also be used to test new design changes in real-world scenarios before implementing them.

Digital twinning is not a new concept. The concept of having identical virtual copies of physical objects was developed by NASA as a means to access a physical asset that is very limited, difficult, and costly. The digital models developed by NASA were used to operate, analyze, and simulate scenarios while the physical object is miles away from the earth’s atmosphere. This is a perfect use case for digital twins.


Digital Twinning and SCADA

Much like SCADA emerged from the technological developments in communication networks and microprocessors, digital twinning emerged from the advancements in IoT and machine learning algorithms. Digital twinning is the logical progression of industrial automation after SCADA systems.

SCADA is a rule-based solution for supervisory control. The data acquired by a SCADA system can be used for explaining a fault after the fault occurs. However, SCADA cannot predict faults occurring and prevent them from happening or detect anomalies by design. 

With digital twinning and unsupervised machine learning techniques, abnormalities can be easily identified and precautions can be taken. Since digital twinning tracks the data from all the installations of the same machine across the globe, abnormality experience at one site can be factored into all machines of the same model across the globe.

Implementing digital twins with industrial plants is a challenge. For deployment of digital twins, the physical blueprint of all machines and parts is required. In reality, the blueprints are not up-to-date at most sites. Another factor is that most plants rely on multiple OEMs and vendors for various requirements. Digital twins have not yet become an industry standard and, until they do,  the benefits of digital twinning will be limited to the OEMs that seriously pursue this technology. While SCADA is still sufficient for most industrial operations, new industrial plants with a long-term outlook are the best candidates for deploying digital twins.


Benefits of Digital Twinning with SCADA

There are benefits that come from utilizing digital twinning with SCADA. Let’s look at what those benefits are and how they look within the system. 


Reduce Unscheduled Downtime 

Unscheduled maintenance costs valuable time, resources, and capital for the machine/plant. With digital twins recording and analyzing data, we would be able to predict a potential fault with the machine, determine different alternatives that can be taken, and choose the most viable mode taking into consideration all the technical and business outcomes.


Model Development and Optimization 

As digital twins are fed with real-time sensor data, models of improved machinery and equipment can be tested with real-world scenarios in simulated environments. Digital twin data can also be used to improve a live machine, test new settings, and delivery of software updates. 

The latest technologies like augmented reality and virtual reality can be incorporated with the digital twin to increase access for inspection to provide a visual aid for on-ground professionals, access from a remote location, and more.


Figure 2. AR/VR can be incorporated into digital twinning to allow for visual aids and remote access and control.


Access to Large Scale Data

As seen in the aircraft example above, Airbus can benefit from the large scale data that Rolls Royce has with the engines in operation across all manufacturers. With SCADA, the benefit comes from the data that can be obtained at one point of operation. With the help of digital twinning, the manufacturer can benefit from the data from all of the plants where the same equipment is in operation. This brings about a huge advantage for the optimization of plant operations.

Digital twinning is the next step in the evolution of industrial process automation. The question is how soon the technology will be democratized to a point where it can be in operation across all industrial operations across the globe to achieve the scale it requires.