6 Common Digital Twin Applications

IoT Analytics, a provider of market insights and strategic business intelligence around IoT (Internet of Things), AI (artificial intelligence), cloud, edge, and Industry 4.0, reviewed 100 digital twin case studies with the goal of classifying these applications into distinct groups. This effort resulted in six clusters of digital twin activity:

Twins for system prediction—A digital twin geared toward predicting complex systems.
Twins for system simulation—A digital twin geared toward simulating complex system behavior.
Twins for asset interoperability—A digital twin geared toward common data formats and streamlined data extraction in complex systems.
Twins for maintenance—A digital twin geared toward assisting with maintenance-related use cases.
Twins for system visualization—A digital twin geared toward visualizing a complex system (e.g., in 3D).
Twins for product simulation—A digital twin geared toward simulating the behavior of new products (mostly during the design phase).

System prediction digital twins

According to IoT Analytics, these digital twins focus on predicting the behavior and future state of a physical system using current data and relevant records of operational history. At the core of predictive digital twins are predictive models used to predict future outcomes.

As an example of this, IoT Analytics points to Korean wind farm operator Doosan Enerbility, which deployed a digital twin to help predict power output. The digital twins incorporate data from IoT sensors on wind turbines along with weather data and a model of expected power output. The digital twin serves two main purposes:

Acting as a performance watchdog where a physics-based model of the wind turbines calculates theoretical output, taking into consideration the current weather conditions, comparing the theoretical output with the actual output and flagging any anomalies.
Acting as an output prediction tool for improved planning that predicts future power generation based on forecasted weather data.

IoT Analytics says the digital twin-based power output predictions assisted Doosan in improving wind farm revenues by increasing the output commitments the company made to South Korean energy grid operators while avoiding fines for failing to meet those commitments.

System simulation digital twins

Simulating complex systems allows engineers to test “what-if” scenarios in a large setting with many more interlinked variables playing a role than in product-level simulations, according to IoT Analytics. Examples include factory simulations prior to opening or making significant changes, rail network simulation, or traffic simulation. One of the key benefits of such simulations is to reduce costs by testing the types of assets used, key operational parameters and other important system variables prior to making the changes.

Train manufacturer Siemens Mobility provides a system simulation digital twin use case example through its work with Deutsche Bahn AG to develop 170 new high-speed trains. For these trains, IoT Analytics says Siemens developed a new car-based train control architecture that had not been deployed before. To reduce development costs, Siemens developed a digital twin that would simulate the entire train’s functionality of 40 different subsystems. The main feature being a functional simulation that used a representation of the electrical schema of the train in the simulated environment. Siemens estimates that the digital twin-based simulation reduced costs by $1 million to $8 million.

Asset interoperability digital twins

These digital twins allow for real-time extraction of data from assets along various dimensions, including asset features, characteristics, properties, statuses, parameters, measurement data and capabilities, says IoT Analytics. Although asset interoperability is an underlying goal of any digital twin project, the main idea when implementing a specific asset interoperability twin is to develop a standardized way to address assets and the ability to promptly integrate new assets to the overall system. Many asset interoperability twin initiatives align on interoperability frameworks provided by major industry consortia, such as the Digital Twin Consortium’s Interoperability framework or Platform Industrie 4.0/IDTA’s Asset Administration Shell.

As an example of this digital twin type, IoT Analytics referenced Formula One team Scuderia Ferrari’s use of digital twin technology to integrate and analyze the various sensor data points coming from a vehicle to enable diverse teams to collaborate on analyzing aerodynamics, power, vehicle dynamics and race engineering. The ability to have all data sources streamlined into a digital twin is helping the team make faster and more data-driven decisions, according to IoT Analytics.

Maintenance digital twins

IoT Analytics says digital twins catering to the maintenance life cycle are mostly geared toward assuring a system’s operational effectiveness, e.g., by assisting maintenance personnel during scheduled downtime or repair tasks by providing them with in-depth information about the physical asset or system. Maintenance digital twins also enable predictive maintenance use cases that aim to prevent asset failures altogether, thereby avoiding costly downtimes.

German utility company E.On is referenced by IoT Analytics for a project that allowed the company to shift to a more preventive, predictive and risk-based method of maintaining its assets. “As part of this project, E.On implemented a cloud-based digital twin used for better prediction of asset failures. The digital twin allows the company to evaluate failure modes per asset type and determine the remaining life of the equipment fleet,” reports IoT Analytics.

System visualization digital twins

Among the most common types of digital twin visualizations are 3D visual elements—often using CAD (computer aided design) drawings—to help create better transparency, e.g., about the current operating conditions of the system, says IoT Analytics.

Ferrovie dello Stato, the national railway operator of Italy, built a digital twin of its rail infrastructure that included more than 10,000 miles of tracks, stations, tunnels, bridges, signals, switches, electrification hardware and IT systems for the purposes of coordinating each system.

According to IoT Analytics, “The digital twin sourced the data from sensors such as cameras and GPS receivers and employed advanced learning algorithms to create an interactive 3D replica of the rail system. ArcGIS was used for 3D visualization of geographical elements, creating a complete visual inventory of the railway infrastructure. The resulting digital twin helps railway managers in the control room visualize infrastructure such as bridges and stations remotely throughout the rail network. It also helps the local staff prevent safety hazards and boost on-time performance.”

Product simulation digital twins

IoT Analytics notes that digital twins are playing a key role in developing new and improved products by eliminating the need to build costly prototypes and allowing for quick testing of thousands (in some cases millions) of product variances. “Product simulation digital twins often reside in or are closely connected to computer-aided engineering software and sometimes CAD software,” says IoT Analytics. “Typical types of simulations include fluid dynamics, mechanical performance or compatibility of electromagnetics.”

German packaging and bottling machine manufacturer Krones was referenced by IoT Analytics for its use of a digital twin to test new product designs for one of its automated beverage packaging systems. “The company wanted to incorporate a dynamic tripod robot into the product design,” says IoT Analytics. “To do so, the friction and force of gravity on sliding packages across a moving conveyor belt and many other similar factors had to be tested. Krones developed a model for the tripod robot by importing geometry and inertia data from STEP files exported from CAD software. The team tested many scenarios, e.g., injecting faults such as extremely high levels of friction during the simulations to analyze the system behavior during such an event. By modeling and simulating the tripod robot, Krones was able to increase the performance efficiency of the robot, product development time was shortened and testing time was significantly reduced.”