What are digital twins, and why should process enterprise stakeholders care?
Process industry manufacturers need to strategize how to enhance their operations to meet ever-increasing demands. This operational enhancement will require plants to develop more elaborate devices, systems, and/or processes, which in turn will require a more complex approach to measuring mission-critical performance indicators and goals.
More and more, the industry will need to use digital twin technology to create predictive analysis benchmarks of asset performance. Using an advanced type of process-model simulation, digital twins provide real-time data analytics that operators can apply in multiple ways across the entire supply chain throughout the industry.
In this article, stakeholders will learn what a digital twin is, its benefits to operational processes, and the role of a digital twin in digital transformation and smart manufacturing.
What is a digital twin?
What is a digital twin?A digital twin is a digital representation of a device, system, process, or even a person. This digital representation mirrors the actual process with full knowledge of its historical performance. By simulating assets to predict their future performance, digital twins drive agility and convergence of operational understanding.
Such enhanced understanding enables effective decision-making and helps determine strategies to maximize safety, reliability, and profitability. Digital twins capture data to determine real-time performance, which can be used across the entire life cycle of an asset for optimization and predictive maintenance.
Are digital twins a type of artificial intelligence (AI)?
While a digital twin incorporates one aspect of AI technology that allows it to improve automatically from experience, there are key differences between digital twins and contemporary AI applications such as machine learning (ML). While ML algorithms observe asset behavior patterns and correlate them with outcomes, they lack the deep knowledge of underlying physical properties, which are fundamental to digital twins. For process applications, digital twins can facilitate better and faster decisions concerning “What if?” and “What is best?” scenarios to test profit-maximizing strategies.
By organically connecting and analyzing information based on digital twins and physical data, an organization can greatly expand its ability to predict and estimate asset performance and productivity. Technologies such as AI can greatly increase the efficiency and sophistication of predictive analytics. A more predictable future augments the ability to optimize, solve problems, and perform various actions.
How does a digital twin mimic the physical world?
Digital twins replicate real-world events and actions by combining live sensor inputs from their physical counterparts with historical-performance data.
Digital twin technology relies on a first-principle model, which is a mathematical algorithm that simulates the performance of an asset; the physical process feeds input into the algorithm, which then uses that data to generate an accurate digital representation of the real-life event.
Applying the digital twin
Digital twins can emulate a considerable number of things. For example, Chevron Corporation is deploying digital-twin technology to predict and resolve maintenance issues in its oil fields to offset the millions of dollars it has to pay each year to prevent equipment breakdowns. Digital-twin trends indicate that 75% of plants implementing Industrial Internet of Things (IIoT) devices either already use digital twins or intend to do so in the near future, according to Gartner.
Statistical processing, AI, and simulation with digital twins are improving the data they collect. These technologies, when integrated with a digital twin, can quickly transform data into more accurate information, which operators can visualize. A digital representation’s ability to process enormous amounts of data into understandable formats enables better decision-making about manufacturing processes; real-time monitoring, predictability, and optimization; and predictive maintenance, while ensuring a process’s performance meets or exceeds expectations.
Digital twins allow data consumers the freedom to experiment with future scenarios. Experimentation to test operational limitations by pushing devices, systems, or processes to physical failure is a costly and potentially dangerous task. Doing so with a digital twin offers insight into the limitations of an asset without the risk of real-world damage.
The benefits of digital twins
Ideally suited for cloud-based collaboration, digital twins offer benefits across a wide range of business objectives in process industries, from plant-process optimization to value-chain optimization and beyond.
Why should process enterprises use digital twins?
Process enterprises use digital twins to meet three key objectives:
|
Improved decision-making: Data consumers use digital twins to experiment with different scenarios and assess the outcomes and impacts of each approach without any real-world risk. |
Improved safety: Process models can help identify potential safety and reliability vulnerabilities; this reduces the risk of employee injuries, environmental contamination, and damage to the facility through streamlined safety processes and improved predictive maintenance. |
Increased profitability: Data consumers can run “What if?” scenarios up and down the supply chain and manufacturing process to determine which strategies maximize profitability. |
Predictive maintenance
A process model improves predictive maintenance by identifying issues that the naked eye may not notice. By combining sensor data with historical performance, the digital twin’s process model can predict catastrophic failures long before they occur. This lead time allows the operator to schedule service when process disruptions are minimal. The results are reduced downtime, fewer service interruptions, increased service life, and significantly reduced operational expenditures.
Collaboration
The data digital twins produce is rarely of use to only one stakeholder. Used in the cloud, this data drives convergence of understanding and action across process enterprises, from engineering to operations, operations to maintenance, operations to supply chain, reservoir to facilities, and shop floor to boardroom. As a result, such data collaboration enables process enterprises to develop new business models.
New business models
Using the cloud, digital twins can engage subject matter expertise and technology from outside corporate boundaries while serving entire enterprises, rendering operational siloes extinct. In addition, they enable applications to subscribe to external data feeds and third parties to remotely offer analytical capabilities while radically expanding agility and reducing infrastructure costs.
Using this new business model to best leverage the subject matter expert domain knowledge and simulation technologies adds bidirectional value—that is, value between corporate assets, central or remote operation centers, support teams, and third parties. Speaking of which, this model also allows third-party suppliers to offer Outcome-as-a-Service. For example, to chemical processing plants, they could offer catalyst performance as a service, rather than simply a catalyst.
Digital twins and digital transformation
What is digital transformation, how does it apply to process manufacturing, and where do digital twins fit within it? At its root, and as its name suggests, digital transformation looks to leverage digital technologies to accelerate business strategy and operations to meet changing market conditions, workforce dynamics, environmental regulations, sustainability goals, geopolitical uncertainties, and other factors.
In a holistic context of process manufacturing, digital transformation focuses on digitalization efforts across the areas of asset life cycle, smart manufacturing, and value-chain optimization. It achieves this in part by exploiting emerging technologies in big data, cloud computing, control systems, automation networks, AI, IIoT, edge devices, and digital twins.
More specifically, digital transformation allows process manufacturers to use real-time data to improve plant processes, perform predictive maintenance, identify production bottlenecks, and strengthen their market position through the use of data analytics. The digital twin concept is a vital tool in the process industry’s drive toward digital transformation. How so?
Why are digital twins central to digital transformation?
Enterprises can apply digital twins throughout the course of digital transformation. Notably, process simulation often begins during the life cycle of an asset undergoing digital transformation: data from simulation can inform the foundation of asset knowledge graphs and by extension the management of asset knowledge throughout transformation. In effect, digital twins support the transition from asset design to operations with ease. The digital twin concept occupies a central position in digital transformation thanks to its ability to handle a wide range of challenges, including the following:
 |
Assisting human operations: |
 |
Consolidating business processes: |
 |
Increasing information security: |
 |
Removing barriers to innovation: |
How are digital twins and IIoT enabling smart manufacturing?
Also referred to as a converged IT/OT approach, the IIoT enables an ever-increasing assortment of cloud-enabled industrial sensors that complete the reflection of a plant operation’s vital signs. Data from sensors informs and drives performance of plant operations.
Digital twins rely on data generation from the IIoT to supply plant operators the decision support they need to perform remote diagnostics, asset management, predictive maintenance, and safety monitoring. Since digital twins rely on the IIoT, they must evolve together.
Smart manufacturing is now at the center of digital transformation in the process industry. An increasingly complex system of smart machines, plants, and operations—each with its own level of intelligent functionality—makes up smart manufacturing enterprises. And the IIoT is needed to connect each smart asset within such an enterprise.
As smart manufacturing enterprises become more complex, the responsibility falls on IIoT technologies to evolve to facilitate tighter integration between individual smart assets and the enterprise at large. To evolve, the IIoT relies on better, more cost-efficient wireless sensors and enhanced cloud connectivity. These in turn heighten digital-twin performance by streamlining real-time data collection and conversion into actionable information.
Which new technological innovations will complement digital-twin technology in the future?
Technological innovation abounds in the process industry. And most innovations are set to impact digital-twin performance in one way or another—there is no way to include all of them within the space of this article. However, here is a selection of technological innovations set to complement digital-twin technology.
Process data analytics and AI/ML with first principle–based process simulators
Advanced analytics include AI-based approaches, which involve machine learning (ML) algorithms that observe asset behavior patterns and correlate them with outcomes such as energy savings or machine failures. Such analytics can detect declines in quality or productivity in early-stage manufacturing. For process analytics, AI/ML algorithms could fall short because they lack the underlying physical properties that a chemical engineer would use. The first-principle model, which is the foundation of a digital twin, applies those physical properties. Combining that model with AI/ML analytics enables process-wide decision-making that accommodates “What if?,” “What is next?,” and “What is best?” analyses.
Robotics with process or maintenance-purpose digital twins
The process industries today can deploy a wide variety of robots, which can perform automated tasks in locations that are inaccessible or hazardous to people. In a manner similar to AR technology, robots can support smart maintenance and real-time decision-making on-site, complementing a maintenance-purpose digital twin’s data-driven decision support. In process-troubleshooting situations, which present safety risks to people, digital twins can direct robots to potential problem areas to facilitate timely remedies.
Yokogawa's digital-twin solutions
Yokogawa’s digital twin solutions offer benefits across a business’s operation, from asset management and manufacturing to supply chain solutions and data analytics. Companies relying on real-time data from operations supply chain technology and plant process applications can benefit from Yokogawa’s digital twin and model applications to meet a wide range of requirements.
What are Yokogawa's digital twin initiatives, and how can they optimize operations?
A digital twin is a digital representation of a device, system, process, or even a person. This digital representation mirrors the actual process with full knowledge of its historical performance. By simulating assets to predict their future performance, digital twins drive agility and convergence of operational understanding.
Such enhanced understanding enables effective decision-making and helps determine strategies to maximize safety, reliability, and profitability. Digital twins capture data to determine real-time performance, which can be used across the entire life cycle of an asset for optimization and predictive maintenance.
 |
Automation and AI to improve performance of advanced controls: A digital twin can function as a virtual representation of a live plant, its automation algorithms, and its process control environment, allowing engineers to perform process control tests and evaluate proposed adjustments before applying them. |
 |
Asset life cycle: Operators can use data that digital twins capture to determine real-time performance across the entire life cycle of an asset for optimization. |
 |
Capability assurance: Operators can use digital-twin simulations to capture real-time work processes, which they then can manipulate to predict key operator actions. Doing so minimizes the learning curve for new operators, who can train within the simulation. |
 |
Enterprise insight: Digital twins can set up a simulation based on existing company key performance indicators. Given a dashboard with information that a simulator-based digital twin provides, operators can use the model in real time and run multiple hypothetical scenarios or predict the future course of a business based on existing data. |
 |
Increased production and predictive maintenance: Breakdowns in any manufacturing system can result in delays along the supply chain. Digital twins make it possible to run an AI/ML model with a first principle–based process simulator to identify predictive maintenance and keep downtime to a minimum. |
 |
Plant process optimization: Operators can use digital twins to create high-fidelity models that they can use for performance monitoring, simulation, and optimization to deliver enhanced yield performance, flow assurance, energy-efficiency improvement, enhanced reliability, and operator-capability assurance. |
 |
Value-chain optimization: Understanding when and where products are in demand allows companies to adjust production and labor needs while exploiting market opportunities. Data analytics that operators derive from digital twin applications is invaluable when predicting market demand. |
 |
Instrumentation and equipment productivity: Real-time and predictive data that digital twins collect reduces the risk of equipment breakdown by improving predictive maintenance outcomes. Stakeholders can reduce operational expenditures through online monitoring and prediction of field device health. Advanced chemistry: Pumps, flowmeters, transmitters, and chemical analyzers are highly intelligent devices that provide asset performance information and live process information to process and maintenance-purpose digital twins. The digital twins inform ongoing performance optimizations and add adaptability to changing duty requirements throughout the intelligent device life cycle. |
Finally, Yokogawa has pioneered several innovative technologies that stakeholders can use as adjuncts to the digital twin applications listed above. The following are two such technologies:
AR technology with maintenance-purpose digital twins
AR technology can support smart maintenance and real-time decision-making on-site, complementing a maintenance-purpose digital twin’s data-driven decision support. In addition, AR technology can help
- improve asset maintenance efficiency and quality, reducing patrol time, maintenance cost, and production loss;
- improve mean time to recover by obtaining necessary information in real time or visually on-site; and
- accelerate decision-making by enabling management to view the necessary data in the preferred format at any time.
Web*Technician™ real-time plant-data sharing service
Web*Technician™ is a cloud service that can deliver highly secure methods of collecting, transforming, and sharing valuable data between the source and destination. It allows plant operators to collect real-time and historical equipment data from plants to detect failures via remote data monitoring before they happen. Used in conjunction with digital-twin technology, this service gives plant operators a more complete overview of plant operations while facilitating a more robust contingency plan to address critical asset failures.
Conclusion
To recap, at the heart of digital transformation, digital-twin technology adds tremendous value to enterprise operations in the following ways:
- It enhances product life cycle management by driving asset improvements through advanced data analytics, AI/ML, process simulation, and operational insight.
- It guides and optimizes day-to-day decisions with respect to an asset using in-depth, accurate data.
- It uses massive amounts of plant information to enable better decision-making and optimization of all aspects of an asset.
- It ensures that actual asset performance meets planned performance.
- It removes operational siloes when run in the cloud, and therefrom adds bidirectional value to new business models by engaging subject matter expertise and technologies from outside corporate boundaries.
- It facilitates future fully autonomous operations with constantly optimized processes, reduced maintenance costs, and minimized risks of process failures and accidents.
On the whole, digital twins improve enterprise agility, operator safety, and profitability by enabling everyone to see inside assets and processes and perceive unmeasured things.
Digital twins are a key element for implementing smart manufacturing and industrial autonomy initiatives in process plants to realize operation optimization, failure prediction of assets, and reduction of process development lead time.
To learn more about Yokogawa’s digital twin solutions, download the white paper: "Digital Twin: The Key to Effective Decision-Making."