A challenging part of working in the Information Technology (IT) field is choosing the right assets for a specific system and getting the desired outcome. This ranges from hardware and software to virtual appliances within the cloud. With every product, there are multiple vendors to choose from, all claiming the best performance on the market. Once an engineer decides on the components and platform, the technology could be obsolete within a couple of years. So, how can IT professionals determine the best course of action using minimal funding with so many options out there?
One option is digital twins. No, this is not some YouTube channel with identical twins vlogging. At its core, digital twins use 3D technologies and sensor data ingestion to recreate a “twin” of an existing infrastructure or physical asset. The digital twin processes sensor data, allowing it to track an investment in real-time, predict its behavior, and perform multiple test simulations. The IT professional can then see any inefficiencies in the system and anticipate any issues that may come later based on certain decisions. Further, assets in the digital version can be swapped out to see if they provide any advantage over the current arrangement. The capabilities of digital twins can offer many benefits to both the private and Government sectors.
Digital Twins Breakdown
So, if digital twins could provide a possible future outcome without committing to purchasing new equipment, why aren’t more businesses doing it? Unfortunately, digital twins are not as easy as downloading an Augmented Reality (AR) app on your phone and aiming it at a device. Many components beyond 3D modeling make digital twins provide valuable data. The figure below displays the high-level overview of elements that make digital twins work.
These elements play a role in how digital twins replicate the physical item, process and store the data from sensors, integrate security, and produce a digital rendition of the system with predicted outcomes.
Sensors are used with physical assets or systems to collect and convert data into a digital representation that 3D software can interpret. Some information that can be collected includes temperature, humidity, motion, location, or vibration. Any calibration data developed in the digital environment can then be transferred back to the physical asset (if applicable).
3D modeling software copies a physical asset to a digital version (twin). Data regarding the asset is considered when creating the model and allows personnel to see how the asset operates in different scenarios.
The digital platform is where engineers interact with the digital twin. The interactions in the digital platform give insights into how the asset will act in certain situations and can even predict issues in the future. Another benefit to this feature is the digital twin will run scenarios digitally instead of a company purchasing the physical equipment and then having to test for compatibility.
Artificial Intelligence (AI)
AI is frequently used for everything from creating realistic pictures of events that never happened to browser applications that hold conversations with people. For digital twins, AI is used to analyze the data from the 3D model. AI performs real-time decision-making, event forecasting, trend detection, and anomaly identification, giving the engineer metrics that can be applied to a company’s system.
The cloud infrastructure portion of a digital twin is used to help store all the data collected. This data can be accessed from anywhere, at any time, by engineers on the project. Digital twins leverage the scalability of cloud technologies as data requirements from system to system will vary.
As one would imagine, data within the 3D model is very sensitive and will require a robust security setup to protect the information. It is recommended that the digital twin be protected just as securely as the company’s production system.
Functions of the Digital Twin
As discussed earlier, the digital twin provides a digital version of your system or environment and performs real-time analysis of data from sensors to determine outcomes. All these steps can be broken down into five main functions.
Monitoring—The digital twin monitors the physical item it was copied from, gathering real-time data for sensors associated with the component. Data monitored include the environment the item sits in, performance metrics, and other characteristics set up with the sensors. Monitoring lets IT professionals know what is happening with the physical asset.
Analyzing—The digital twin uses Machine Learning to determine possible scenarios for the component and infrastructure attached to it, any patterns a human analyst would not notice, and ways to increase the element’s efficiency within an infrastructure. The data being analyzed is pulled from the sensors on the physical device.
Forecasting—With the data analyzed, digital twins can provide possible scenarios for the monitored component and infrastructure. These scenarios could be failure points, the resources required to start a new system, or finding ways to make the current system more efficient. The forecasting portion makes a digital twin so valuable to a company. The twin provides a probable outcome without investing large amounts of funding and human resources. Further, forecasting can be used to decide whether to continue or start a specific project path.
Recommendations—The digital twin’s recommendations are tied to the forecasting functionality. Once the digital twin determines an outcome(s), the scenarios can be used to decide a direction for the business or Government entity. The suggestions can start the groundwork for a new project or make the current process more efficient.
Enablement—Digital twins can enable decision-makers within an organization. Data and reliable forecasting will make decisions quicker and more successful. Instead of using resources to test a situation on a smaller scale (e.g., buying and setting up a new server dedicated to a specific function), a digital twin can be created to see if the server would be appropriate and what would be needed to give the most value to the organization.
These five features are what make digital twins so valuable. Previously, an entity would need to have a location, purchase the equipment, hire the personnel to run and maintain the systems, and troubleshoot after the fact if any issues arise. Digital twins allow all these steps to occur virtually, providing possible outcomes and leading to more reliable decision-making.
Digital Twins and the Government
If there is one group that could benefit from digital twins, it is Government entities. According to the United States Government Manual, the Government comprises roughly 96 independent executive units and 220 components of the executive departments. These numbers are not even all-inclusive, as there is no complete list of every organization within the Government. Within these groups are thousands of teams with their projects and priorities, all needing financial, human resources, and physical equipment. It is estimated that only 24 percent of Government agencies are experimenting with digital twin technologies.
One organization taking advantage of digital twin technology is the Navy. The Navy is currently creating digital twins of their full shipyards. This includes the buildings, infrastructure (IT-based and structural), personnel, and ongoing projects at the shipyard. A complete digital view of every aspect of the shipyard helps the Navy see inefficiencies at any level of operations. Do they need more personnel here? Would another power source be beneficial? Are there any slowdowns with their communication technologies? All these questions can be answered with the digital twin and guided in a modern environment of agile decision-making without additional funding.
Another possibility is for the local Government (city-level) to make a digital twin of their city. The data gathered from this could include traffic flow, dangerous intersections, quality of life for those who bike or walk and forecast any potential industrial or urban development issues. Some cities use digital twins to determine sustainability efforts and any effect population growth will have on them. There are many benefits to forecasting a future outcome before committing resources to a project.
Digital twins can even be used to predict the future of Earth based on environmental issues plaguing our planet. The National Oceanic and Atmospheric Administration (NOAA) uses digital twins to do just that. NOAA uses thousands of land- and air-based sensors worldwide to gather data. The massive amount of data is then fed into the digital twin for simulations and outcomes to be predicted. When the question of what will happen to the environment if we cut greenhouse gas emissions by 22.5 percent comes up, scientists can run the simulation and give an accurate answer backed by data.
Will Digital Twins Take Over for Personnel?
Any new technology that assists decision-making will always be perceived as threatening people’s careers. Digital twins should not be one to be worried about, though. Incorporating digital twins will require more specialized personnel to help set up the technology and ensure it is maintained for accurate results. At the same time, the rest of the team gathers data to plug in and document the outcomes for end-user consumption. Digital twins should be used as a tool like Jira or AutoCAD, where personnel use it to provide better results for the end goal, not as a threat to their career.
The Future of Digital Twins
As mentioned earlier, only 24 percent of Government agencies are using or beginning to use digital twins for their projects. With budgets not increasing—if anything, they are decreasing—and the expectation of more work being accomplished, digital twin technology seems like the future route the Government may need to take. There will be an upfront cost for the technology and training; however, the long-term results could save the Government millions of dollars and make projects more efficient. No longer would the Government have to make an educated guess regarding an outcome from previous lessons learned; they could plug in the data and get an accurate prediction, allowing for more agile movement to complete a task.