One of the main reasons I am excited about my ongoing column ‘Data and Trending Technologies’ on TDAN.com is that I get to explore the latest technologies such as blockchain, Internet of Things, and illuminate how data plays a significant role in all these technologies. The technology I am exploring in this article is no exception to that and I am doubly excited about this technology. The term Digital Twin hasn’t hit the popular press until 2017 when Gartner named it in their top 10 strategic technologies trends list.
Since then, Digital Twin has appeared in every other major analyst’s ‘must have’ technology.
So what is a Digital Twin anyway? To put it simply, a digital twin is a virtual replica of a physical object or a process or a system for that matter across its entire life-cycle. You are probably wondering which virtual models have been in place for quite some time, whether it be a flight simulator or NASA’s models of their spacecrafts, etc. You are right but the differences this time are around the data that can be collected, analyzed, and used to either reduce costs (maintenance, proactive monitoring, etc.) or increase growth opportunities.
Sensors placed on the physical asset (a jet engine, car, a train car, etc.), collect data for the virtual model (twin) and this data enables engineers / technicians to see how the physical asset in the field is doing, and more importantly how it will do in the future. This ability to apply data analytics in all their forms (i.e. descriptive analytics, diagnostic analytics, and predictive analytics) is what is really exciting. As you can imagine, the amount of data that needs to be collected, curated, managed, and analyzed can only be dealt with latest technologies like big data and cloud computing.
The reason why digital twin is exciting for me is that my area of practice— i.e. data analytics, data management, and data governance—becomes crucial for a whole lot of companies to implement digital twins. For example, a car manufacturer or any of the myriad number of car part suppliers now need to plan for collecting data, analyzing data, and providing these analyses to the engineers for help with crucial decisions. These companies need to be smart about software and data analytics. In the process, the quality of data, data governance, and data analytics become important areas to be concerned about.
Digital Twins and Data Management:
Now let us dig into what digital twin technology requires.
First and foremost, digital twin technology requires an Asset model which describes the asset framework and components as a hierarchy. This hierarchical representation requires a solid data management framework.
Secondly, Digital Twin technology requires a knowledge base of asset data (ex. time series data), relationships of data, and a repository for ongoing insights. This knowledge base becomes a crucial foundation for a data analytics engine.
Thirdly, it requires an analytics engine or analytics hub to describe, diagnose, predict, and prescribe the behavior of the asset. This is where each asset class might have its own proprietary set of algorithms and an opportunity for the vendors that operate in this space to differentiate themselves.
One common thread to the data collected and managed in digital twin is the need for a ‘single source of truth’ regarding the asset’s information, data regarding its behavior, and other data insights. Data management professionals can quickly realize that this need for ‘single source of truth’ calls for leveraging established data management / data governance practices, and extending those practices to cover the new ground in these areas.
As you can see, all these critical components call for investments in big data, analytics, artificial intelligence/machine learning. Organizations already using big data, analytics, or AI/ML will have an advantage in kick-starting their digital twin initiatives.
Digital Twin Implementation and Use Cases
Digital twins can be implemented at various levels. Given that the recent popularity of Digital Twins is driven by Internet of Things, a Digital Twin can be made up of a component within a product or infrastructure setup (ex: a heat pump), the complete product itself (ex: an autonomous car), or of a process. In varying degrees of complexity, the digital twin of a process is much more specialized and the most complex implementation.
As with any technology, the business use cases form the foundation of value proposition for digital twins. Typical business value proposition for digital twins tend to be:
- Increasing equipment effectiveness by monitoring the effectiveness of physical assets in the field.
- Improving Operational efficiency by closely coordinating and analyzing the efficiency of all connected assets in varying regions / temperatures / environments.
- Reducing maintenance costs by predicting maintenance requirements and proactively maintaining equipment and/or intelligently maintaining / servicing least used equipment.
- Improving safety by analyzing for abnormalities and predicting unsafe issues.
Some of the industries that are investing in Digital Twin technology are:
- Production / Manufacturing: Process control industry is probably in the forefront of using digital twins to proactively service / maintain and reduce operational costs by leveraging digital twins for the far-flung equipment / assets / products
- Retail: Digital twins of retail products / services, and even entire retail stores is a very efficient way to manage retail operations. It is a well-known fact that all established franchises in the food industry have digital twins of their stores and kitchens to improve operations.
- Utilities: Utilities is another industry with remote operations and digital twins help them manage their operations and reduce overall costs
- Farming and Healthcare: Farming and Healthcare are also emerging as other industries where Digital Twins are increasingly being used to improve business operations.
Conclusion
Even though the term Digital Twins is relatively new, the technology behind it is being used for quite some time. The popularity of Internet of things made digital twins an exciting area to participate in. Interestingly, data management, data governance, and data quality play a significant role in digital twin technology since data is its underpinning and a single source of truth is crucial. Companies that are already investing big data, AI/ML, or data analytics will have a head start in digital twins. As a data management and data science professional, I couldn’t be more excited about technologies like digital twins.
