Achieving the Value of Digital Twins
I have to imagine that everyone that works in an industrial setting has heard about digital twins at this point. Perhaps they’ve heard more than they care to, and would rather see more digital twins in practice. While the promise is clear, and we see a growing number of case studies proving the benefits of the approach, many companies struggle to create and maintain digital twins. So, are they achievable, from a practical perspective? And if so, what can companies do to ensure their success, particularly smaller companies that don’t have the resources that some of their bigger competitors have?
I recently moderated a panel to investigate these questions. I shared perspectives from my research and was joined by two experienced industry leaders who shared lessons learned from working with manufacturers on their digital twin efforts. My co-presenters were:
- Prashanth (Prash) Mysore, Global Strategic Business Development Director, Dassault Systèmes DELMIA
- Fabien Roger, Manufacturing Industries Sector, Dassault Systèmes
Here’s a summary of our conversation, please feel free to watch the replay of “Virtual Twin for All. Is it Attainable?” to hear for yourself, including some great questions from our audience.
Defining the Digital Twin
Let’s start with some definitions to make sure we have a common understanding of what we talked about.
Tech-Clarity defines the digital twin as “a virtual model of a physical item”. The model represents a specific product, configuration, piece of equipment, plant, city, or other physical asset with enough fidelity to predict, validate, and optimize performance and behavior. It’s connected and kept in sync with its physical twin over its lifecycle to collect, aggregate, and analyze actual field data to monitor performance, gain intelligence, and close the loop between designs and the real world.
Dassault Systèmes has a unique perspective on the digital twin, the “virtual twin.” Fabien Roger, who has worked with leading companies such as Boeing on digital twins, explains the difference. “We talk about the virtual twin because we want to extend the usage of the digital twin not only to past data but also to project future possible scenarios on this data. That’s why we call it virtual twin.” Core to this definition is the ability to simulate and evaluate different scenarios to select the best potential future outcome.
Prashant Mysore, who has almost 25 years of experience working in a variety of countries with customers from various industries, further clarified the point. “Virtual twins are not static models, but rather a responsive and connected system between the physical and the digital world. It’s an executable system, and it has a closed-loop connection between physical and digital systems.”
Our views are complementary, and the panel strongly agreed that the purpose of the digital twin is to solve business problems. Digital twins should not focus exclusively on technology but on providing tangible business value.
Digital Twin Value is Available, and Varied
The panel shared examples on the use of digital twins to provide value in companies ranging from an international fragrance retailer, a construction equipment company, a company that develops data centers, a steel company, and a line builder focused on commissioning industrial robots. One compelling story details how an automotive OEM uses virtual twins to anticipate costs based on knowing the cost structure of raw materials inside their product. Fabien describes how Dassault Systèmes helped the manufacturer create a virtual twin that extends their engineering data with procurement and supply chain information in order to monitor and anticipate predictive costs.
I encourage you to listen to the webinar replay to hear more about these examples.
What Challenges do Companies Face Adopting Digital Twins?
But not all companies have been as successful yet. We polled the audience and confirmed what we all believed, that we are in the early days of digital twins. The poll showed that some companies are gaining (and expanding) value from digital twins, but that just about one-half of companies participating in the webinar are still “experimenting and learning.”
Some of the typical challenges we’ve encountered include not having a 3D model to start with or having connectivity issues with remote equipment. Prashant pointed out that it’s a broader issue. He points out that some companies have issues around data availability, ant that it’s not necessarily 3D data, depending on the type of digital twin and the problem they’re trying to solve. He points out the need for both structured and unstructured data, which is likely scattered, and the lack of real-time data available from sensors and machines, and there is data at all. Fabien also pointed out data challenges, explaining that most of the time data is not available or it may be in a spreadsheet or in several different types of information systems. This makes it very difficult to gather, consolidate, and aggregate information for digital twins.
One of the other key areas that companies face challenges with digital twins was highlighted by another poll question. Many webinar participants reported that “Having the right people and skills” was a challenge. Many participants also reported a related challenge, “keeping the digital twin up to date.” Maintaining digital twins can be as big, or bigger of a challenge than creating them in the first place. Lastly, from a technology perspective, we pointed out that companies often struggle with the technology side not because there’s too little technology, but maybe because there’s too much to choose from (and integrate, if it’s not a platform).
Advice from the Experts
Given the value available, and considering the challenges, what advice did the panel offer? Here are some final words of advice from the webinar:
- Be business driven
- Have well-defined and clear objectives and goals
- Collaborate with your extended enterprise and with your stakeholders
- Use 3D scanning to help digitalize existing assets
- Build smart technology systems, invest in sensors
- Bring data from different sources, including IoT data, into a data lake or a data historian
- Contextualize your IT and OT data
- Invest in platform technology
- Start small and scale fast
- Take a step approach and get your first result quickly
The last piece of advice is to get help from experienced resources. Fabien introduced Dassault Systèmes Value-Driven Engagements model. With this approach, Dassault Systèmes does more than supply technology. They develop and maintain virtual twins for their customers as a service. He described how they apply both technology and industry expertise to help develop virtual twins and then train company resources to use them. This is an innovative model that can help companies get started and reduce the risk of project failure.
Overall, it’s time to move beyond learning mode and start leveraging the digital twin to drive better business outcomes. Please take the time to watch the webinar replay to learn more from research, case studies, and a very valuable question-and-answer session.