Gartner analyst Marc Halpern looked a bit worried when talking about digital twins during this year’s PDT Europe conference in Gothenburg, Sweden. He voiced concerns regarding the turn things have taken.
“It really scares the heck out of me,” he said, pointing to the seemingly unfazed attitude surrounding this complex concept in the wake of the hype of recent years.
“There’s a naïveté about the possibility of bringing together digital twin concepts in terms of cost and time," Halpern claimed. Although he was positive about the basic content and thought structure, he warned the audience, “It will take longer and will be more resource-consuming than anyone can imagine to get these solutions in place.”
It is not impossible, he added, but there are a couple of crucial factors to consider. Which are they? Halpern discussed eight success points under the headline, “Busting the Myth of Digital Twins and Planning Them Realistically”.
Marc Halpern is the spitting image of the genius professor. The way he verbalizes complex structures in combination with his depth of scientific knowledge and extensive real-life experience gives him an air of credibility. What he says counts.
Transformation is on its Way, But Few Tools are Ready
It’s not surprising that both Halpern and Gartner Group are widely cited when the world’s biggest companies consider how to build and support their PLM strategies. The firm is considered trustworthy because they are largely independent of the PLM vendors. Gartner has a large clientele representing industrial producing enterprises and organizations; enough to keep their opinions objective.
Right now, much of the discussion in both large and small company environments circles around IoT, digital twins and smart, “intelligent” solutions.
A transformation is on its way.
The commercial PLM developers all claim to be well on their way towards a digital twin. All of the major players, including Dassault Systèmes, Siemens PLM, PTC, Aras, SAP, and IBM are working to create tools and digital threads as backbones that can keep it all together. It is not an easy road.
The customers who are using PLM solutions from these developers—which encompass practically all major OEMs, as well as other organizations such as those in the defense industry—are currently looking at how to "attack" issues surrounding the IoT and digital twin.
The Digital Twin Underpins Much of the Future of Products
The development of these solutions relates strongly to "smart" products. On the one hand, the physical products themselves are becoming more complex, including sensors, electronics and software that controls their operation and their connection to the Internet. On the other hand, these products are often transformed from being sold as products to being sold as services (Product-as-a-Service). This means, for example, that an OEM that manufactures trucks is evaluating how to sell transport services rather than trucks. Or engine manufacturers are providing “power by the hour,” rather than aircraft engines.
One effect of this trend is that ownership remains with the producers, who are increasingly responsible for larger parts of product life cycles. Producers are no longer only expected to develop and manufacture products, cars, aircraft, construction machines or the like, but also must ensure—either themselves, or through partners—that the products being sold are working effectively for end users.
This, in turn, increases the requirement that these products—besides being ever more functionally sophisticated—also be manufactured with higher quality materials, and are robust enough to last longer in the hands of the end users without breaking down. With this shift in responsibility, reliability now becomes the responsibility of the product manufacturer.
This often means that the products will be equipped with communicative components, such as sensors, antennas, electronics and software, that enable the products to be subject to predictive maintenance. Machines, cars, tires, airplane engines and more will be able to "know" when parts need to be replaced—before a break-down incurs expensive downtime.
WHAT ARE THE PLANS FOR DIGITAL TWINS? In your IoT solution implementation, is your organization also adopting digital twins? Gartner asked this question in a 2017 survey. According to the results, the majority of organizations are using or plan to use digital twins in the next year. Here’s the breakdown:
- 24% Already using digital twins
- 24% Not using, but planning to use in the next year
- 19% Not using, but planning to use in the next three years
- 7% Not using, but planning to use in four or more years
- 20% Not planning to use digital twins
- 8% Not familiar with this technology
Staggering Opportunities, But Still “More of a Vision”
The scenario is mind-boggling because the possibilities of IoT, digital twins and other technologies seem endless, both technically and commercially.
Unfortunately, all of these opportunities are linked to efficient product development platforms and processes, which in all cases require rational solutions—and therein lies the challenge. These tools are generally still in the early stages, and need further development.
Marc Halpern specifically addressed digital twins in his presentation during PDT Europe. He warned of excessive optimism regarding digital twins because, "there are no simple solutions; neither in terms of product development nor the product or service in the hands of the end user."
"There are a number of myths," asserted Halpern. "Certainly, digital twins are a tickling story. But so far, the concept is more about vision and promises than about finished solutions. In short, the risks of digital twin failures are great.”
What is a Digital Twin—and What is it Not?
To highlight the problem, Halpern defined what a digital twin is, and also what it is not:
- Not just a 3D model
- Can check the "thing" that it is a representation of
- Is a virtual model of a "thing"
- Each physical thing has at least one unique twin
- Has an identity and context in which it should work
- Monitors, requests state/status and receives status information
- Can simulate the real physical case
- Includes analysis, regulation, prediction and algorithm solutions
Overall, Halpern claimed, "It is extremely complex to get all this together into a working whole, and I find that there is often what can be described as a naiveté about the opportunities for cost and time to achieve this."
He also noted that these concerns were only about the "thing" in operation. There are also a number of other critical aspects, not least about the digital twin in product development, systems engineering and manufacturing, where the twins are to interact with other twins, such as the digital twin of the production line.
The Hazards of Proprietary Formats
Despite his warnings, Halpern is positive and believes in the digital twin’s future. However, it will take considerably longer than we think to get this trend rolling.
Halpern said that if you consider the consequences of the following points, there are good opportunities to succeed in these projects:
- Product data in formats that lock-in the digital twin raise questions related to standards. “Proprietary formats are dangerous,” Halpern said, and was very clear about the traps for product developers. "Almost daily I see problems that relate to lock-in effects among my clients, and it scares the heck out of me. Think of standards in this."
- What is the life expectancy of the technology and models?
- Who owns product data and how does it relate to the safety aspects of the digital twin?
- Unclear value with regard to aggregated data. Halpern advises clients to ask, “Is it the correct data and the right analysis that can be done?”
- The digital twin is always developed by someone with certain knowledge and skills; how can you keep and “store” the information that contributed to the creation of the model? “What was the thought chain around the solution? Can it be documented to be reused if the person who has all this in his or her mind quits?” Halpern asked.
- Regarding exchange of data with others, he says, “Ensure that it is feasible, with as few limitations as possible.”
- “Keep in mind that the digital twin represents increasing intellectual capital as the years go by, as more information is added to it. How can this be secured?”
- Start your digital twin project with a minimum level of complexity necessary to provide a useful level of capability
In summary, it seems that Halpern’s advice to clients is:
a) Be careful when evaluating the claims of PLM vendors when they speak about the digital twin
b) Understand that while the future opportunities are exciting, they are almost certainly further out on the time horizon than you think
c) The inherent complexities of a digital twin are difficult to comprehend, particularly when you consider the twin in product development, systems engineering, manufacturing and in use
d) The data within the twins will become ever more valuable over time, so make sure that you understand who owns that data
e) Interoperability isn’t sexy, but it is necessary to ensure continuous access to all of your data, so beware proprietary formats.