Like many technologies at the peak of its hype, the industry is still determining the full potential of digital twins. Creating a digital twin of a thing, device, or enterprise asset enables businesses to enhance their visibility, gain situational awareness, and make educated business decisions.
“We’re on a steep learning curve with it [digital twin technology] now,” said Marc Halpern, Gartner vice president of research for engineering and design technologies. “We’re at the peak of hype with digital twins, where I think it’s going to happen and it is happening.”
As use cases develop for digital twins, Gartner forecasts the technology will become increasingly prominent for IoT systems and devices; business transformation initiatives, particularly digital transformations; evaluating enterprise performance and cost; improving the customer experience; manufacturing; risk management; and more.
While there are traditional software techniques and manual processes that can address these use cases, a digital twin can eliminate the more error-prone and delay-prone workflows. The primary benefit here, said Atul Mahamuni, the VP of IoT applications at Oracle, is “that the digital twin brings many separate concepts into a single interconnected concept. This allows for creation of a single integrated workflow.”
Oracle is working to establish a digital twin framework for IoT. Mahamuni said Oracle considers the following three aspects vital to their definition of a digital twin. First, it seeks to use digital twins as a virtual representation of physical assets. Second, it uses it as a predictive and behavioral model of the asset. And lastly it will integrate the insights from the former two points directly into business applications.
Software AG Digital Twin
Software AG this month launched an Enterprise Digital Twin framework built on its management and monitoring platform Aris. Its framework helps enterprises execute and evaluate business processes on a strategic level by gaining end-to-end visibility and visualization of their operations.
For Software AG, a digital twin is a “digital representation of an organization unit,” said Helge Hess, Software AG SVP of product management and marketing. This includes replications of the processes, employees, and assets. “Everything that you need to know to understand what’s going on within the company or within a specific unit of the company. When you have such a digital representation you can play with that as well.”
Software AG’s framework on Aris uses process mining and intelligence for pattern and anomaly detection to deliver intelligence on process improvement. While process monitoring isn’t a new concept, the digital twin compiles all the relevant and specific information related to the customer journey, including bottlenecks, strategies, policies to give more impactful insights, said Hess.
In addition, Gartner analyst Halpern said that with monitoring and digital twins, “it has more to do with the ergonomics of how you’re interacting with the data. A digital twin brings all of the data together in an orchestrated way so you’re going to get much better insight into how the data interacts with each together, affects each other.” That, and there’s the potential for simulation in applying performance solutions, as you can interact with the model in real-time.
Additionally, Hess said it can help enterprises execute new strategies. This includes changing the product portfolio or regional presence and entering a new market. “You can really bring that into a reality as you see, really, which aspects of the business are being affected,” said Hess. This allows enterprises to manage the supply chain and evaluate the end-to-end processes.
Other use cases for Software AG include gaining customer insights and managing risk and compliance.
The Steep Learning Curve of Digital Twins
While some organizations have begun to think about implementing digital twins, Halpern said there are still a lot of challenges before the framework is widely used. But, he added, the appeal of the technology is obvious.
“It’s appealing in the sense of keeping information very well organization and ergonomic when it comes to supporting products through a lifecycle,” he said. A digital twin holds all of the information and records regarding products and components, thus making it easier to access all the content you need, quickly and easily.
However, there are challenges. The first is that building the models is risky and difficult. If they’re created with proprietary design software, they could become unreadable in their life. Second, the analytics are hard to “get correct.” Many of these issues revolve around data, such as the quantity, quality (including gaps in data), and types of data collected.
In addition, Halpern noted that as the technology gains its foothold, artificial intelligence (AI) will begin to get embedded to create fully predictive enterprise models. However, this could be five to 10 years out. Halpern said companies have to first understand what is happening with a digital twin, begin to establish patterns in the data, and be able to perform root cause analysis on the data, so that they can begin to correlate the data, and in turn implement AI.