Technology on the factory floor is providing innovative opportunities to improve operations — and managers and owners need to be aware of what the future holds.

One emerging innovation that may soon spread widely is digital twins —a transformative technology that bridges gaps between the physical and digital worlds.

Manufacturers continually face operational challenges, and replacing costly equipment, building new lines, or duplicating the most experienced and knowledgeable employees is likely impossible.

But firms can create digital replications that simulate, predict, and optimize operations.

Manufacturers who have not gone digital yet may wonder what a digital twin is. In simple terms, it is a digital copy of a physical asset, process, or product with a defined purpose driven by data.

Digital twins are a valuable tool to help improve safety, quality, throughput, and savings.

On factory floors, skilled labor is often the most valuable asset. Workers are the artists of manufacturing organizations, blending raw materials, experience, and intuition to produce high-value products. But if a valued worker is suddenly gone, or a manufacturer’s workforce can’t keep up with demand, an operation could collapse.

Digital twins help capture and retain an organization’s knowledge and can then assist in upskilling and training a growing workforce. Through technology, product quality stays high, as does retained knowledge.

MxD, the Digital Manufacturing and Cybersecurity Institute,^[a] has been developing guidelines that help manufacturers establish building blocks for digital twins that can be layered atop existing systems.

Initial steps include:

• Defining the objectives of what needs to be accomplished in a firm, such as improving quality and safety in an operation;
• Conducting an assessment of current systems, with a review of existing available data;
• Identifying key stakeholders, such as the operator and controls engineer of an existing system;
• And defining what the digital twin will replicate.

As an independent and neutral storehouse of resources, MxD has guidelines that have been established after working with clients, and through demonstration examples on our testbed factory floor in Illinois Factory Floor in Illinois.

MxD operates a Cyber Process Testbed, and we recently used that testbed — which includes a clean-water loop and a contaminated-water loop — to evaluate how to establish a digital twin on an existing system.

In the testbed’s original state, water pumps were controlled only through relay switches but were not capturing data or integrating it into operations. For the digital twin use case to improve preventative maintenance, we added sensors to generate data. This allowed us to feed data into our digital twin model. From there we were able to see the digital twin providing feedback to the operator and displaying the pump’s behavior in a variety of conditions.

We also needed to develop a method for data acquisition, settling on using locally available communication protocols to transfer and move data from machine to the digital twin.

By working with operators, we clearly identified use cases for digital twins.

Undertaking this exercise provided valuable lessons that can be applied to other manufacturers.

Digital twins can be deployed even in challenging environments. MxD often works with manufacturers involved in Department of Defense contracting, with stringent security needs. In these settings, a digital twin must be “air-gapped,” meaning it cannot have any connection to the internet. Our work has shown that strategies are available to develop digital twin applications fully on-premises, with resources being built and contained within the infrastructure of the manufacturer.

Manufacturers exploring digital twins as part of their problem-solving face some initial decisions, such as whether to start with an in-house approach or retain a commercial vendor. Commercial providers can offer off-the-shelf, prepackaged solutions that get plugged into an existing system, or they may create a digital twin for the operation.

Digital twinning is not a short-term solution, and a commitment of resources is required. Manufacturers need an infrastructure to embrace digital solutions for them to be effective.

A phased approach can guide maturation of a company’s digital adoption plan:

Phase 1 involves identifying and assessing an existing manufacturing system.

Phase 2 comprises an analysis of challenges and needs and trying to identify some early and intermediate wins — like improving safety and getting ahead of predictive maintenance. This phase also involves establishing baselines in processes using data. With those baselines set, digital twinning will then yield data that can help operators identify abnormalities.

Phase 3 is when manufacturers start piloting some digital twin projects, analyzing data and operations.

So, is it certain that manufacturers will see a return on their investment if they take these digital twin steps?

The answer is almost certainly yes. Company leaders don’t know everything they need to know about their processes. They are not in the weeds of their operations. Digital twins allow managers and leaders to be aligned with workers, improving insights.

Digital twins become one part of a firm’s digital thread. Technology is providing data from everywhere, and all those threads must be connected for optimal safety, quality, and profit. Digital twins are a good way to get started thinking about digital thread.

Our manufacturing world is never stagnant. Advances in artificial intelligence, machine learning, and 5G connectivity will enhance the capabilities of digital twins, making them more powerful and accessible.

Digital twins come in all shapes, sizes, and capabilities. Manufacturers differ, and every digital twin is tailored to an operation, product, staffing resource, or business needs.

The digital twin is purpose-built and will help make decisions backed with data.

In the near future, adoption of digital twins in small and medium-sized enterprises is sure to grow as costs decrease.