How Digital Twin technology is bridging the manufacturing talent gap

This article was first published in SupplyChainBrain

Previously, digital twins were created using spreadsheets to input data and see potential outcomes of various scenarios. Thanks to advances in artificial intelligence and virtual reality, digital twin technology has progressed with faster, more accurate projections, based on endless streams of data that can be accessed from anywhere in the world. It’s essentially creating a “metaverse” for manufacturing companies.

By using a direct feed of data and applying machine learning, digital twin technology allows a manufacturing plant to utilize the data in real time to become smarter and more efficient. It does this by identifying gaps in the manufacturing process, predicting the timing of maintenance or fault-production issues with increasing accuracy as it gathers information. Managing these issues early on or before they happen prevents downtime and productivity losses.

As the manufacturing sector grapples with a shortage of skilled labor, digital twin technology is helping leaders maintain a productive, safe factory floor despite resource limitations.

Supporting lean operations

With experienced manufacturing workers retiring and few skilled workers available to replace them, the U.S. industrial sector has been facing a skills shortage for several years. The Great Resignation, fueled by the COVID-19 pandemic, has only added to the challenge.

As manufacturers struggle to attract, onboard and train new talent, digital twin technology is serving as a lifeline to many manufacturers. Companies are using it to test different equipment setups to find the most productive options and prevent downtime. In addition, automation is allowing many manufacturing plants to continue production despite reductions in available labor. A single machine operator can run multiple machines autonomously, as the digital twin analyzes data to see how different inputs and processes will impact productivity.

Digital twin technology is especially useful in the prototyping stage of manufacturing, where teams are developing multiple iterations of a product to determine which design and materials will produce the best final piece. Digital twins can streamline the prototyping process to improve time to market while reducing costs. Digital twin prototypes are digital renderings that allow design teams to see how different models will perform in the real world, without devoting materials and production time to multiple possible iterations. The protype can contain information pertaining to the physical attributes, properties, operating parameters, bill of materials, part numbers and more. With a digital prototype, products can be built much earlier on in the process, and simulations can be run using virtual reality. Changes can be made quickly to improve designs before the physical machines are built.

If traditional time to market requires two to three prototyping cycles, the use of a digital twin in the early- to mid-stages can eliminate one whole prototyping cycle, reducing time, resources and cost.

Optimizing workplace conditions

By factoring in how many employees will be working at a given time, as well as their skill levels and equipment needs, a digital twin can simulate how productivity will be impacted by moving certain employees to different areas. It can balance the workflow and predict which combinations of workers and equipment will be most productive.

A digital twin can also predict failure based on factors such as age of equipment, materials used, maintenance records and other data points, allowing workers to prioritize preventive maintenance to avoid unnecessary downtime.

In the event of an equipment breakdown, digital twins enable virtual troubleshooting, allowing staff to investigate causes of failure and possible resolutions. This reduces the amount of unnecessary time and labor it would take to diagnose the problem and find a solution manually. Leaders can then enlist the help of the correct specialist to fix the problem, rather than going through each potential equipment issue. Because the digital twin runs on a high-powered computer, analysis and simulations can be done in minutes rather than hours or even days.

Steps for implementation

Once senior leaders are on board with the potential gains of digital twin technology, manufacturing leaders can move forward with the implementation process. As the new technology is introduced, it’s important to take these key steps:

• Establish your goals. Before beginning the technical process of creating a digital twin, establish what the digital twin will be doing. What business-related problem is it meant to solve? Will it be monitoring equipment to report failure? Predicting failure and recommending preventive maintenance? Or running simulations to see how different activities will impact productivity? Digital twin data can help the organization solve the issues it needs to, such as predicting preventive maintenance, but can also indicate other areas of improvement as the digital twin data continues to get smarter and more data is collected. There’s also the question of what needs to be monitored. Do you need a digital twin for one machine or production line, or a larger one of the entire plant? While it might be tempting to monitor every aspect of the plant, focus on areas that will bring the greatest return on investment. Digital twins can always be expanded later on.
• Find the right software. For manufacturers, choosing a digital twin creator will be like selecting any other tool in the shop — think about the application, the goals and what you’ll need to achieve them. If you only plan to monitor equipment for operational efficiency, a limited digital twin should suffice, while more advanced technology will be needed to run simulations.
• Import CAD models and data. This is where things really start coming together. Importing computer-aided design (CAD) models and other streams of data into the software will allow you to build out the digital twin. Depending on how complex the digital twin will be (monitoring versus simulations, single machine versus full operating plant), the data needed to create an accurate virtual representation will vary widely. In general, expect to export as many data streams as you can to see the best results.
• Create a secure connection. As digital twins require the constant transfer of data between systems, secure networks are vital to protect the main manufacturing plant from hacking or other interference. When choosing a digital twin platform, ask about the security measures in place to minimize risk.
• Plan for scale. Even if you’re starting small with a digital twin for just one machine or line, the computing power needed is likely to be beyond any standard system in the shop. As you consider tools and platforms, plan for scale and draw on cloud technology to ensure there’s room for growth if you choose to add other machine lines later.

Setting up a digital twin can help manufacturers improve productivity and safety on the shop floor, but a few conditions must be met to ensure success. Providing access to accurate data is crucial, as is ensuring a secure network. Beyond the technical specifications, leaders need to provide sufficient training, communication and incentives to make sure all employees are onboard with the process. Effective change management is key to successful implementation. Providers need to show the benefits of digital twins while training staff on how to use the platform. It’s important to understand that while accuracy isn’t always 100%, especially in the beginning, even at 80% to 90% the digital twin can easily improve outcomes, and will continue to get better as it receives and analyzes more data.

Working with an experienced partner in this field also helps manufacturers realize value. A digital twin provider that understands the ins and outs of how a particular plant is set up, what types of equipment are used, and industry-specific challenges will be better able to create a digital twin that meets the company’s needs and goals.

CV Ramachandran is a digital transformation and operations improvement expert at PA Consulting. Vignesh Ramesh is a digital health expert at PA Consulting.