Digital Twins Are Shifting to Favor End-to-End Business Problems, Underpinned by AI

The concept of a digital twin has been around for several years, but the sentiment around its benefits and use is changing. Originally, management said, “we need to have a digital twin,” and Industry 4.0 teams hustled to assemble the various components from the ground-up, and then looked for practical applications. Now, the prevailing sentiment is to work backward from business challenges with a problem-led, rather than technology-led innovation stream. Additionally, the market is coming to terms with the fact that there are different levels of fidelity when it comes to digital twins—from basic to comprehensive—due to the desire for more complete capabilities that apply to more stakeholders, from design and manufacturing to marketing, sales, and service. These more comprehensive digital twins are essential to maximize investments in Artificial Intelligence (AI).

Rusch Maschinenteile is a German family-owned manufacturer of workpieces. Over the last few years, it has needed to produce increasingly complex parts in a shorter amount of time, while meeting stringent quality requirements. To meet such demands, the company employed a comprehensive digital twin that spans its DMG MORI Computer Numeric Control (CNC) machines, Siemens SINUMERIK ONE digital machine controllers, and customers’ workpieces.

This approach allowed Rusch Maschinenteile to virtualize machining processes, simulate and test operations before manufacturing, and minimize downtime. The result was 40% faster production ramp-up, less scrap and rework, and better utilization of equipment by minimizing unproductive machine time by up to 75%.

Rusch Maschinenteile customers benefitted from higher quality products delivered faster, plus the ability to view and address issues before they occurred by working with digital models ahead of workpiece production. For Rusch Maschinenteile, the use of a comprehensive digital twin—spanning all engineering domains—meant 80% fewer manual programming tasks and a 20% increase in overall performance due to reduced cycle times, and a digital backbone ripe for future innovation in areas that include AI.

A comprehensive digital twin unifies the entire workflow for operators and engineers to help them see the entire picture. For example, by testing and simulating programming and machine performance in a virtual environment before real-world deployment, engineers can minimize programming errors that cause scrap and machine damage. Engineers and frontline workers can also collaborate on process sequence steps to improve the performance of CNC machines across the fleet by scheduling orders from multiple customers in the context of capacity versus delivery commitments.

Basic digital twins that convey telemetry data are still being entertained, but such installations should only be pursued with a more comprehensive digital twin in mind. One of the core reasons for this new approach—for comprehensive, rather than basic digital twins—is to prioritize standard data models that promote synchronicity across the organization.

Data synchronicity is critical—such as the linkage between Product Lifecycle Management (PLM) and Application Lifecycle Management (ALM)—because, these days, both products and processes are largely driven by software. Furthermore, the availability and prevalence of AI means that a common data foundation is critical to enabling agentic workflows that free up time for employees to perform more meaningful tasks.

One example of an agentic workflow would be to manage the linkage between PLM, ALM, and the supply chain, so that when a requirement changes in one area, that change request is synced in other areas, without the need for manual coding or updating. Such agentic workflows mark a major step forward in terms or productivity gains, and comprehensive digital twins that are created to solve business challenges are the best way to maximize these investments.