Digital twins help to reshape manufacturing

Digitized production and use of a digital twin improves the amount and effective use of valuable data about the manufacturing process. See two methods to create a digital twin.

By Ashok Kumar October 4, 2017

Over the last few years, digitalization has brought about a fundamental change in our everyday life and the manufacturing industry is no exception. Factories are transforming manufacturing processes to produce customized products and systems and are leveraging key technologies to digitize production and the supply chain. One such technology that has changed the face of the manufacturing industry is the digital twin model.

Among the top 10 strategic technology trends in 2017 as noted by Forbes, digital twin technology has helped manufacturers use tools like conceptualization, collaboration, and comparison. Together, these form the substratum for next-generation solutions and innovation. Since the inception of digital twin technology, the manufacturing industry has seen a tremendous surge in the amount of valuable data that is being collected and the fidelity of information about physical and virtual products. This allows manufacturers to look beyond clay models or even physical prototypes, helping bridge the gap between the virtual and the real space.

Analysts predict that connected products, backed by actionable real-time data, will be represented by digital twins. With the unprecedented growth of the Internet of Things (IoT), there is a sudden burst of interest in this concept. By 2018, organizations investing in IoT-based cognitive situational awareness or operational sensing will witness improvements of up to 30% in critical process cycle times, as Forbes cited. Consequently, manufacturers will be able to compare both actual and anticipated scenarios to make informed business decisions.

Product environments are becoming increasingly complex with new feature requirements, stringent regulatory guidelines, and rapid development cycles. Although enterprises like to collect troves of data, they lack a comprehensive digital plan to transform their capabilities radically.

It is often the case that legacy simulation frameworks fail to consider equipment aging, fault and failure conditions, wear and tear based on material properties, and the environment. Moreover, incumbent simulation software packages haven’t evolved quickly. The ideal scenario is to develop a robust roadmap that can accurately predict system performance, incidence of errors, and ascertain the ROI. So, to lay the foundation for connected products and services, digital twins are becoming a business imperative.

Defining the manufacturing ecosystem

Manufacturers today often grapple with the challenge of designing efficient products. Interestingly, a recent survey revealed that while companies invest in analytics, DevOps remains the top priority. (DevOps is a conflux of development, operations, software development, and delivery process that emphasizes better communication and collaboration.) Enterprises that embrace the digital twin model can analyze their data and monitor systems to detect problems in advance, preventing downtime. As simulating new innovations is integral to the digital twin model, "what if" scenarios are generated to boost time to market, eradicating cost-intensive product quality issues.

At the onset, smart components with sensors gather real-time data to enable seamless integration with the physical asset. These components also are connected to a cloud-based system which gathers and processes the operational and environmental data monitored by the sensors. 

Digital twins: two methods

Creating the digital twins involves two crucial methods.

1. Outline the digital twin process and information requirements pertaining to the product life cycle.

2. Create the technology that integrates the digital and physical asset for real-time flow of sensor data.

Critical inputs from the sensors are analyzed against contextual data to identify areas of improvement. Additionally, the digital insights are applied to create new services for the business. This next-gen model yields an evolving digital profile of a physical object or process, helps in analyzing incoming data, and identifies deviations from an optimal point. 

Creating digital twins

As the Industrial IoT (IIoT) continues to gain prominence, applying predictive analytics will become integral to meeting business goals. The IIoT involves harnessing industrial data and using insights to generate accurate predictions. Digital twins increasingly are being deployed in predictive analytics to sort this data.

In the automation industry, most companies are using digital twins for optimization and virtual commissioning. However, they usually are not leveraging it for preventive or predictive maintenance. One exception is a subsidiary of a U.S. multinational conglomerate that created a digital twin of a steam turbine at a conference held in November last year to highlight possible areas of predictive and prescriptive maintenance.

NASA has adopted the digital twin model to develop road maps, next-generation vehicles, and to help decide on new product recommendations. The digital twins of spacecraft and stations ensure that the systems run effectively, and the crew is out of danger. This paradigm shift to the parallel software model has ensured greater reliability and safety. 

Embracing digital twins

By 2022, 85% of all IoT platforms will embrace digital twins, according to the Research and Markets report, "Internet of Things (IoT) Digital Twinning: Market Outlook for IoT enabled Physical to Virtual Mapping and Management 2017-2022." Subsequently, companies will reduce investment in physical prototypes and shorten the go-to-market period. The product life cycle will become more efficient.

Furthermore, in the manufacturing sector, predictive engineering analytics (PEA) will leverage the digital twin model to provide operational data. This model also is being applied to achieve closed-loop product life cycle management (PLM), which streamlines the digital thread and facilitates smarter enterprise asset management.

The digital twin model will soon function as a proxy for users with specialized skills, including engineers and manufacturers. Clearly, it is the next technological wave in simulation, conceptualization, and optimization.

Ashok Kumar is general manager, industrial products at L&T Technology Services, a CFE Media Content Partner. Edited by Mark T. Hoske, content manager, Control Engineering, CFE Media, mhoske@cfemedia.com.

Key concepts

  • A digital twin allows changes to be tested prior to switchover.
  • Preventive or predictive maintenance could benefit from digital twin architecture.
  • IIoT platforms are embracing use of digital twins. 

Consider this

Would justification such as fewer or no outages be enough to expand use of digital twins? 

ONLINE extra

See other L&T articles linked below.

See more on digital twins from Control Engineering.

Link to more information from Forbes and Research and Markets, provided by L&T Technology Services.