A New Study Finds No Clear Consensus on the Future of Digital Manufacturing

In this first entry of a two-part report, we examine why industry experts think there’s a lot of uncertainty when it comes to digital manufacturing between now and 2030. Part two of the report will drop next week.

An engineer uses a tablet to model, design and monitor a 3D model of an aircraft engine while on an automated robotic line at an aerospace factory. (Image: Siemens)

A recent study surveyed international experts across industry and academia to assess how and where interconnected digital twins (IDTs) will impact digital manufacturing up to 2030. Through their study design, the researchers came to an interesting conclusion: no obvious consensus exists for where digital manufacturing is headed, leaving plenty of room for ongoing innovation over the next few years.

A conventional digital twin is a simulation model of a single asset or machine that an engineer would trust as much as the device itself. Since digital twins are designed to behave exactly the same as the physical system, any data generated during simulations is usually considered with the same scrutiny as physical experiments. Today, most conventional digital twins are used to create knowledge and support decision-making and are often siloed and coveted like a trade secret.

In contrast, interconnected digital twins (IDTs) support a more open data ecosystem by integrating data from multiple devices from within or between companies. This type of simulation strategy results in more comprehensive solutions and can help facilitate innovation across an entire industry. By integrating data from devices amongst manufacturing facilities and even between companies, engineers can use the expansive information to drive system-level innovation. The aggregate model generated by an IDT can then make information more accessible across or even between companies to support decision-making and optimization of operations.

Unlike traditional digital twins, IDTs need to carefully consider where and how data is collected and integrated into the aggregated simulation. IDTs usually rely on data spaces to ensure governance and security regulations compliance. This means that the data “stays” in its native location, and an app (i.e., the IDT) is brought to the data—a stark contrast from conventional digital twins, which bring the data to the simulation.

But if IDTs seem like an obvious choice to support an innovative, open data future, why are they not being more readily adopted? This and other questions were the main focus of the recent study published by van Dyck et al. in the Journal of Product Innovation Management.

In their 2023 publication, van Dyck et al. used a forecast-based survey, known as a Delphi study, to evaluate how IDTs might shape the future of digital manufacturing. To begin, the researchers asked a panel of experts to share their opinions on several topics related to digital manufacturing between now and 2030. They conducted more than 100 interviews to probe industry leaders about future trends associated with Industry 4.0 and the adoption of IDTs. From these interviews, they developed 24 projections and presented them to a separate set of 35 experts to gauge their opinion on the likelihood of a given prediction. For example, how likely a certain technology will be adopted by 2030. These rankings and qualitative feedback were then collected and aggregated to identify any trends that emerged across the industry experts.

Interestingly, although the researchers were hoping to characterize an industry-wide consensus for certain topics and projections, the main takeaway was a lack of agreement across experts regarding the future of digital manufacturing. The study’s lead author, Dr. Frank Piller, hopes these results can be used to spark important discussions as companies progress towards the realization of Industry 4.0.

Despite a lack of clear consensus among the surveyed industry leaders, the researchers were able to identify four major trends that will likely impact the future of digital manufacturing, focusing on IDT implementation between now and 2030:

1)      Decentralized Data Generation

First, the panel of experts expressed an ongoing trend towards decentralized operations and data management. Specifically, when it comes to IDTs, the adoption of this type of simulation strategy can facilitate data systems that are decentralized, with data-driven insights more readily shared across locations, use cycles and technical stakeholders. By decentralizing data generation and analytics, insights for digital manufacturing can be scaled to achieve company-wide or industry-wide improvements more efficiently; a key strategy to remain competitive in an increasingly data-driven world.

2)      Transparent Operations

Governments, industry stakeholders and consumers increasingly demand more transparency around operations and environmental sustainability. By design, adopting IDTs is a clear strategy to increase transparency in manufacturing operations and expand data sharing within and between companies.

3)      AI-Assisted Automation

An unsurprising but nevertheless consistent trend that emerged from the survey was an increase in AI-assisted automation within digital manufacturing. The industry experts agreed that AI-driven decision-making will continue to expand between now and 2030, especially as AI-powered tools continue to mature.

4)      Subscription Models for Equipment

The final and most interesting trend to emerge from the study is an ongoing shift towards subscription models for production equipment that are outcome-oriented. This means that a company will guarantee a certain level of performance for equipment in exchange for a periodic fee. Performance data will therefore be shared through an ongoing process between the original manufacturer and the subscribing company to support real-time improvements in key manufacturing equipment.

For years it seems that the promise of Industry 4.0 has remained just out of reach. Yet, digital manufacturing is anything but stagnant; it’s simply moving at a slower pace than our projections for the potential of emerging technologies. For example, the value of a company’s data is often overestimated, leading companies to hoard data under the guise that it will one day lead to earth-shattering insights and innovations. Although big data will be a key player in the ongoing development of digital manufacturing, it’s certainly not the full picture. This is especially true when you consider data on a small scale: a single machine, operating plant, or even a full enterprise. Instead, the main findings from this study seem to indicate that more diverse data can often lead to greater advances in innovation, and the creation of an IDT can facilitate a more robust, reliable simulation of assets. What remains to be determined is how a company can continue to gain a unique advantage over their competitors when sharing and accessing this disparate data. The authors highlight this as one of the major concerns holding companies back from developing and investing in IDTs.

It seems that between now and 2030, digital manufacturing will best align with a “hybrid intelligence” model, whereby human decision-making is supported by AI-driven data analytics, with data hopefully coming from as many sources as possible. Although completely automated decision making in digital manufacturing remains enigmatic, most experts seem to agree that a sliding scale of “hybrid intelligence” will allow the use of leading technologies, without limiting the utility of human critical thinking. Hopefully as more and more open data philosophies are adopted, the use of IDTs will also expand, helping companies realize the full potential of big data in manufacturing.