From Physical Products to Digitally-engineered Meta-Products: How will Digital Twins accelerate the transformation of industrial processes?

Smart factories – Once again, a booming prospect

The digital twin concept is not new, especially in the manufacturing space, but its intrinsic value is expected to take a leap in the coming years. Forecasts of the global digital twin market hint at massive growth for the next decade, approaching $110Bn in 2028, up from $3Bn in 2020 – which is a 36x increase in 8 years!

This is driven by the multiplicity of use cases derived from digital twins, from automobiles to smart cities, to basically any industry and geography. Energy, mining, manufacturing, healthcare, aerospace… every key player is entering the race to leveraging digital twins. Robot manufacturers such as Siemens, Schneider, or ABB, are even further committed to that?shift, having a dedicated Digital Twin strategy.

Digital twin of a plant?

Consumer-goods giant Unilever announced in 2019 it was building virtual versions of its factories, using data streaming from its sensor-equipped machines.

The goal??Create digital models of plants that can track physical health of assets and enable testing of operational changes.

How does it help the business??Digital models let the company make real-time changes to optimize output, use materials more precisely, and help limit waste of products that don’t meet quality standards – ultimately saving costs (Unilever’s project saved $2.8M at their Brazilian plant, cutting down on energy use and driving 1-3% increase in productivity).

Business value – Primarily, cost savings

Digital twins of plants gather data from multiple sources (physics-based and model data, analytical models, logs, transactional data), and make the information available to various areas of the business. This vastly improves decision-making through a shared understanding of operational status and reduces the overall lifecycle cost of operating and maintaining a plant.

Typical use cases of a plant digital twin are:

• Increasing uptime through predictive maintenance:?traditional maintenance can identify root causes of failure and replace broken assets when failure has already taken place (corrective maintenance), which is costly in downtime, labor, and unscheduled maintenance requirements. Preventive maintenance allows to perform maintenance before the actual failure happens, by determining the lifetime of assets and their parts and scheduling maintenance at regular intervals before product end of life. With the advent of digital twins, businesses are moving towards building predictive maintenance models that optimize maintenance cycles and use?just in time?replacement of components. Predictive maintenance can support replacing components that are close to failure and extend component lifetime by reducing unscheduled maintenance and leveraging massive amounts of data to predict maintenance needs. Thereby achieving cost saving and substantial competitive advantage.
• Reducing development time through increased transparency:?in complex systems, development process is often arduous since the various stakeholder groups (sales, engineering, finance, etc.) may have different or contradictory product requirements. Physically testing design or material changes is a long process and can prove very costly. On the other hand, digital twins allow for easy visualization and modeling of technical or material changes, as a result the implications for meeting customer requirements and product cost targets become clearer. As a digital platform, the technology allows for proper capitalization to refine tests over time.
• Avoiding bottlenecking through design optimization:?a plant digital twin can model the structural complexity of industrial processes, key operational constraints, product shelf life, labor skill levels, multilevel bills-of-materials while considering the probability of failures across all machines within its ecosystem. Compared to traditional plants for which bottlenecks can only be removed once identified (meaning they are already impacting production level), digital twins can be used to evaluate, and optimize the process design before implementing it or even before allocating capital, increasing overall ROI.
• ?Optimizing throughput via real-life rescheduling:?traditional planning systems only generate the schedule on a weekly (or at best daily) basis, therefore creating a discrepancy between planning and real-life state of operation. A factory digital twin on the other hand, can enable scenario planning, consider trade-offs as well as possible outcomes while still capturing operational requirements and scheduling rules. Therefore, determining the best labor shift requirements and optimized material arrival times at corresponding operations; effectively orchestrating all intertwined processes traditionally siloed.?
• Improving capital allocation decisions:?COVID-19 pandemic and the most recent Ukrainian crisis are major examples of international events that put a dire pressure on supply chain. To lessen future risks, many manufacturing companies are thinking about bringing production back closer to home. Such investments can be significant and digital twins can ensure that they yield the expected returns by pressure-testing all variations. The value created for the business lies therefore in the?long-term reduction of operational expenditures, especially on maintenance and R&D even considering a one-off cost of larger CAPEX during development & deployment of the digital twin.

To minimize Digital Twin Capex spend, clear specifications need to be defined, focusing on the primary value generating?assets?needing to be digitally twinned. To avoid over-engineering and over-analysis, requirements need to determine the?accuracy and robustness?of the digital twin as well as potential relationships with other twins.??

Large scale implementation challenges will revolve around?accurately assessing physical assets and value chains current state and opportunities,?evaluating data management practices?(determining the volume, variety, velocity, value, veracity and variability of data), and?establishing a company-wide architecture spanning both IT and OT.

A clear emphasis should also be put on the team driving the project. One dedicated team should own the digitalization initiative, hopefully sponsored by C-level executives, but all functional groups concerned must be heard, and their requirements collected.?

Using the Metaverse to go further

By replicating a complete factory as-is in the Metaverse, from the earliest stage of the production chain to the end-product (with current 3D, photorealistic rendering design), a transformational shift occurs. Indeed, in traditional industrial processes, the value lies in the physical end-product. On the other hand, in digitally twinned factories, the physical product becomes a mere manifestation of a single version of the?virtual meta-product, designed specifically for a targeted customer. The underlying value lies in the fact that the?Meta-product is accessible and viewable remotely from everywhere, as well as infinitely customizable.?Therefore, selling the Meta-product allows for an increased engagement of the customer in the production phase, since the customer is able to discover their future product early on directly in the virtual shop floor and ask for the enhancements they want.

Such product design and new sales processes provide reduced inventories due to delayed differentiation, reduced development times and simplified industrial processes. The overarching target is the end-to-end product creation and design directly in the Metaverse, and a “simple” 3D-printing of the product in the real world, that could even be done directly in retail spaces to minimize sales disruption.?

Business value – the monetization potential unlocked

Apart from reducing the cost baseline, digital twins can also be used to generate new business models, such as:?

• Immersive experiences in the sales and customer service processes:?using digital 3D modeling, it is possible to recreate photorealistic 3D visualization of the plant and/or product, boosting sales by generating customer adoption and engagement.
• Monetization of product lifecycles:?Tesla is already selling access to advanced features that go further than the standard Autopilot at cost for customers when selling the car ($6k-12k). Those features are only enabled by the continuous integration and development of the digital twins of the car. Other carmakers (like Volkswagen) offer such autonomous driving features as pay-per-use through a subscription model rather than as one-off purchases.
• Monetization of production insights from the digital twin:?robot manufacturers such as Siemens, ABB or Schneider Electric, can sell more expensive maintenance service packages by ensuring higher uptime and reduced maintenance cost to their end-consumer.
• 3rd party access:?smart cities can monetize the access of their digital twins to architecture, telecom, or energy firms to allow better integration of the building in the city, for more accurate targeting and positioning of 4G/5G antennas, and for more accurate positioning of solar panels.
• Print on-demand:?in the luxury goods sector, there is an emerging trend of last-minute finalization of products such as handbags, leveraging all the data of the digital twin to customize bespoke products for clients, paving the shift from global factories to proximity 3D printing workshops.

Future of the digital twin

If digital twins are already a reality, their usage is only expected to grow further, driven by the manufacturing industries aiming to reduce costs and improve operations.

How will the digital twin model evolve in the future?

Immersive features of digital twins are forecasted to improve, as is already showcased by the partnership between Siemens Xcelerator (the partner ecosystem of the automation conglomerate) and Nvidia Omniverse (the platform for 3D design of the chip manufacturer). Siemens and Nvidia believe that connecting those two platforms will facilitate industrial automation, using photorealistic real-time 3D graphics and real-time AI-driven physics to simulate manufacturing applications more accurately. Augmented reality smart glasses will also be used in conjunction with digital twins to facilitate the interface between man and machine. By doing so, the real features of any operator (height, stress, movements, etc.) will be accounted for, hence increasing the workstation’s ergonomics and reducing the risk of work-related accidents.?Metaverse is actually a real opportunity to put the human back in the center of industrial processes.

Digital twins will also form an interconnected and interoperable ecosystem. Right now, digital twins are specifically linked to one asset, product, or system and there is no prevalent standardization protocol to ensure interoperability across vendors. One key challenge to unlock maximum value potential of the digital twin ecosystem will be to enable a vendor-agnostic and open information exchange along product lifecycle and across organizational boundaries.

They will be used across an increasing number of industries.?Use cases can be found in:

• Healthcare?- creating digital twins of hearts and lungs, that are customized to personal health data, allowing doctors to better monitor and anticipate the heart and lung’s behavior
• Sustainability?- the European Union is currently developing a Earth’s digital twin to monitor and predict interaction between natural phenomena, human activity, climate change, extreme weather events, and their socioeconomic impact as well as possible mitigation strategies
• Local city planning?- digital twins of smart cities can be modelled to optimize energy consumption through digital imagery, adapt shops’ opening hours to the flow of pedestrian traffic etc.

Though investments are heavy, and technical challenges can be encountered, digital twins appear not to be a trendy technical fad but rather a key stepping stone to enable a successful digital transformation model.

Authors:

• Vanessa Lyon, BCG, Managing Director and Senior Partner?
• Mika?l Le Mouellic, BCG, Managing Director and Partner
• Jean-Christophe Laissy, BCG, Partner and Director, Tech transformation@scale
• Laurent Alt, BCG, Associate Director, Enterprise Agility & Systems Engineering
• Alexandre Toureh, BCG, Senior Associate
• Victoria Guérendel, BCG, Associate?

Related links: