Tougher market conditions, rapid technological innovations, and changing customer needs are requiring companies to become more responsive to customer desires and expectations. From a manufacturing point of view, the shift from selling to servicing a product is the future if manufacturers want to increase revenue and profit margins.  Underpinning this business model shift is the trend towards Internet of Things (IoT) or the world of connected digital twins of products and processes, which ultimately form a “Digital Shadow”. Whilst IoT illustrates a systemic view of the interconnected world between many assets and systems, a digital twin is a narrow view of a connected singular asset or group of assets.
The adoption of these technologies is a key enabler for the fourth industrial revolution, where possibilities and opportunities are endless. What if you could have real-time visibility of how a factory is performing all the way down to the machinery used to conduct individual steps in an assembly? What if you could analyse the health of a fleet of vehicles and identify an issue before it has even occurred? What if you could improve the flow of goods and inventory to deliver higher service levels? This is the bridging of the physical and digital world to facilitate more direct models of personalised production, customer interaction, and improved efficiencies.
Empowered by data, 5G connectivity and artificial intelligence, the transformation of business models will be coupled with the adoption of digital twins. Driving the business case of digital twins is the opportunity for manufacturers to realise operational cost savings, increase productivity, improve customer experience, and define new revenue streams. Research indicates that by 2030, 30% of G2000 companies will be using data from digital twins of IoT connected products and assets to improve product innovation success rates and organisational productivity, achieving gains up to 25%. Underlining the importance, SAP Senior Vice President of IoT recently stated that “digital twins are becoming a business imperative, covering the entire lifecycle of an asset or process and forming the foundation for connected products and services. Companies that fail to respond will be left behind.”
The digital twin is a virtual representation of an asset, which can be in the form of a physical product or a process. The technology records the real-time performance of an asset and its deviations from its optimal performance – creating a repository or a “twin” of the asset in the virtual world. It comprises of three key elements: physical assets in the real world, virtual assets in the virtual world, and the link for data flow between the real and virtual world. 
The main objectives of the digital twin are:
These applications translate into substantial benefits:
To produce a digital twin of a product or process it requires the collection and centralisation of data across many sources. In the context of manufacturing, the four main sources of data are from sensors (attached to products or processes), enterprise data systems, RFIDs and machines or robots. Once the data is collected, it is relayed it through edge computing interfaces and integration layers, which collect and store data, to be deposited into the cloud. As data is continuously collected over time, the cloud becomes a centralised repository of an asset’s current or past state – cultivating a “single source of truth”.
The centralised repository provides the opportunity to apply analytic and human interface applications for organisations to monitor an asset in real-time. Capitalising on this wealth of data, the digital twin can generate recommendations if an issue has been identified. Supporting manufacturers to diagnose and solve issues quicker, the technology stack ultimately allows organisations to reduce asset downtime and keep production running smoothly.
Taking a step further, with the use of artificial intelligence, organisations can now not only diagnose current inefficiencies, but also identify future issues and tap into “predictive maintenance”. In order to pre-empt issues from arising, predictive maintenance is added intelligence that can predict when or if functional equipment or processes will fail. Thus, optimising operating costs by reducing maintenance costs and increasing profitability.
The digital twin has shown that it can be applied in different levels of complexity and across the value chain. Leveraging the digital twin’s wide range of applications, manufacturers have used it to support in operations and facilitate asset management. This has translated in bottom-line gains and top-line growth.
With increased competition and changing customer demands, getting visibility of one’s logistics can enable manufacturers to better manage inventory levels, increase production efficiency and deliver higher service levels. SKF, the world’s largest bearing manufacturer, built a digital twin of its entire distribution network to create an integrated planning model vision. Rather than planners organised by regions, they would have global responsibility by production line. Enabled by the visibility of their entire supply chain, each “global planner”, responsible for one product, would have for their own forecast, inventory and end-to-end plan. This has enabled SKF to achieve end-to-end optimisation of their production and logistic network, whilst improving overall lead times.
The arrival of 5G provides the opportunity for manufacturers to realise the potential of IoT within their warehouses in order to optimise assembly time and shorten time to market. Tight delivery times, misrouted containers, and extensive manual supervision limited a large German OEM from operating their warehouses effectively. Therefore, they implemented a digital twin of their warehouse, which included robots and processes. By improving visibility of their warehouse operations and communication between assets, the OEM addressed problems quicker, reached higher levels of process automation and increased service levels.
Going up the value chain, and into the customer’s world, the digital twin not only enables manufacturers to respond to assembly related issues, but also to customer or end product issues – highlighting the shift towards servitisation. Tesla, for example, have a digital twin for every VIN or vehicle that is in customer hands. Through the use of software updates and remote diagnostics, Tesla have leveraged predictive maintenance to reduce asset downtime. Since May 2019, Tesla vehicles can now predict when future issues will arise and automatically order parts on behalf of the customer. This all contributes to Tesla’s ability to deliver a superior customer experience.
Looking beyond cost optimisation and with the explosion of big data, manufacturers can extract customer insights to develop new products or services. Stara, a Brazil-based tractor manufacturer, used the digital twin to monitor their customers’ products but also to understand how they used them. Whilst Stara were able to improve product uptime for customers, they indirectly discovered a new revenue stream. Soon, Stara launched a profitable new service that provided farmers with real-time insight about the optimal conditions for planting crops to improve overall productivity. As a result, farmers reduced seed use by 21% and fertiliser by 19%. Echoing this benefit, Catapult High Value Manufacturing, a UK government thinktank, supplants “through increased customer intimacy greater insight into product or process... Digital twin can be used to influence the next generation developments in physical products or processes”.
The adoption of the digital twin is becoming widespread in the manufacturing sector because companies are utilising this technology to solve business problems across their value chain. According to IDC research, in 2016, manufacturing was one of the industries with the largest use case for digital twins, accounting for more than $102 billion in total expenditure in this technology. Accentuating its impact, researchers estimate that by 2025, improvements delivered by IoT and digital twins could be worth more than $470 billion per year in the manufacturing sector.
Despite the enormous benefits, there are challenges that lay ahead for manufacturers that want to adopt this technology, such as uncertainty of the ROI or business case, lack of qualified employees, and data management issues. These are all the type of areas that need to be addressed before and during the implementation phase of this technology. A key differentiator for manufacturers in the future will no longer be whether they want to adopt this technology, but how they adopt it and enable this digital transformation. Scope, scale, and speed will be the key drivers for manufacturers that want maximise the benefits of this transformation.
Maximilian Poupon, Jeremy Hammant, Emile Naus