A ‘digital twin’ of the supply chain lets a business model the potential effects of reconfiguring its processes, workforce, assets, and data before actually putting material and information through it.
- P.S. Easwaran
With COVID-19 bringing unparalleled disruptions, supply chains, multinational corporations, and local exporters are now focusing more on diversifying risk. By exploring a partial shift of manufacturing ecosystems out of concentrated zones, they aim to build resilience even while responding to the pandemic’s short-term challenges.
This means India—with a similar global value chain participation index as manufacturing-intensive countries such as China—can leverage the shift without major changes to operating models. Global companies planning an India angle can seek out asset-light models, with contract manufacturing, sourcing, and strategic partnerships. At an organisational level, firms can become globally relevant and amass exponential impact by redefining competitiveness, enhancing quality, and optimising costs.
Making the shift in competitiveness – time for digital twins
This shift in competitiveness requires organisations to consider revamping and recreating through digitisation. A ‘digital twin’ of the supply chain lets a business model the potential effects of reconfiguring its processes, workforce, assets, and data before actually putting material and information through it.
Put simply, a digital twin is a digital copy of a physical object or process. By continuously updating itself on the object’s behaviour, it helps in optimising performance. It also aims to identify deviations from an ideal and highlights possibilities to optimise costs, enhance quality, and bolster productivity and efficiency.
Digital twins are creating a ‘Midas effect’ across industries—be it in optimising automotive value chains, aircraft innovations, or digital oil-drilling models. Forward-thinking manufacturers are already benefiting from digital twins. For example, manufacturers of passenger vehicles create digital twins of the vehicles sold. While vehicles are simulated in factories, data transmitted via sensors from vehicles on the road are interpreted by artificial intelligence (AI) to evaluate their functioning. Even software updates can resolve maintenance issues (for example, adjusting hydraulics to correct rattling doors), so customers need not visit workshops. In automotive manufacturing, digital twins are incorporated into the stages of product development, manufacturing, and after sales services. Digital twins enable manufacturers to gauge market demand in real time by sensing customer needs. Unsurprisingly, Deloitte’s 11th Annual Tech Trends Report of 2020 also highlights digital twins as a major tech trend.
Now, imagine if the digital twin was extended to the entire value chain (sourcing, after sales, supplier collaboration, planning, asset maintenance, contracting, logistics, distribution, product development, value added services, and others), this could give businesses a complete, end-to-end digital print of operations. Building a digital twin broadly involves the following six steps. These are:
Creation: Outfitting the physical process with sensors to measure critical inputs.
Communication: bi-directional, seamless, real-time connectivity between physical processes and digital platforms.
Aggregation: Feeding data into a repository, processing, and preparing for analysis.
Analysis: Analysis and visualization of data to develop models generating insights.
Insight: Presenting findings, and
Action: Feeding actionable insights to the asset and processes.
As with any technology solution, digital twins have a few watch-outs. First, digital twins require a comprehensive roadmap of investments in platforms, models, and maintenance, especially for organisations that have not invested in process-enablement technologies. It implies relatively higher near-term investments, despite the prospect of higher Returns on Investment (ROI) in the medium term. Second, achieving precise representations of an asset’s physical, chemical, thermal, and electrical features can be difficult, thereby obstructing visibility into such processes. This can be navigated by incorporating assumptions, modifications, and simplifications into models. Third, data quality poses a challenge. Especially in India, with multiple aspects of the supply chain (for example, logistics and remote operations) data points often come from sensors in uncontrollable conditions, transmitted through patchy networks. Companies must plan for inconsistencies in the data flow. Finally, since a digital twin’s success relies on the availability of data, insufficient data can make the models ineffectual. To mitigate this, owners need to plan an architecture that enables the access and use of a wide range of data sources.
As seen, the challenges of digital twins can be manoeuvred with better foresight and planning. With the long lifecycles of manufacturing and the interlinkages between functions and resultant impacts, the deployment of a digital twin requires strong governance and a plan for evolution. This will enable agility and flexibility vis-à-vis the shift in marketplace competitiveness.
- P.S. Easwaran is Partner and Leader, Supply Chain, Deloitte India. Views expressed are the author’s own.