TechTarget defines a Digital twin as "a virtual representation of a real-world entity or process" that "functions as a proxy for the current state of the thing it represents ... in the early stages of product development" and "is retained and updated for later stages of the product's lifecycle such as inspection and maintenance."
Digital twins are used in R&D to design effective systems and, increasingly, during operations to assess the impact of proposed changes and gain insight into effective operation. Not every process or system is suitable for a Digital twin, and the cost of developing one must be offset by a degree of complexity that makes the insights that can be gained valuable.
Organizations often use digital twins to make choices about how best to configure and run the system or process. The digital twin is used to simulate the likely impact of changes, to evaluate possible improvement projects and, when linked to real-world data, to provide insight into how the process really operates. This last point is critical: A two-way interchange of real-world data flowing into the digital twin and insights flowing back out to improve the real world is central to driving continuous improvement using a digital twin.
Some processes revolve around an operational decision. A claims process depends on decisions about claim eligibility and claim payment. Mortgage or loan processes depend on the credit and Risk decisions made. Benefits processes rely on the eligibility decision, delivery processes rely on staffing and scheduling decisions. A digital twin allows the impact of a change on these decisions to be assessed. For instance, if credit Risk is tightened so that more loans need manual approval, how does that affect wait times and customer satisfaction?
The extent to which decision-making is built into the digital twin is a critical constraint on the value to be gained. If the decision can only be varied at a macro level, little value will be gained. If, however, there is a component in the digital twin specifically focused on decision-making (a digital decisioning twin), much more value can be gained.
Risk management can be improved. A digital decisioning twin can allow the impact of different risk tolerances for credit risk, fraud risk and supply chain risk to be assessed at a very granular level.
The regulatory impact can be predicted. When regulations on things such as consumer privacy, consumer credit or government benefits drive decision-making, future changes can be guaranteed. A digital decisioning twin can show the impact of future regulations and how other changes can be used to mitigate these impacts.
ML and AI can be more readily adopted because the impact of ML and AI algorithms on decision-making can be simulated in the twin. This can allow a concrete discussion of the trade-offs, replacing both fear-based rejection and blind adoption.
Digital decisioning twins can be combined with a broader digital process twin to improve simulation accuracy. Instead of using "average" decisions in the process simulation, transaction-by-transaction changes can be simulated.
As with any digital twin, the first step is to understand your business. This can be challenging because so much decision-making is only in people's heads.
