What it is: A Digital Twin is a software-based model of any physical entity or system, from an appliance to a human heart, to an entire building. Once created, a Digital Twin offers a way to monitor a physical system, or subject it to predictive calculations, or otherwise leverage data. Applications are widespread: sectors that could be affected include the automotive industry, the medical industry, manufacturing, and many others.
What it does: A Digital Twin outputs the functioning of the physical system it models and can be reconfigured remotely, in (almost) real-time. For example, a Digital Twin of a tractor could report that one of its parts is malfunctioning; or a Digital Twin of an organization’s department could instantly report bottlenecks in production. The focus is on detailed, real-time logging of integrated data.
Why it matters: Data can be useless unless contextualized. By modeling a physical system, a Digital Twin can take individual pieces of data and place them within representative models. This is already valuable for monitoring purposes, and such data can inform machine learning algorithms to produce detailed predictive models. For example, a Digital Twin of a malfunctioning vehicle could be used to feed a model that predicts the circumstances in which another such vehicle will fail.
What to do about it: Consider whether your enterprise could benefit from more detailed modeling of specific assets or processes. Look into how feasible it would be to collect large amounts of data about this asset or process, to create a Digital Twin. Note that the creation of a Digital Twin is often costly, for example requiring management tools and processes, though the long-term advantages can be striking.