Digital twin technology for maintenance
Introduction
A digital twin is a computer system programmed in such a way that, by receiving the same inputs experienced by the object or physical process of which it is a twin, it provides the same outputs.[1] For example, a gas turbine has as inputs methane at 97% purity and air at 20 °C, and drives an alternator that produces electrical energy. A properly programmed digital twin of this turbine can predict the gigawatts of power the alternator will produce if the methane purity drops to 96% or the air temperature rises to 21°C.
When it is said "when it receives the same inputs, it provides the same outputs" it does not mean that physical air enters the digital twin of the turbine, but rather that the data of that air enters (quantity, temperature, etc.). The same goes for exits.
Origin
Digital twins are the result of the evolution of simulation, which has been around for centuries. The construction of small-scale ships, to test the buoyancy and maneuverability of the real ships that would be built based on them, reached its peak in the 19th century.[2].
Digital twins were anticipated by David Gelernter in his 1991 book Mirror Worlds.[3].
The first practical definition of a "digital twin" is attributed to NASA, which was trying to improve physical aircraft model simulation in 2010. However, the concept is older and is believed to have been launched by Michael Grieves in 2002, when he worked at the University of Michigan.[4] Grieves, from the Florida Institute of Technology, was the first to apply the digital twin concept in the manufacturing.[5][6][7][8][9][10].
The concept and model of the digital twin were presented publicly in 2002 by Grieves, then of the University of Michigan, during an Association of Manufacturing Engineers conference in Troy, Michigan.[11] Grieves proposed the digital twin as the conceptual model underlying product lifecycle management (PLM).
The concept, which was given several names, was later called the "digital twin" by NASA's John Vickers in a 2010 roadmap report.[12].
Parts
The concept of a digital twin consists of 3 different parts: the physical object or system, the computer object, and the connections between them. These connections are the data and its transmission media (cables, radio frequency, laser, ultrasound, etc.). Data is taken from the physical object and its environment using sensors and meters and entered into the computing object. This object does not have to present a visual appearance on the screen similar to that of the physical object. A perfectly functional digital twin may simply consist of sensors, wiring, a computer, and lines of computer code. The important thing is that it takes the necessary variables from the physical object and its environment to faithfully reproduce its behavior, and that the treatment of these variables is appropriate.