Autodesk has sponsored this post.
Unlike a biological or identical twin, a digital twin does not have a universally accepted definition. In application, a digital twin will mean different things to different industries. On an assembly line, a digital twin of a robot may look identical to the physical robot, especially if it is photorealistically rendered. The digital twin can mimic the physical robot’s movement, for example. The digital twin may not pass a close inspection for similarity, however. It cannot have the internal minutiae and complexity nature routinely provides. A robot’s digital twin may lack fastener threads, weld details, etc. found on the physical robot.
But unlike nature’s twins, digital twins need not replicate every bit, part and function of their physical counterparts in order to be effective. If the digital twin can determine the reach of the assembly line robot and prevent interference with the production line or other robots, then the digital twin can consider itself complete enough for that particular mission.
In fact, it is the differences between a digital twin and its physical counterpart that prove to be the most useful. A digital twin can make its physical twin follow its every move. You won’t get that with identical human twins. A switch thrown on a digital twin of a robot can be programmed to turn on the real robot. This can also happen in reverse: the digital twin can copy the movement of the physical twin. And so, a digital twin of a production line could either control the physical production line, or for the purpose of monitoring the production line, the digital twin would show its operation.
Outfitted with sensors and data streaming, the position of robots, production machines, parts and assemblies in a manufacturing environment can be relayed to their digital twins and shown on remote screens. A lights-off, fully automated production line could be seen through the behavior of its digital twin anywhere in the world, in 3D and in real time. This makes the digital twin the next generation of visual monitoring of systems, easily surpassing current methods of visual monitoring. For example, compare the realistic image of a digital twin of an assembly robot in motion providing instant understanding of what is happening on the factory floor to an instrument panel with only flashing lights to indicate a few of the robot’s operational parameters.
Another example of the practical value of digital twins is machine simulation. Using a digital twin to validate a manufacturing process can prevent actual, real-world machining delays and catastrophes. A digital twin of a CNC tool can prevent machine crashes and exceeding the machine’s limits halfway into a job. The digital twin will predict the machine’s performance and allow machining immediately at full speed, bypassing the usual very slow first run.
A digital twin starts to take shape during the design phase, before the machine or assembly goes into production. Engineers will start the creation of the digital twin of an automobile, for example, to produce a stunningly accurate computer model of every single part, inside and out. It may not have fuel in the tank, a charge in the battery or the scent of leather, but engineers can rest assured that all parts will fit together perfectly. There will simply be no clashes or loose fits discovered too late, after parts are manufactured. The attention to detail that 3D modeling allows will reduce the act of first physical assembly from suspense to anticlimax.
The virtual assembly lets designers and engineers assemble wherever they may be, together—or as is often the case these days, apart. An engine can be digitally mated with the chassis even though the physical counterparts are made in different locations.
After the physical counterpart leaves production, its digital twin can keep a watchful eye on it, aware of what it is doing and what is being done to it through its end of life. With data being collected while in use by the physical counterpart, the digital twin can show duty cycles and issue maintenance alerts. For example, a 3D printer can have its print heads cleaned before it starts making faulty parts.
With 3D printers and products of all sorts, a digital twin can be made to save time, money and resources.
For more information, see CIMdata’s eBook Digital Twin: Weaving the Digital Thread.