“The digital twin can make predictions about part quality, when maintenance will be needed and more,” says Dawn Tilbury, Ph.D., a mechanical engineering and robotics professor at the University of Michigan. “However, lack of standardization makes digital twins difficult to implement in manufacturing.”

Currently, Tilbury says most digital twins are for single devices. If a factory has a digital twin for an entire system or line, it is likely a custom piece of software that has to be rewritten if a part of the system changes. Meanwhile, many companies are providing digital twin software for individual machines.

According to Tilbury, linking them up with twins of the previous and next machines on the line would enable better coordination both up and down the line.

“Everyone’s building digital twins, but we’re trying to build the glue or connectivity that enables [the technology] to work together—to be composable, reusable and maintainable,” explains Tilbury. “[We] will work on issues that can bring broad benefits to everyone making and using digital twins, helping expand the use of these great technologies to improve manufacturing performance, quality and uptime.”