Digital Twins – Who Do You Like Better?
The study of twins (digital twins being a whole other ball game) has long been an interesting topic for psychologists, genetic researchers, doctors, and scientists. Most of these studies have focused on the potential of varying the environment for one of the twins while keeping the other unchanged to see how traits or genetic disorders can be passed on or treated. If Hollywood is to be believed, this can all be done using a simulation that a geeky computer whiz can type up in 5 minutes, right? Wrong!
What started as science fiction in the early 90s in the book Mirror Worlds, Digital Twins, became a technological reality when it was introduced to the manufacturing process in 2002 by Michael Grieves. A digital twin is a computer representation of a physical component or process that can be used to run predictive models and tests with real-time information. Although this sounds like simulations, a digital twin uses real-time data from sensors on the part to update the digital representation, as opposed to simulations that use a point-in-time reference of the part. Digital models also differ from digital twins as they don’t contain a physical counterpart and just exist digitally.
The name “digital twin” didn’t exist until NASA used the word in one of their reports in 2010, and the concept had been called many things before that. Consisting of three distinct parts, a digital twin has a physical object, a digital object, and the connection between these two objects. Apart from the obvious use cases of modeling wear and tear on products, there are several use cases of digital twins that make it ripe for large scale investment by technology and manufacturing firms:
- Preventative Maintenance: Using machine learning algorithms, engineers can predict when a part is going to fail and have a replacement strategy in place to reduce downtime.
- Fashion & Retail: Enhancing customer experience by helping customers model different designs and styles for both personal use and shop floor layouts.
- Efficient Packaging: Modelling how packages will be delivered will help with more efficient delivery box sizes and prevent damage that currently happens.
- Construction: Construction firms can assess the efficiency and strength of their current designs to design buildings more optimally in the future
- Remote Repairs: Similar to NASA twinning their spacecraft, companies can model their parts that are remote and inaccessible to engineers frequently
With several more use cases in the healthcare industry for personalized drugs and the automobile industry for connected car enhancements, digital twins are all the rage in the market. While this allows for better research funding, it also creates unrealistic expectations that sometimes keep the technology from reaching its true potential. Where this technology leads, only time will tell. If you notice any potential use cases for this technology around you, do let me know in the comments.
