The term digital twin is overused and misstated because for many manufacturers, it’s expensive and not accessible. Read why digital twin manufacturing is a buzzword:
- There’s the hype of what a digital twin can provide and then there’s the reality. The two are not the same.
- It’s an overused term with unrealistic expectations, making it inaccessible for most manufacturers.
- A digital thread is more beneficial for most manufacturers.
- A digital twin has gone through the “hype cycle” and has plateaued in popularity.
What is a Digital Twin in Manufacturing?
In theory, a digital twin is a digital model of something that exists in the physical world. It uses the design of a physical object and all of the data attributed to that object to create a digital model that allows for simulations to be done, simply. Basically, you’re using data science to recreate the physical world in a digital sense. The reason for this? A lot of times it can be very difficult, complex, or expensive to simulate scenarios in physical form.
The idea is to make predictions and tests easier to do by using a digital twin. But, there’s the hype surrounding the digital twin and then there’s the cold reality. The two often differ greatly.
Digital Twin Manufacturing Examples Hype vs. Reality
The Hype of the Digital Twin
The digital twin concept originated at NASA, long before ‘digital twin’ became a buzzword. At a time when space exploration was at its peak, NASA engineers struggled to figure out how to fix issues that arose once a shuttle had launched. How could they fix a problem when a shuttle was millions and millions of miles away in outer space? Enter the ‘digital twin’ idea. While the ‘twin’ was originally a physical model of the larger shuttle, it enabled engineers to look at the shuttle as a point of reference – no longer guessing as to what could be causing a problem.
For years, the digital twin lived in the form of spaceship recreations.
In the early 2000s, the term digital twin became more well known, especially in the manufacturing industry (see Siemens MindSphere IoT and Digital Twin). Companies like Boeing and Rolls Royce used digital twins for the production of aircraft components to collect real-world data and make improvements or understand where failures could happen. They followed in the footsteps of NASA, but this time, with actual digital models.
Essentially, Boeing and Rolls Royce (and similar companies) added sensors to the production lines making jet engines and collected data in real-time. Then, they used the insights gathered from the real-world data to test new engine designs, understand maintenance activities, and steps needed to improve production processes. That’s the dream of a digital twin.
However, the reality isn’t quite the same.
The Reality of a Digital Twin
Creating a digital twin is expensive, complex, and time-extensive; most companies can’t afford the cost or have the manpower to create one digital twin, let alone multiple. For all of the physical objects a manufacturer wants to look at, each object will need to have its own sensors and separate data set. Think about how complicated and costly that could get depending on how many products a manufacturer makes. Then, it gets even more difficult when you consider the part being produced – is it complex or simple? Are you producing an insulated cup or a car? The idea of a digital twin really starts to break down when you take into consideration all of those factors.
Let’s say a manufacturer makes a component of something, say a small part for a jet engine. How is it cost-effective or useful to use a digital twin for a small part?
A digital twin is really only applicable to certain industries or use cases, specifically for manufacturers who make big components of products or have capital-intensive equipment. It’s not realistic to think every manufacturer is going to have digital twins of the things they make or that they should.
It’s just not cost-effective.
Why is ‘Digital Twin’ an Overused Term?
The term digital twin is overused and misunderstood because for many manufacturers, it’s simply not feasible yet is regarded as “the future of manufacturing”. Why is the industry pushing a concept when many won’t be able to benefit?
While it can be useful in particular use cases, a digital twin is not an option for just anyone. Like we said above, it’s expensive and complex – not something every manufacturer can use. For that reason, it’s an overused term with unrealistic expectations.
Another term that’s used less (although, it can also tend to not be clearly defined) is ‘digital thread’. This term is better aligned with what most manufacturers are thinking – capturing data through the entirety of design, manufacturing, and delivery of a process.
The concept is less about a ‘digital twin’ of a physical object, and more about tying together the entire production process which for manufacturing, will provide a much larger benefit.
Essentially, creating a digital thread means complying with all of the data for a single product from start to end.
For example, you can tie quality metrics of an insulated cup into a single system so the manufacturer can understand how design changes affect the manufacturability or quality of a product, using real data. A digital thread is tying together every little piece of data known about a product so there is a history. That history of data can be used in the future to make decisions.
In an industry where context is key, having a ‘thread’ of data is incredibly beneficial.
A lot of the time, the term ‘digital twin’ is confused with ‘digital thread’. When people are referring to the concept of a ‘digital twin’, they’re actually referring to the idea of a ‘digital thread’. Confusing, right? It’s a common mistake to blend the two terms, but the confusion has led to the mass overuse and “buzzword” status of the digital twin.
The Difference Between a Digital Thread and a Digital Twin
While a digital thread is more beneficial to most manufacturers, creating a digital thread is still difficult, even though it could provide a lot of very useful, very strategic data for manufacturers of all sizes, and without the inevitable time requirements or cost of a digital twin.
More people aren’t utilizing a digital thread because typically, the systems that do it are expensive and complex to install (a lot like manufacturing analytics used to be!). Today, to create a digital thread effectively, a manufacturer would need to pull data from a variety of different systems, aggregate it together, and contextualize the data to make sense of it or find insights. Unfortunately, it’s not very accessible for many manufacturers.
That said, there is a much larger opportunity for the development of digital thread software than there is for digital twin. It’s possible for the software to be developed that makes the creation of a digital thread accessible and affordable for all manufacturers. For the manufacturer’s sake, we hope that’s the case. (We believe there is great benefit in the use of a digital thread.)
Digital Twin Technology is and Continues to be a Buzzword
Rounding back on the digital twin – it was and continues to be a buzzword. It made its mass entrance into the world of technology around 2018 and continuing into early 2019 – the topic of many presentations, the next ‘big thing’, but that fanfare has largely died down due to the realization it isn’t an option for every manufacturer.
There’s this idea of a “hype cycle” coined by Gartner. Essentially, when a new technology is developed or rising in popularity, there is an “innovation trigger” that causes it to reach a “peak of inflated expectations”. Then, it drops to a “trough of disillusionment”. Finally, through the “slope of enlightenment”, it reaches a “plateau of productivity”.
This is exactly what happened to the digital twin term when marketers, IT, and manufacturing professionals alike realized it couldn’t live up to the hype, at least not at this time.
It will take someone coming in with a new vision to make it more applicable and accessible for everyone. Until then, a digital twin will continue to be a buzzword that will remain out of reach for many manufacturers.
