Keith Reilly - Digital Twins

With a digital twin, you can conduct high end simulations using real time data from your real production line and make changes to the twin to find out exactly what will happen without incurring the cost of shutting the environment down.”

“There is widespread agreement in manufacturing that digital twin technology is enormously important, but at the same time, there’s also disagreement about what exactly it is. I find that fascinating. For us, at least, digital twins are real-time replicas of physical production environments.”

“They don’t need to be exact 3D images of an entire environment, just the bits that matter. Ultimately, we’re more concerned with what we can learn from a twin than with its completeness.”

Keith Reilly is a product design engineer with DMI, and he has a deep understanding of the role that digital twin technology can play in delivering competitive advantage in manufacturing. But he’s also a realist, and he knows that, like any tool, it’s only as useful as the experience of the people who wield it.

“In a non-joined-up production environment, it’s common for data from machines to be downloaded every couple of weeks or even once a month. That historical data is then analysed to try to glean insight. However, because a digital twin uses a constant stream of real-time data, it is in every respect an exact copy of the real production line it’s based on,” he said.

“For example, let's say you have a production process, and you are considering making some changes, but you don’t know how they might fully play out. With a digital twin, you can conduct high end simulations using real time data from your real production line and make changes to the twin to find out exactly what will happen without incurring the cost of shutting the environment down.”

The idea of creating virtual models of complex production processes is not particularly new. In the 1950s, mainframe computers were used in the aerospace and automotive industries to model how airplanes and cars interact with the world and by the 1970s, computer aided manufacturing (CAM) was already being used to design and build virtual prototypes of all kinds of products.

But as technology has progressed, so too has the range of possibilities that exist with what can be done through modelling.

“Computer modelling in the past was limited by its use of historical data and a disconnect that existed between the software and the real world of the production floor. However, the advent of increased computer power and the decreased cost of sensor technologies means that the state of the art manufacturing facility of today is a much more joined up place,” said Reilly.

“And with all the data generated by these digital ecosystems, it has become possible to create a new kind of model, one which isn’t working retrospectively but with real time information generated by a physical environment.”

Like when any new tool becomes available, it’s important to ascertain whether it will add real value. Reilly points out that it’s possible to create extremely high fidelity digital twins, complete with on-screen, AR or VR CAD environments where a viewer can inspect every nut and bolt used to create a production line.

“Seeing a digital twin like that in operation is undeniably cool, but is it going to add real value, or will it just over complicate things?"

"Part of our role is to identify where tools like this can deliver real value by knowing where gathering real time information will pay actual dividends,” said Reilly.

“Because we are technology and vendor agnostic. We have a lot of freedom to start with a customer’s problems or challenges and then help them decide on the best solution, rather than starting with a certain toolkit and trying to squash that into whatever solution we recommend out.”

When it comes to manufacturing technologies, Reilly has some interesting thoughts. To start with, he is firmly convinced that the industry is at the start of a new era of data utilisation.

“More and more companies are producing larger quantities of data, and the concept of big data has been around for some time. But what is likely to change and even become a defining feature of the next few years is a maturing of the way companies actually make use of this data.”

“We’re seeing a new drive from companies that have large amounts of data asking our data scientists to help them uncover hidden trends from it. They want to know, can this data be used to help inform decisions in ways that weren’t possible before?”

He also reports a mainstreaming of additive manufacturing or 3D printing.

“The actual tech has been around for a long time, but we’re seeing it being used with more confidence now. It’s really powerful for specific use cases, such as, for example, in the concept of ‘batch of one’, which allows manufacturers to offer bespoke options and greater flexibility.”

This is part of a general overall trend Reilly identifies in technology that sees processes and production systems move from ‘slow, large and expensive’ to ‘fast, small and cheap’ as adoption rates increase, and economies of scale come into play.

“I think it’s also fair to say that there is a significant amount of re-shoring going on, where factories that moved out of Ireland and away from Europe in the past are returning. Because of that, there’s an opportunity for manufacturers based here to make products that maybe they haven’t made before.”

“The reason for this is that in the past, certain kinds of manufacturing moved away because of cost considerations. It basically became cheaper to do things elsewhere, but as time has passed, the situation has changed. Ireland has a much higher skill base now than it had, and in addition, certain countries that offer low cost manufacturing have become less politically stable over time.”

Brexit, the Covid 19 pandemic and even the blocking of the Suez Canal in 2021 showed how delicate and vulnerable international supply chains can be. It’s no longer possible to be absolutely sure of being able to ship goods from around the world without delays.

“And also, there is growing awareness that possibly shipping a product from one side of the world to the other to be assembled and then shipping it back isn’t a great idea. It might be commercially viable, but concerns around sustainability and carbon footprints are becoming more important too.”