Digital twins – Introduction

“Digital twin refers to a digital replica of potential and actual physical assets, processes, people, places, systems, and devices that can be used for a variety of purposes” – Lee Beadmore, Vice President and Chief Innovation Officer, Capgemini’s Business Services

 

Necessity is often said to be the mother of invention, and the story of Apollo 13 in 1970 is a prime case in point. Watch Ron Howard’s film of the accident, and you’ll see.

After lift-off, an oxygen tank explodes in the side of the spacecraft, depleting not just its oxygen supply, but its power. The ground crew issue guidelines to the astronauts, but those instructions prove to be irrelevant, because they bear no relation to the real-world circumstances.

At the Mission Control Center in Houston, a fellow astronaut realizes the problem, and organizes a team to replicate as exactly as possible the conditions being experienced out in space. They equip their own version of the spacecraft modules only with the tools and materials available to the Apollo 13 crew. Between them, they develop a new understanding of the issues, and find ways round problems that aren’t in the manual.

The ground-based Apollo 13 replica was effectively a physical twin of the active spacecraft, enabling experimentation in a safe, offline environment – and now, almost 50 years later, we’re increasingly seeing the development of non-physical, digital twins for the same purpose.

Digital twins are quickly becoming established in IoT heavy domains. In manufacturing, for example, they enable planners to gauge the effect of changes in production runs before taking them live on the factory floor. Digital twins can be used to avoid bottlenecks through problem prediction, increasing efficiency, and reducing downtime.

What is perhaps less obvious are the potential benefits digital twins can bring to the information processing domains of finance and accounting (F&A), human resources (HR), and supply chain management (SCM). In this realm, the data-heavy process itself is the asset – in effect, it’s a production line for processing data.

And, of course, as we enter the AI-infused era, digital twins provide a playground for human and artificial intelligence (AI) minds to meet, pulling in the data needed to train AI models. When fully harmonised with the physical world, the impact will be ground breaking.

We have already explored this topic in Capgemini’s TechnoVision as a fundamental building block called “Twin Worlds.” In this paper, I will build on some of these ideas and add a perspective on what can be addressed, and what we can expect to achieve.