Digital Twin - Pathway to Automation

The Digital Twin brings a generational change to the way software addresses business  


A Digital Twin is described as a digital replica of real-world entities. What does that really mean?

Digital Twin technology originated at NASA, from their experience with Apollo 13. Today it is most commonly used for engineered artifacts such as jet engines and intelligent buildings, but the same logic applies to business - its structure, activities and resources.

Using a Digital Twin for business compared to traditional enterprise software is like viewing a video versus a photo – one lets you picture what is happening, the other is a snapshot of what happened at a point in time. We call it 3D versus 2D, where the third dimension is time.

A Digital Twin provides a richness of information that is vital when automating business activity. In short, it will change the way business interacts with computing.

Automation needs Digital Twin

Automation can be truly transformative. if you call a ten-fold productivity increase transformative.*  One of the reasons its potential is not being seen is that it relies upon what the 3D Digital Twin brings: the integration of processes across an organization and a 100% accurate definition of business.

An integration framework shows how processes fit together in a value chain. Its absence means that automation projects focus on creating simple unconnected processes such as replacing spreadsheets – termed “islands of automation”.  One Fortune 100 CIO described processes that lack an integration framework as ’like a plate of spaghetti’.

With 2D software, process definition usually falls short of the 100% accuracy required for successful automation. It’s like a high-speed rail line with two metres of track missing: the whole system fails for that single shortcoming. An automated process can be programmed to make a decision based on two options - but if the situation doesn’t fit either option the process halts. Getting workers to reveal unexpected conditions requires the deep level of engagement engendered by a Digital Twin.

Real world is the key

Until now, software development has created conceptual solutions to real-world situations. Focus has been on data and trying to extrapolate the real-world using data architecture. But in the real world, activity has primacy. New staff will first ask what to do before enquiring about what resources (such as data) they should use. In the real world, without activity, data is static. Once activity commences it needs data to support it, then it generates data from its actions.

By replicating the real world of an organization a Digital Twin makes it easier for people to relate to what they do and to describe this when processes are being defined. Uncertainty can be addressed by seeing or touching the real-world equivalent. There are no Actors in the factory, no Swim Lanes in the office. Digital Twin refers to real world artifacts and actions using the language of the real world.

It cannot be understated that the quality and completeness of a digital definition of business activity directly affects the success of automation. While some in the 2D world think ‘80% is good enough’, the further away a process is from 100% accuracy the worse the outcome, be it automating, making improvements, customer experience (CX) or just understanding the business.

When it comes to quality, 3D Digital Twin delivers vastly better results than 2D software.

Digital Twin is a model solution

Saying Digital Twin is a generational change in software solutions delivery requires an explanation.

Digital Twin software is a prebuilt model of business. All that users have to do is to populate it with the specifics of their own organization. The model contains rules that define how business behaves and how its components relate to one another. It requires no software development.

Perhaps the most important feature of a model is that it enables rapid change. It is self-validating, which means that it checks the validity of new processes or changes as they are defined. Testing isn’t required and this means a small change can be made and put into production within an hour.

In fact this feature is the best way to ensure a 100% accurate process definition. It is not possible to get a truly accurate definition from what people say, but once a process is automated the errors and omissions quickly become apparent. With conventional software development this would be highly problematic as with every stoppage the software has to go back to the development group. With a Digital Twin model fixes can be applied on the fly, typically leading to fault free systems in just days.

Digital Twin Outcomes

A Digital Twin’s components are aware of the state of other components. For example, if a worker is not available the model knows what activities are affected. In a business restructure in which tasks or even complete processes are being eliminated a model shows who is affected. And if headcount is being reduced the model shows which tasks and processes will be impacted.

This is a great example of where Digital Twin automation comes into its own. Because the software replicates real life, work can be reallocated at relatively low risk. Automation presents work to be done in a sequence based on the decisions that it makes so staff don’t need to memorize what to do next, and it can provide step-by-step instruction for completing work.

The competitive landscape

As automation becomes increasingly strategic, business is turning to Process Mining software to try to convert their 2D view of business to 3D by deducing the sequence of a process workflow from the changing state of the 2D data. This cannot deliver the required level of process accuracy. Workflow has to be specifically defined, not deduced. Process Mining can’t detect what it doesn’t know about, such as manual activities and the exceptions that derail automation.

With a focus on real-world activity, the Digital Twin model provides a far clearer and comprehensive approach to understanding process workflow. This is what finally enables automation to address the value-chain activities of business - demand, value-creation and supply. A stark contrast to the simple 2D islands of automation initiatives.

What is in a Digital Twin model?

First, a Structure in which to operate – an organizational hierarchy of people (as in positions and groupings e.g. departments, companies) and locations (as in relative space e.g. GPS).

Then comes Activity – operating within the structure, passing from position to position as the work is completed (defining the flow of work between people and the sequencing of tasks at each stage).

Finally there are Resources used in the activity – some consumed and passed on, such as materials, others ‘rented’ for all or part of the activity, such as staff, facilities or equipment. External entities such as customers and suppliers are also resources, although neither consumed nor rented.


When asked to compare Digital Twin to products that purport to automate or define business, our response is simple. Is the product 2D or 3D? Can it capture reality? Because without a systemic picture of reality how will you a) understand, b) simplify and c) automate your business? This is what sets the Digital Twin apart - a generational change to the way business employs software.

* One of the results experienced by a clothing company using XSOL Digital Twin Automation