Arrow DownArrow ForwardChevron DownDownload facebookGroup 2 Copy 4Created with Sketch. linkedinCombined-Shape mailGroup 4Created with Sketch. ShapeCreated with Sketch. twitteryoutube

Frontier decision-making is here

Find out more

What is a computational twin?

At first glance, a computational twin may sound like an alien concept, yet things like them have been around since the 1960’s.

It’s only now though they are starting to show their promise for most organisations.

So, what are they?

Simply put, a computational twin is a digital version of operational processes in a business.

They are also fundamentally different to a digital twin, which is a virtual representation of an object or system used for anything from understanding how a building’s airflow works, to overseeing a manufacturing process. 

Computational twins are broader than this. They look at your end to end process – for example, your entire S&OP – and tell you everything from what is influencing customer behaviour, through the manufacturing process, to where to send individual SKUs. 

A computational twin captures key organisational processes to make predictions about what will happen in the future. They use machine learning and other techniques to help decision-making. 

At a basic level, a computational twin can operate just as a digital twin. Instead of a supply chain manager having thousands of spreadsheets detailing how their manufacturing processes operate, a digital model will show them how it’s running in an integrated dashboard, and automate things if desired. 

But a computational twin goes further. It lets businesses understand the present and predict the future. 

They are updated from multiple sources of real-time data, and integrate with existing systems. With the right type of advanced machine learning, they tell you what is happening now, what will happen, and crucially what should be done next. They give businesses that precious commodity – decision intelligence.

What’s changed to make this possible?

Well, we’ve had computers since the 1950’s – and people have regularly tried to create digital representations of organisations.

These didn’t work very well though, because businesses have complicated processes that are hard to describe in a small number of rules. But over the last few decades we’ve got much better at machine learning.

At a basic level, machine learning involves learning the rules of how a system works from data. So, rather than having to programme the rules by hand, you can learn them directly from your data.

This is game changing for digital representations of organisations, since it means processes previously impossible to codify are now possible – through machine learning.

The result of this process is a computational twin.

How do they help businesses in the real world?

Well, say you run a manufacturing business, and you want just the right amount of stock.

You want enough stock to serve your customers, but not so much that you have lots of cash tied up in inventory, or too many old products if there is a newer one to ship.

Getting inventory right depends on correctly predicting customer demand and manufacturing capability.

Connecting these into a computational twin means you can immediately understand the repercussions of demand, and manage inventory to a level where you balance risk and reward strategically.

Ultimately, a computational twin lets you test different outcomes, and see the reasons behind those outcomes.

When embedded in your business, you can change one element of your process or workflow, there and then, and test the result. You can do true scenario planning with predictive accuracy of around 98%.

There is no impact on your day to day activities. There is no need to “reverse” your decision if it turns out to be wrong. Computational twins allow for intelligent, evidence-based decisions – and mean there is no need to rely solely on the past to predict the future.

Computational twins improve organisational performance by enabling better decision-making.

But this is not an ‘off-the-shelf’ product.

It requires a combination of technology, proprietary software and algorithms, aligned with a deep understanding of how to embed a computational twin in a business and build user trust. The end result is an optimal combination of human and machine intelligence.

To find out how our decision intelligence software, Frontier, builds a causal model of your organisation through computational twins to optimise organisational decision making, click here.

To find out more about what Faculty can do
for you and your organisation, get in touch.