Computational twins; the key to merging operational and strategic planning
Organisational performance is the sum of decisions a business makes - both large and small.
The problem is, often a local decision looks optimal at that level, but can be sub-optimal when you zoom out to the operational or strategic level.
When you make a decision, it’s all about context - but that can be difficult across a supply chain or business process. Especially if it is a global enterprise.
Making a sequence of good operational decisions compounds into growth; the ability to do so repeatedly results in a sustainable competitive advantage. A series of bad decisions does the opposite.
So how should businesses optimise their decision making? The answer could be computational twins.
Computational twins are models that look at whole business processes (specifically operational processes). In the simplest terms, they tell you what is happening in your business, in real time, and by analysing multiple data sources.
That is different to a digital twin, which is a virtual representation of a single object or system. A digital twin model of a train, for example, would model what the object does digitally, so performance can be optimised and improved. Both digital and computational twins use machine learning to help decision making,
But it is not enough to simply have a computational twin and forget about it.
It is a representation of a process. It must be optimised to get the best results.
Computational twins are just like a human, and therefore require users to work with the twin to test, experiment and predict different outcomes. It is a striking example of human and machine learning together.
When optimised, computational twins tell you the impact of a business decision, and why that change will happen.
Crucially, it does this without impacting on your day-to-day business activities. You can test and iterate in the background. You can even automate many of these micro-decisions. The machine will quickly learn how to improve them, freeing you up to apply that human plus machine intelligence to the more complex decisions.
It is decision intelligence that really adds value and helps businesses grow - and it is one reason why Gartner predicts that by 2023, more than 33% of large organisations will have analysts practising decision intelligence.
It is why the very best, most profitable businesses don’t have people running around with spreadsheets, second guessing what will happen by looking back on the past.
They have custom built, digital models to tell them what is happening, what will happen, and why that will happen.
They have decision intelligence, giving them a description of what is happening, a prediction of what will happen, and a prescription of what they should do next.
It is decision intelligence that allows businesses to merge operational planning with strategic planning.
For example, this means a manufacturing business knowing not only how much stock they need to service their customers, but also avoid having too much cash tied up in inventory, or too many old products when there is a new product ready to ship.
Both operational and strategic scenarios can be tested and planned for, thus improving organisational performance by enabling better decision-making.
At Faculty, our technology, software and algorithms, and understanding of how to embed in a business and build user trust, means we can combine the best of human and machine intelligence. We take a scientific approach to demand forecasting to show not just what is happening, but why.
Knowing what will happen is helpful but knowing why something will happen is fundamental to any good decision making.
Author:
Ilya Feige, Director of AI, Faculty