Digital Representation & AI

We have already walked through the door in a period of great change in the way we build and use new applications in supply chain. There are two main drivers that are powering this change: digital representation and artificial intelligence (AI). They are both closely connected, and while the more cautious operations research (O.R.) professionals may feel like their skills may be aging, business is, in fact, moving toward O.R. and bringing with it a fresh language and some cute packaging that will likely pave the way for far wider adoption.

What is a digital representation? Digital representation is a simple concept at face value. Build a model of your supply chain in a suitable tool to accurately represent the various facets of your business, such as production assets, storage and value-add assets, capabilities and capacities, suppliers and customers. The model can also utilize a digital twin of specific production assets to allow deep analysis of areas impacting the machine without disrupting the machine itself. Sounds simple, right?

Now throw in a requirement that the digital representation must be structured to be leveraged across any number of applications, technologies and mathematical techniques to improve supply chain decisions in ways that may not yet be defined. This is possible. We know because we’ve done it, and it requires a new generic mindset: “What data do I need to solve any problem in the supply chain?”

What does this mean for O.R. professionals? The implication of digital representation is that there will be a voracious appetite for new applications from the businesses with a very explicit requirement that they meet functional needs and are connected to other apps both vertically (across functional areas) and horizontally (across multiple planning horizons, from years to real time). These applications will all feed and be fed data from the common digital representation, so that they are faster to deploy and more powerful in their output. Users will experience a common design language across a diverse array of apps. Equally, traditional apps that use a dedicated data set will find themselves orphaned and often targeted for retirement.

What about AI? AI is clearly hot right now. In fact, it's overheating to the extent that it’s becoming a hammer looking for a nail. Many companies are looking to embark on AI pilots when instead their business would be better served identifying important use cases that are now technically within reach. The digital representation can greatly help us here; it does not care which mathematical techniques are leveraged across its framework. A number of them can be leveraged simultaneously.

An example of this is sales and operational planning, an exercise to have demand stakeholders collaborate with the production and fulfillment side of the business to create great plans that increase both revenue and margin. Digital representation allows these stakeholders to use predictive techniques to create a statistical forecast. They can leverage probabilistic reasoning in the form of machine learning to normalize inherent optimistic/pessimistic traits in event planning and capacity predictions and optimization to drive a multitude of scenarios to expose an agreed plan – all enabled across the same digital representation. This is a high-demand use case that will have a measurable impact on revenue and gross margin, while piloting various facets of AI. 

Summing Up

These two major changes are going to make users much happier through the power of insight. The work of application developers and deployers will be far more interesting as the work becomes more creative, with less time sacrificed to endless data collection and clean up.

Gartner emphasizes the importance of digital representation and AI, yet suggests that it will take some years for even large organizations to adopt these techniques. Business is definitely moving in the direction of operations researchers, and given your incredible skills, I suggest you turn and run toward it so we can all enjoy these new capabilities, sooner rather than later.