The term AI has been around for years and most companies have been leveraging analytical techniques for just as long in order to improve their business. So, what has really changed to make AI a hot topic and more importantly, how can manufacturers benefit today?

Let's go straight to an example of improving the production yield - the illustration below captures how you can (today), efficiently analyse different types of data from 5M factors :

And find patterns/benefits you couldn't before -

So, what makes the above analytics possible ?

1. Speed: the processing cost/speed of collecting and analysing data has come down significantly which means you can consider data sets far bigger than the traditional methods you would use.

2. Processing unstructured data: now you can take into account data such as voice or images as well as words and numbers and hence understand the full context for making a decision - which you could not really do before.

3. Scope: you can now combine data from different functions or industries, which means you can study problems and their contexts in a way that you couldn't in the past.

As the Mckinsey report shows, the manufacturing and supply chains could be amongst the biggest beneficiaries from leveraging AI. 

The ideal scenario of course, is to combine the technical skills with a strong domain knowledge, in order to create "Machine Learners" (as below) whose output business users can understand. We know many forecast/supply optimisation projects never took off, simply because the organisations could not align the necessary resources and capability in terms of people and processes.