Digital Twins & the World of Warehouse

Data Science
Back in the late nineties the only people who knew how to make really good computer models were building them to fight dragons in.

The graphics were clunky and the dialogue laughable but these early 3D worlds were the start of something huge. Fast forward 20 years and virtual models are used everywhere, from skyscraper design to Hollywood blockbusters to big business enterprises, and those nerdy computer modellers are selling their expertise for big bucks to make it all happen.

What’s a Digital Twin and how do I use it?

A typical warehouse will catalogue its inventory in a database, they’ll be able to tell you where everything is and how much of everything they have, but they won’t know how long it will take to pick an order or when a product is about to be overstocked or out.

With a Digital Twin you build a virtual warehouse and use the systems you already have to populate it, you can monitor deliveries and shipments in the real warehouse and add that data as it happens. If you can track your people as they move around then they go in the model too. Ultimately, you can watch your virtual warehouse functioning in real time, all from the safety of your desk.

Digital Twins aren’t just limited to warehouses though, as businesses finally start implementing their internet of things we’ll see the opportunities to model and predict increase exponentially. Once your merchandising, stocking and distribution systems are in sync and visible you can eliminate almost all the uncertainty and bullwhip through your supply chain. So long as your model is accurate you can use predictive algorithms to work out how to respond quickly and set long term strategies.

Imagine being a fashion retailer, one of the Kardashians posts a photo of herself in an interesting new outfit and before the day’s out you’ve found a supplier, added ‘the look’ to your website, updated your back office, stocked your warehouse and are shipping orders to the masses. All before your competitors have even decided what to do.

Some more examples:

  • A digital twin of your store will show you what’s making it to the changing rooms but not the tills - perhaps it looks nice on the rack but fits poorly… note to supplier.
  • Your twinned distribution network tells you a delivery van has broken down - if the nearest two others go pick up the stock you could still fulfil most of your orders today.
  • One of the complicated printers you lease to an office is reporting irregular stats and has a 75% chance of breaking down in the next six months – There’s an engineer in the area next week, better have him pop in for a check-up.

Having your Digital Twin or Internet of Things on hand makes it a lot easier to identify and respond to issues or make basic improvements to your processes, but most critically you can now start probing the model. You can configure future events and speed things up to find out what’s going to happen. If you let your data scientists loose on your Digital Twin they can run all kinds of experiments to optimise your business, and a suite of AI algorithms can feed your people suggestions and make links that human minds can’t see.

This is going to cause some serious change in the way we run our businesses. Experimentation will replace speculation as the basis for predicting ROI and decision making, our business cases will be stocked with regression values and peer reviews and we’ll all be expected to speak the language… If you don’t then you won’t be able to use the science, and only those that can will reap the rewards.

So get started now, change your culture to promote creative curiosity, get your recruitment teams looking for inquisitive and playful minds and build a huge Digital Twin world for your geeks to explore.

If you’re interested in this kind of thing check out our video on 5 ways data science is creating the future.

Joe Chamberlain