Retailers love to burble on about how cool and smart their supply chains are, yet come a crisis like COVID-19, toilet roll and hand sanitizer are suddenly rarer than strong political leadership. The truth is, the weak link remains with the in-store reporting of sales back to local or regional HQ and an inability to respond to fast-moving changes through the supply chain. AI and Bots will help out next time a similar event comes around by dynamically responding to changes across the supply chain.
Bots and AI are already well entrenched in our supply chains, but not to the extent retail business leaders would like to think. Even a minor shock can impact individual stores and there are plenty of examples beyond the current crisis.
Take a local store in a town near any university. The minute the students show up, the store runs out of ready meals, pizza and household cleaner items in a day or two. And it takes “the system” a couple of weeks to catch up, leaving locals feeling grumpy and students looking hungrily and curiously at aubergines. You would think “the system” would have those dates plugged in and plan accordingly, but according to retail operators, no.
There is some seasonality, take big sporting events when crates of beer appear ahead of big games and tournaments, or Pimms or other more refined beverages for tennis and other summery spectator-driven events. But the systems remain slow to respond after the event has passed by and there are big gaps all over the store.
This partly explains why, when COVID-19 struck, even though there was plenty of toilet roll in the supply chain, it was proving impossible to get the product to the stores that needed it, along with other essentials. These supply chain systems like a neat, steady flow of data, and are not designed to react to a toxic shock event.
Enter the bots and AI
As mentioned, there are already AI systems used in supply chain, providing machine learning to improve demand forecasts with various impacts across the business. They are used, along with traditional in-store methods, to help sell promoted or high-profit items, to work out what deals or special offers are successful, but there seems to be no way for one store to provide a business alert “we’re running out of toilet roll.” or for the business to respond. Especially if all stores start sending the same message.
Most people think of chatbots as something that humans talk to in a customer service role, saving them having to wait in the call queue. But increasingly bots are talking among themselves as part of the growth in machine-to-machine business. Already, vending machines can order fresh supplies of drink when they run low, and scale this up and we will see enter supply chains talking to each other without human involvement. While a future scenario may differ in the specifics, be it an environmental crisis, political or governmental, the ideal business response would be:
In-Store Bots reporting from the EPOS system “Sudden spike in demand for products x,y,z”
Head Office AI – “Pause next delivery shipments, source more of x,y,z from local providers and dispatch as a priority”
Supply Chain AI – “Source wider supply of x,y,z, alert manufacturers of demand spike. Redirect supplies from non-affected areas”
Manufacturer AI – “Note demand spike and reprioritize manufacturing, alert suppliers of time to deliver new stock and by how much.
Supply chain AI – “Create space in the chain by deprioritizing less important items for when the new stock is available to provide smooth delivery to depots”
Head Office AI – “Prioritise more delivery capacity for items x,y,z to stores, along with other items likely to become key depending on the type of scenario.”
And between them, these bots can handle all the digital paperwork, pass around invoices, perhaps with a dash of blockchain for authentication, all before a manager has had time for their morning coffee.
The human/bot relationship
For a complex system to arrange all of that without human oversight will be an impressive achievement and something that suppliers are aiming for. Not only to respond more flexibly to events but for overall efficiency and operational capability.
Humans will be involved for several reasons, firstly to check that a few weirdos aren’t just randomly buying all of one product because they heard a rumor on social media. They’ll also double-check the numbers before physical orders are approved because they might know of other sources that the AI isn’t aware of. But overall, the AIs could deliver a workable best-effort plan within a matter of minutes, where it would normally take people hours or days to establish a chain of events and get all the necessary confirmations. In-stores, workers will also help realign the store layout to put the must-buy items together to minimise impact across each shop. Although in the further future, drone pallets or floating carts could bring them out automatically as stock depletes.
There are plenty of day-to-day benefits offered by an AI supply chain during normal operations, helping businesses keep tabs on competitors and analysing the huge amounts of big data that can lurk within these systems. But when it comes to a crisis, the bots will be able to take the reins of a fast-moving situation and deal with it faster and more efficiently than one person putting the brakes on a complex system and trying to adjust things manually.
As the supply chain industry undergoes increasing automation, the AI could take control of individual vehicle routes, or autonomous cargo vessels to get the job done faster. They could take control of automated production facilities and, at the other end, link with retailer systems. With the rise in online shopping, they can work together to handle prioritised delivery to customers by drone or delivery bot based on their need to ensure fair delivery, taking the “panic buying” out of the equation and helping retailers and governments calm people down during an evolving situation.
So, when you’re next at your local minimart or supermarket and there’s a rush on, ask your friendly store worker what’s going on behind the scenes and you might get some interesting insights.