More and more supply chain leaders are considering incorporating artificial intelligence (AI) into their business, but are still skeptic. That’s because they’ve been on the receiving end of too many sales calls where companies say AI can change the world and magically do all sorts of things.
It turns out that it's easy to sell on hype, but it's harder to actually deliver results. AI isn’t a crystal ball, but it can be a powerful tool for organizations throughout the supply chain. From deciding what to keep in stock to absorbing the shocks that are bound to happen, here’s how a flexible AI can help meet your business goals.
Providing advanced warnings
Our brains can only handle a limited amount of information. It’s not our fault; that’s just how we’re wired. We struggle when trying to make a decision that involves more than 4-5 factors. And, because the global supply chain is so complex, managers often have to account for more factors than that.
For example, let’s say a customer had 40,000 SKUs and five purchasing managers. That means each person was handling just under 10,000 SKUs each. There’s no way they could possibly pay attention to each one.
A well-trained AI system, however, can handle that level of detail. It can see that a shipment has been stuck at customs for three days when it was supposed to clear in eight hours. And, it can notify the purchasing manager about the issue and the risk of missing the delivery date. Rather than be caught off-guard later on after the delivery date comes and goes, managers can get 3-4 days advance warning to proactively take action.
Deciding what to keep in stock
The past 18 months have put an unprecedented amount of pressure on businesses to manage supply and demand. Shipping is overwhelmed, and supply shortages are everywhere.
The industry saw this problem play out with a retail automotive group focused on a particular type of service, with other competitors that try to get that same business. The question they faced was, we know we’re going to stock out of some products, so which ones do we let go?
Using AI models, they were able to double-down on the main product and let everything else stock out. The idea was tied to the importance of customer loyalty. They spun the shortage into an opportunity by focusing on the core products that people loved, and those were the ones that kept them coming back.
Defining a new equilibrium
Before the pandemic, the toilet paper market was pretty steady. But, no one, not even AI, could have predicted how demand would spike in those early days. However, where AI can be helpful is in mapping out what future scenarios could look like, and under your best assumptions, what your business’ best move would be.
By looking at “what ifs,” even those that seem unlikely, AI can set a new equilibrium and help organizations make smarter decisions when the unthinkable happens. In the case of toilet paper manufacturers, AI can examine whether it makes sense to retool everything in the process or retool the 20% it takes to meet the initial demand with a projection that everything else is going to return to a base level.
Absorbing the shocks
When things are running smoothly, most business’ supply chains are in a fairly steady state. You might sell 5,000 units one month, 5,500 the next and 4,500 after that, but it’s generally pretty predictable.
The challenge that the supply chain faces is that all of that falls apart when shocks come through. All the nicely controlled, just-in-time production systems are thrown into chaos. The result is that companies fall back into manual decision-making. They go into triage mode and begin guessing to find the right next step. But, people aren’t good at making thousands of interconnected decisions on the fly, and that’s where automated systems can come in and help steady the ship.
One AI horror story from the supply chain world comes when a consulting group told a global logistics company that the AI ran overnight and that it could tell them where every parcel was going to be, where every trailer was going to be, by the minute, throughout the day.
That’s ridiculous, of course, because it shows a complete lack of understanding about the supply chain industry, where conditions change by the minute. If someone calls in sick or even takes a restroom break that wasn’t accounted for in the AI model, the entire model breaks down.
It’s possible, however, to take a more flexible approach, updating those AI predictions as often as every 15 minutes. That’s the type of AI that learns on the ground and adjusts accordingly. It’s not so much a look into the future but a dynamic tool that reacts minute-by-minute.
To make the most of any AI tool, companies must put their strategy first. Leaders should agree on what they’re trying to achieve and then apply AI to help them reach that goal. Rather than concentrate on which parcel ends up at which scanner at a precise time, shift the focus to reducing overall costs and increasing operational efficiency. If you set a higher-level goal and hone in on the opportunities, you're often going to get better results than trying to get everything perfect.