![ONLINE SHOPPING SHEIN AMAZON EBAY iStock-Robert Way-1499148111.jpg A HAND HOVERS OVER A SMARTPHONE SCREEN BEARING LOGOS OF ONLINE SHOPPING SERVICES, INCLUDING SHEIN, AMAZON AND EBAY.](https://www.supplychainbrain.com/ext/resources/2024/04/29/ONLINE-SHOPPING-SHEIN-AMAZON-EBAY-iStock-Robert-Way-1499148111.jpg?height=100&t=1714396546&width=150)
Visit Our Sponsors |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
For most distribution operations, the application of artificial intelligence is a journey that is far from complete. But leading companies are well on the road to that goal, according to Brad Radcliffe, vice president of sales, sortation and distribution with Beumer.
SCB: What are the primary applications of artificial intelligence in the modern distribution center?
Radcliffe: I don't know that we're actually there yet. The journey has just started. I think everybody recognizes that this is a transformative time for supply chain, and that the need to make faster decisions to meet the expectations of the customer has grown exponentially.
Twenty 20 years ago, it was OK to make decisions in a vacuum without all the information — it wasn't as impactful. Now, most companies recognize that they have to start the journey. A.I. is the end state where we can make decisions in microseconds, taking in multiple variables. There are massive amounts of data that have never been part of the decision cycle.
SCB: It sounds like A.I. is essential in the age of big data and the internet of things. Otherwise, how can we make sense of this massive flood of data that we're being inundated with?
Radcliffe: You're exactly right. Companies are investing a tremendous amount of resources to figure it out. To move from collecting data to making decisions to true decision engines that drive efficiencies in your supply chain is an interesting challenge.
The first step in the journey is what we've labeled big data. It's taking disparate pieces of information from multiple databases, and pulling it together in a format that starts to make sense. The processing power and software to make it possible are coming to fruition.
SCB: In the early stages of A.I., we're asking the system to provide us with information for humans to make decisions. Ultimately, do we expect A.I. to be making those decisions for us, or will there always be humans to do that?
Radcliffe: In my opinion, there is always going to be a human in the loop to set the business rules and decision criteria — but decisions have to happen in microseconds. Once you've set the parameters for the decision engine, you've pretty much got to trust it to go forward. There will always be human interaction, it’s got to happen in real time. It has to be instantaneous.
SCB: What’s the role of A.I. in the next generation of fulfillment and sortation systems?
Radcliffe: The three biggest cost drivers for any supply chain are inventory carrying cost, transportation and labor. We’re trying to understand which decisions in the cycle impact those three things. If I can make decisions faster, based on historic data as well as actual demand patterns, then I can start reducing overall inventory levels and transportation cost, while making sure that product is in the right place at the right time.
Within the four walls of the warehouse, we’re trying to impact the labor component. The role of automation is to make labor more efficient. Ultimately, decision cycles have to be faster and faster. Inbound receiving, storage, replenishment, picking: all of it has to happen flawlessly. By optimizing an automated solution, we can understand the local and regional impact of demand. That influences the use of labor in the warehouse — when I fulfill an order and how I move product around.
SCB: To some extent, hasn’t sortation long been highly technologically enabled, with the help of sophisticated conveyor systems?
Radcliffe: Upper-level sortation has always been a little bit smarter. We've built algorithms and decision engines within our sortation process. It's just that now we're getting better at helping people understand the journey we've been on. And we’re trying to understand how we take that next step, to optimize labor upstream and downstream. When I look at the same machine used in hundreds of locations, I start getting predictive analytics based off run times and heat maps. Instead of waiting for something to break down, I can go out and troubleshoot, and get ahead of the cycle.
SCB: The system is telling you that something is going to fail?
Radcliffe: Right. And that's the easiest part. The next is trying to marry the sortation operation with the demand piece — customer orders coming in — and getting ahead of the cycle.
SCB: The job of sortation has become so much more complex in the age of e-commerce. You have so many small packages being delivered direct to customers. That’s got to be an entirely new challenge.
Radcliffe: It is, and that is what’s driving this transformative activity. It impacts every aspect of the supply chain, from manufacturing to transportation to the four walls of the warehouse to the final mile. You’re going from a retail environment where you might have 30 or 40 items per order, to one with similar volume but eight to 10 times the number of orders. Anytime you have a unique order, it increases your touches. It's a different labor component, and I have to be faster and more efficient.
SCB: Give us some guidelines as to how companies can evaluate A.I. for their operations.
Radcliffe: It's like any application of an automated solution. You don't apply automation or A.I. unless you can first define what the problem is. You start with the end in mind. The next step is to decide whether you want to build an internal skillset, or go out and lease it — find a consultant who understands big data and its role within your warehouse. You need to do an exhaustive amount of research to pick the right partner who will help you execute your vision.
RELATED CONTENT
RELATED VIDEOS
Timely, incisive articles delivered directly to your inbox.