Automated Distribution Centers: Justification in a Same Day World
By Dean Starovasnik
It used to be that we located DCs in the US based on a desire to minimize the total logistical costs of providing a specified service level to a certain percentage of customers within a defined number of days. But now, in many industries, delivery in a number of DAYS has been replaced with a demand for delivery in one or two HOURS! Consequently, there’s a shift in strategy being considered by many companies regarding the size and number of DC locations required to fulfill orders in this ‘same day world’ of delivery.
What does this have to do with automation in the distribution center, my usual topic? Well, quite frankly, the shift in strategy can make justifying automation tougher.
Why would a strategy that positions your DCs closer to your customers make automating the DCs more difficult? On its own, such a strategy doesn’t. However, it typically requires that you locate more DCs in metropolitan areas around the country in order to reach such an aggressive service level. More DCs reduces the volume in any one DC to a fraction of the total. That’s a problem because, in general, the higher the volume in a facility, the easier it is to justify automation. And if you can’t automate, by extension, it also means you will need more labor.
But wait! Isn’t labor, its availability, cost and quality one of the top three issues facing distribution leadership? So, are we working ourselves into a bigger fix by making automation even more difficult to justify? Yes. We are. So, what do we do?
As usual, there are a couple of ways to address this. The first is to recognize that the automation necessary to reduce the labor content in lower volume facilities cannot be justified solely by labor cost reductions. Rather, just as a website, customer service representatives and advertising are all necessary to generate sales, so too is the infrastructure necessary to deliver the products once they’re sold in the time required to satisfy customer expectations. So, evaluating automation based on its ability to meet broader demands is the first step.
Secondly, identifying automation options that are scalable is essential. Normally, when we say “scalable” we mean devices that can handle increasingly larger volumes without dramatic alterations or exponentially increasing costs. However, now we need devices that are scalable downward, where the first few elements, the first few pieces aren’t the most expensive. You see, when a sorter or some other “whatever” is designed into a system, the first destination is the most expensive. This is because supporting this destination or “whatever” means buying the induction, the scanner, the software, any prime mover and such; all the things needed to create the desired behavior. Every additional destination or “whatever” further amortizes those initial costs making the 20th “whatever” a fraction of the cost of the first.
However, the issue now is that we need automation without the high upfront costs. We need to be able to scale downward to the first five or ten “whatevers” without the costs increasing dramatically. And while this is easy to state, it is immeasurably harder to achieve.
So how can we realize this objective? Well, this is just a blog, not a research paper. And while I do have some ideas, I’m not quite ready to postulate an answer – just yet. If you’re facing this challenge and have some ideas, I’d welcome the opportunity to discuss them with you. Just drop me an e-mail at firstname.lastname@example.org. Hopefully, working together, we’ll get all of us closer to a solution.