Deciding where received goods physically end up is partly possible to take over with AI. The calculation behind that decision — which location matches rotation speed, space and product type — lends itself well to software. The physical movement of boxes, pallets or crates to that spot does not. That distinction determines the entire answer.
The task itself is highly structured: a wms can calculate a location using fixed rules — rotation speed, dimensions, shelf life, compatibility with neighboring products. This is not creative work, nor is it a task with much room for judgment: the logic can largely be captured in rules, not something that needs to be reconsidered case by case. Compliance also plays a minor role: there are rarely legal obligations that require a separate assessment per storage location, except perhaps for hazardous substances or refrigerated goods.
What tips the balance towards 'partly' rather than 'fully' is the physical side. The volume of this task is high — every pallet, every box, every day again — and that volume makes automation attractive. But the execution requires physical actions: picking something up, moving it, setting it down, often in places that are not always accessible to a robot without adapting the warehouse itself. As long as that physical step is human work, oversight on the floor remains necessary too, even though the calculation is automated.
A warehouse receives a pallet of fast-rotating items. The wms calculates, based on rotation speed and available space, that this should go to a location close to the dispatch area, rather than at the back with the slow-moving stock. That calculation happens automatically, immediately upon receipt. A warehouse worker gets the location on their scanner and physically moves the pallet. The thinking work has been taken over; the lifting and driving has not.
At a company with an automated warehouse — with shuttles, AS/RS systems or other robotics for physical movement — the balance shifts. There, the entire cycle, from calculation to placement, can proceed without a human hand. At a company without a wms with location logic, or with an assortment that varies greatly in shape and weight, it is the other way around: then the calculation too remains partly human work, because the rules cannot be properly captured. So the outcome depends not only on the task itself, but on what is already in place: systems, robotics, and the predictability of the assortment.
In companies with a wms that supports location logic, the assignment decision is often already automated; what remains is the physical execution and the oversight of it. That is a different kind of shift than, for example, monitoring inventory levels, where mainly the signaling and deciding is taken over without anyone having to physically move anything. It is also different from generating and assigning pick lists, where the planning can be fully automated but the execution again lies with a human. Warehouse work consists of a series of such tasks, each with its own ratio between thinking work and physical work, as can also be seen with picking and packing goods or with placing orders with suppliers. An overview of how these tasks relate to each other within the logistics process can be found at which work in logistics AI can take over today.
This is not a statement about how much staff a warehouse needs or should have. It is about the nature of one task: which part of it can be calculated according to fixed rules, and which part continues to require physical action. What a company does with that outcome — scheduling differently, redesigning layout, investing in robotics — is a choice that lies with the organization itself and, where it affects staff, is bound by its own applicable legal requirements. We do not provide staffing advice or justification for dismissal decisions; we describe what the work entails.
The classification into eight axes — from structuredness to cost of error — that underlies this kind of judgment is explained step by step on the page about how a task is assessed. That explanation makes clear why physical work, despite high volume and clear structure, still produces a different outcome than purely administrative work with the same characteristics.
If you want to know how this applies to your own warehouse, the first step is not this one task, but the broader work pattern: how much of the work in your warehouse consists of deciding, and how much of physically moving things. The free quickscan from FTE TO AI consists of twelve questions, can be done without an account, and gives an indication of what part of the hours in your profile can be taken over by AI today. The full work scan, which maps the work of an entire company task by task including fte calculation, is still under construction and not currently available.
Vraag maar. Ik ken de kennisbank van deze site; wat ik niet weet, zeg ik erbij.
Answers come from this site’s knowledge base. Not tailored advice, and not a scan of your company.