Based on bills of materials and order quantities, it is determined which materials and quantities are needed, and whether replenishment is required. This is work for the planner or buyer, carried out in the ERP system or inventory system. It is calculation work with a fixed source and a fixed demand: bill of materials times order quantity, compared to what is in stock.
Three axes are decisive: structuredness, volume and error costs.
The structuredness is high. A bill of materials is a fixed list of parts and quantities per end product. An order quantity is a number. The calculation that follows — quantity needed times number, minus what is in stock, is replenishment — is a calculation rule with no room for interpretation. There is no moment where someone has to decide what a customer "actually" means, as with explaining a vague customer request. The rules are fixed in the bill of materials management.
The volume is high. This calculation must be redone for every order, every day. With hundreds of orders per week, this quickly adds up to a substantial part of a planner's or buyer's hours. Work that recurs often and varies little is exactly the work where an AI agent saves time: not because it is clever, but because it is a lot.
The error costs sit in the middle, and that is the reason for oversight. An incorrect material calculation does not immediately lead to a dangerous situation or a legal problem, but it can lead to an order getting stuck on the floor because there is too little material, or to unnecessarily high inventory because too much was ordered. This costs money and time to fix, but it is usually reversible. This is why an AI agent that performs the calculation and proposes a replenishment, with a human approving the outcome before an actual reorder is placed or a purchase order is placed with a supplier, fits here.
The remaining axes show why this is not a task that can be left to AI without oversight.
Judgment scope scores 3: there are situations where a planner makes an exception, for example because a supplier delivers unreliably or because a customer has a rush order that breaks the normal sequence. Compliance scores 4, meaning that few legal obligations attach to this task itself, but that the consequences of an incorrect calculation — too little material for a critical delivery — can have knock-on effects on agreements with customers.
Customer contact, physical work and creativity are not relevant here: this is an internal calculation, not a conversation and not a physical action. This does not weigh against automation, it is simply not a relevant threshold for this task.
For this task, an AI agent can be deployed: a system that independently combines the bill of materials with the order quantity, consults the inventory level and makes a proposal for replenishment. The human reviews that proposal, approves or rejects it, and intervenes in exceptions.
This only works under two conditions. First, the bills of materials must be up to date: if a bill of materials is missing a part or lists an incorrect quantity, the AI will calculate flawlessly on a faulty basis. Second, the inventory levels must be reliable: if the inventory system does not match what is actually on the shelf, the calculation of the replenishment will not be correct either. Companies where incoming goods are structurally received and booked in usually have that condition in order; companies where this is still done manually and irregularly run into a wall of inaccurate data before the calculation itself ever becomes a problem.
At a company with a limited number of fixed products and clear bills of materials, this task can be almost fully automated, with a light check afterwards. At a company with a lot of custom work, changing suppliers and bills of materials that differ slightly per order, more work remains where a planner personally assesses whether the calculation is correct and whether an exception is needed. The quality of the underlying systems also makes a difference: a company that consistently maintains its ERP and inventory system gains more freed-up hours than a company where that data lags behind.
This shift is also connected to other parts of business operations. Those who automate material calculation often see that follow-up steps such as ordering and receiving are also due for revision, and this applies more broadly to production environments — see what AI can take over in the installation sector for how this type of task relates to planning and material logistics. The picture is also shifting in administration itself: which work in administration AI is already taking over today shows that similar calculation rules often apply there as well.
This is not personnel advice and not a reason to draw conclusions about positions. What an employer does with the freed-up hours is up to the employer itself; separate legal requirements apply to that, on which this page makes no statement. What is stated here is a description of the task: which part is structured enough for an AI agent, and which part continues to require oversight.
Whether material calculation is largely transferable in your situation depends on how up to date your bills of materials are and how reliable your inventory registration is. The free quickscan of twelve questions, without an account, gives a first indication of which part of the hours in this profile can be taken over by AI today. The full work scan, which breaks down your company's work task by task, is still under construction.
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.