Why Large ERP Systems Like SAP Aren't Enough for Make-to-Order Production Planning

ERP systems such as SAP reliably cover standardised business processes: accounting, warehousing, procurement and core production flows. Daily capacity planning in make-to-order manufacturing, that is, allocating people, machines and tooling according to current workload and disruptions, is typically still handled manually, often in Excel or over the phone. The gap between what ERP offers and what operations actually need can now be effectively closed by a combination of custom software and artificial intelligence tools.

What is make-to-order production, and why is it harder to plan than series production?

Make-to-order production means a company isn't manufacturing one product in large volumes, but smaller batches or individual units based on specific customer orders. Each order can have different dimensions, deadlines, and material or machine requirements. While series production is planned in advance over a long period, in make-to-order manufacturing demand arrives continuously, and the company must assess daily whether it has spare capacity, when it can realistically deliver an order, and how a new deadline affects the ones already scheduled. As soon as the human factor comes into play, illness, a machine breakdown, or a delayed material delivery, the entire plan has to be reshuffled, and doing this manually is very hard to keep up with.

Why don't large ERP systems like SAP cover capacity planning in manufacturing?

ERP (Enterprise Resource Planning) is a software system that connects key business processes, from accounting through warehousing to production, on a single platform. Its strength lies in providing a unified view across the whole company or even an entire multinational group, which is why it tends to be configured generically, so it works across branches and countries. The detailed operational specifics of a single production hall, such as the availability of a particular mould or the current staffing of one trade, usually never make it into such a system. In companies with a foreign parent, the ERP is often configured centrally, and the local branch has no room to adapt it to its own needs. As a result, digitalisation initiatives often emerge from the bottom up, at the level of an individual branch that finds its own solution to a specific problem rather than waiting for a change across the whole group.

How does artificial intelligence help with predicting deadlines and planning capacity?

If a company has historical data on how long different types of orders take and how schedules have shifted in the past, this data can be used to build a language model, a type of artificial intelligence capable of processing large volumes of textual and structured data and finding patterns within it. The model takes current demand and the operation's workload status as input, and based on that, proposes a realistic deadline or flags a clash with another order. In the early stages of deployment, the model mostly makes recommendations that a person then approves; only gradually, as its reliability is confirmed, does the level of automation increase. The advantage over manual planning in Excel is that the model can recalculate the impact of a disruption on dozens of related orders within moments, a task that takes a person considerably longer.

How does artificial intelligence check technical documentation against applicable standards?

Newer language models can process images and documents as well as text, which means they can also be used to check technical drawings or production photographs against applicable standards. The model is given the requirements set out in the standard on one side and the actual drawing, photograph or PDF on the other, and compares whether the output complies. This replaces manually looking up the current wording of a standard and comparing it by hand, a process that is time-consuming and error-prone. The accuracy of such a check improves the more clearly a rule is formulated in the standard, so for more complex or ambiguous requirements, human verification is still needed.

How do mobile apps and QR codes speed up order tracking in manufacturing?

A QR code is a two-dimensional barcode that, once scanned with a mobile phone, immediately displays the associated data, for example information about materials, order status, or the progress of a service task. In practice, it's most useful where technicians or operators work in the field or directly on the shop floor and have so far recorded information on paper. A mobile app linked to QR codes lets staff log an action on the spot, with the data immediately reflected in the central system, without anyone needing to transcribe it manually afterwards. Companies typically use this approach to digitalise service and maintenance processes, warehouse material records, or tracking work-in-progress on the production floor.

Excel-based production planning vs. AI-supported planning

• Data currency - Planning in Excel: Manual updates, risk of outdated information. AI-supported planning: Data is linked automatically from source systems.
• Response to disruption - Planning in Excel: Manual recalculation of impact on other orders. AI-supported planning: The model recalculates the impact on related orders within moments.
• Visibility across the company - Planning in Excel: The file is often held by just one person. AI-supported planning: Shared overview available across teams.
• Error resilience - Planning in Excel: Highly dependent on one individual's attention. AI-supported planning: The system flags clashes or discrepancies.
• Scalability - Planning in Excel: Complexity grows with the number of orders. AI-supported planning: Handles a higher volume of data without losing clarity.

Signs your ERP isn't enough for production planning:

• the key production or capacity plan lives outside the main system, typically in Excel or a shared document,
• one specific person is responsible for keeping the plan up to date, and their absence slows planning down,
• information about breakdowns, absences, or delayed materials reaches the plan late or not at all,
• checking compliance with standards or regulations involves manually searching through databases,
• service or production data from the field is transcribed into the system from paper.

Frequently Asked Questions

What is an ERP system? ERP (Enterprise Resource Planning) is a software system that connects key business processes, such as accounting, warehousing, procurement and production, on a single platform with a unified view of data across the company.

Why do companies with an ERP still use Excel for production planning? ERP systems tend to be configured generically so they work across the whole company or group. The detailed operational specifics of a single hall, such as the availability of a particular mould or shift capacity, usually never make it into them, which is why a parallel manual record-keeping system emerges.

How long does it take to deploy an AI tool for capacity planning? It depends on the scope and quality of the available historical data and the complexity of integrating it with existing systems. A realistic approach is to start with a small pilot solution in one specific area and expand it gradually.

Can an AI tool be connected to an existing ERP such as SAP? Yes, a custom solution is usually built as an add-on to the existing ERP, drawing data from its interfaces and adding functionality that the standard system doesn't cover, without needing to replace the ERP.

Will artificial intelligence replace production planners? No, the model acts as a support tool. It proposes deadlines and flags clashes, but the final decision and approval of the plan remain with a person, especially when accounting for operational exceptions.

How does artificial intelligence check technical documentation against standards? A language model capable of processing images or PDFs as well as text compares the content of a drawing or photograph with the requirements set out in the relevant standard and flags any discrepancies, speeding up and improving the accuracy of manual checks.

What is make-to-order production? Make-to-order production is a model in which a company manufactures smaller batches or individual units based on individual customer orders, as opposed to series production of a single product in large volumes.

Is custom software worthwhile for a smaller manufacturing company too? Yes, if it addresses a specific bottleneck that neither off-the-shelf software nor the ERP system covers. A solution can be built as a small pilot tool for a single area, without needing to change the company's entire information system.

Summary

ERP systems such as SAP cover standardised business processes well, but they typically don't cover detailed daily capacity planning in make-to-order manufacturing, which is why companies handle it manually, often in Excel. Artificial intelligence can fill this gap, whether by predicting realistic deadlines based on historical data, checking technical documentation against standards, or digitalising order tracking through mobile apps and QR codes. The key is not to build the solution as a replacement for the ERP, but as a targeted add-on addressing a specific operational problem. This approach is faster to deploy and easier to evaluate for impact than trying to replace the entire information system at once.

If you're dealing with a similar gap between your ERP and your operation's real needs, we'd be happy to discuss your specific situation, with no obligation.