Assessing the assembly sequence early on: How a new model saves planners time

Author: Timo Miebach | Last update: 26.05.2026

Not every assembly step can be performed at any given time. Certain operations must necessarily take place before others—either because access is otherwise lost, because a screw cannot be tightened until a housing component is secured, or because a fixture requires a specific initial setup. These dependencies are called precedence relationships, and they determine the extent of flexibility available when planning the sequence.

In a flow assembly line, where products pass through fixed stations one after another, this flexibility is particularly valuable. The cycle time of the entire system is determined by the slowest station. To ensure that all stations are utilized evenly—a process known as assembly line balancing—operations must be able to be flexibly distributed across stations. The more precedence relationships constrain the sequence, the more difficult this balancing becomes.

This becomes even more complicated when multiple product variants run on the same line. Different variants entail different dependencies, and the point at which variants begin to differ from one another (the so-called variant origin point) should occur as late as possible in the assembly process so that production can remain “anonymous” for as long as possible. However, restrictions imposed by equipment such as fixtures, workpiece carriers, or special screwdrivers can unintentionally move this point forward, with direct consequences for efficiency and predictability.

Research approaches for evaluating sequence flexibility have existed since the 1970s. They typically measure the number of precedence relationships relative to all possible entries in a precedence matrix. The more relationships there are, the less flexibility there is. The problem is that all constraints are treated equally. It does not matter whether a dependency shifts the critical variant creation point forward or not. And these models do not even capture which specific resource causes the constraint.

This means that while a production planner can see that flexibility is low, they cannot see exactly why, and certainly not which adjustment would have the greatest effect. Until now, this has been of little use for strategic decisions in early planning phases, where precise time data is still lacking.

Our solution is based on the precedence matrix, a tabular tool that illustrates dependencies between assembly operations. In its classic form, each cell contains one of three values: 1 (this operation must come first), -1 (the other must come first), or 0 (no dependency). The new model adds additional information to each cell: Which product is affected? Which resource causes the restriction, or is it inherent in the product itself? And how severe is the restriction?

We distinguish three levels for this weighting. The most severe restriction is one that forces a variant-specific operation to precede a general one, as this shifts the variant creation point forward and most severely impairs planning efficiency. Dependencies between two general operations are of medium importance. Although they affect all products, they leave the variant point untouched. The least harmful are restrictions that influence the variant creation point in the right direction or not at all.

These weighted priority relationships result in the Assembly Sequence Flexibility (ASF)—a metric between 0 and 1. A value close to 1 means: plenty of leeway; the sequence can be easily adjusted. A value close to 0 means: little leeway; the sequence is largely fixed. The model also calculates how much flexibility is lost due to the product alone (ASFp) and how much additional flexibility you

Zum Test wählten wir einen Modellhubschrauber. Ein Produkt mit 80 möglichen Varianten, zusammengestellt aus sieben binären Merkmalen. Die Standardvariante umfasst 17 Montageoperationen, darunter viele Schraubverbindungen und einige Rastverbindungen. Produziert wird in einer Fließmontage mit sechs Stationen. Drei Betriebsmittel (bezeichnet als E1, E2 und E3) wurden auf ihren Einfluss analysiert.

Das Ergebnis: Die Gesamtflexibilität ASFtotal liegt bei 0,71. Das Produkt selbst würde eine Flexibilität von 0,90 erlauben. Die verbleibenden 0,19 gehen auf das Konto der drei Betriebsmittel. Der größte Einzelverursacher ist E1 mit einem Verlust von 0,08, gefolgt von E2 mit 0,06 und E3 mit 0,05.

Was bedeutet das praktisch? Ein Montageplaner, der E1 durch eine flexiblere Vorrichtung ersetzt oder konstruktiv anpasst, würde den größten Flexibilitätsgewinn erzielen und das lässt sich quantifizieren, bevor auch nur eine Zeitaufnahme stattgefunden hat. Die Studie schlägt vor, diesen Gewinn gegen den Aufwand der Maßnahme abzuwägen: Eine Kosten-Nutzen-Analyse auf Basis der ASF-Verluste kann helfen, Investitionen gezielt zu priorisieren.

The model solves a real-world problem in assembly planning. It identifies sequencing flexibility at an early stage and enables informed decisions about equipment before time-consuming time studies are available. This provides real added value, especially for products with many variants where planning resources are limited.

At the same time, there are limitations that should be kept in mind when interpreting the results. So far, only a single product variant has been tested in a training factory—a controlled academic environment that is not directly comparable to a real mass production system. We have identified as an open research question whether the model remains equally manageable for more complex products involving dozens or hundreds of operations.

Furthermore: The model operates without assembly times. This is its advantage in early phases, but also its limitation. It does not provide insights into how well the line can actually be balanced or what the station utilization will ultimately be. It evaluates the structural flexibility, not the concrete outcome. Furthermore, the weighting factors (1, 2, 3) have not yet been validated by experts. A more systematic methodology, such as pairwise comparisons in workshops, would be desirable for practical application. And the question of how the model can be extended to mixed-model assemblies with multiple concurrent products remains open for now.

✓   solid methodological expertise and modern planning tools,

✓   innovative and practical solutions,

✓   professional communication and structured project processes,

✓   many years of experience in research, teaching, and industrial projects.

Our experts are always happy to assist you with project inquiries, collaborations, or further information.