Requirements for configuring components are derived from the following objectives for configuring a modular system:
- Universal applicability: A modular system should be designed in such a way as to ensure that as many variants of customer-specific machines as possible can be configured using the components.
- Simple configuration: Minimal effort should be expended in configuring and parameterizing the components.
New requirements that cannot be met with existing components lead to further development of the modular system. This process involves two different principles, both of which are illustrated in the following diagram.
Two variants, A and B, are depicted on the left side of the Variant generation diagram. They differ only in terms of the alternative subcomponents D and E. The modular system in the example contains two component classes, A and B, for the two variants. In order to reduce variance, the modular system can be further developed by applying the following principles:
- Generalization: The difference between variants A and B lies in the installation angle. Thus, components A and B can be generalized into a powerful component F by being assigned an angle parameter to describe the installation position.
- Decomposition: The components are assembled out of smaller components. In the example presented above, three components are admittedly created from what were originally two components; redundancies are avoided, however. Furthermore, the process yields fine-grained components (for example, C) that are reusable in further variants.
Depending on the focus of the principles being applied, the configuration effort is divided between configuration and parametrization tasks. Should the parametrization of powerful components become too complicated, decomposition of these components is often necessary. If the number of components to be managed in the modular system threatens confusion, it is helpful to generalize.