Abstract
Mixed-model assembly lines are of great practical relevance and are widely used in a range of industries, such as the final assembly of the automotive and electronics industries. Prior research mainly selected and discussed isolated problems rather than considering the whole planning process. In this article, mixed-model production planning is decomposed into five steps: initial configuration of the line, master scheduling, reconfiguration planning, sequencing, and re-sequencing. This article reviews and discusses all relevant planning steps and proposes general planning instruments as well as formalised decision models for those steps, which have not been thoroughly investigated in the literature thus far.
Notes
Notes
1. Related planning problems, such as maintenance planning of machines, calculation of safety stocks for just-in-time delivery and staff deployment, which can also be subsumed under production planning in a wider sense, are not in the scope of this article.
2. To keep the model simple, it is assumed that deviation costs c it do not depend on the shift in which a model i is produced and that any additional shift is equal in length and incurs constant costs of K S . The model can, however, easily been extended to cover the more general cases, so that the extensions are not provided.
3. The weight α thereby determines whether the resulting car sequencing problem can be solved without rule violations. Lower bound computations can be employed in order to set α appropriately (see Fliedner and Boysen 2008).
4. Note that re-sequencing can also be employed to decouple multiple production levels. In the automobile industry, re-sequencing buffers are, for instance, employed in between the paint shop and the final assembly, so that the sequence fed to the paint shop can be modified before the final assembly starts. By means of this, the total sequencing problem is decomposed into a paint shop sequencing problem, where batches of the same colour are sought (Lustig and Puget 2001, Spieckermann et al. 2004), and an assembly sequencing problem, which focuses on capacity violations (Inman and Schmeling 2003).