The lot-sizing problem concerns a manufacturer that needs to solve a production planning problem. The producer must decide at which points in time to set up a production process, and when he/she does, how much to produce. There is a trade-off between inventory costs and costs associated with setting up the production process at some point in time. Traditionally, the lot-sizing model focuses solely on cost minimisation. However, production decisions also affect the environment in many ways. In this dissertation, the classic lot-sizing model is extended into several different directions, in order to take various environmental considerations into account. First, items that are returned from customers are included in the lot-sizing problem, within the context of reverse logistics. These items can be remanufactured to fulfil customer demand. In another extension, a minimum is imposed on the size of a production batch, in order to reduce the pollution associated with producing many small batches. Furthermore, a lot size model is considered in which there is a maximum on the amount of pollutants, such as carbon dioxide. This model can also be seen as a bi-objective lot-sizing problem. The mathematical models that arise from these extensions are fundamentally harder to solve than the classic lot-sizing problem. Several approaches to solving these problems are developed, based on mathematical optimisation techniques such as mixed integer programming, dynamic programming and fully polynomial time approximation schemes.
|Award date||3 Oct 2013|
|Place of Publication||Rotterdam|
|Publication status||Published - 3 Oct 2013|