With the increasing importance of means of recovering items, practices such as reuse, recycling and remanufacturing have been gaining increasing importance in production systems. In this thesis, we aim to explore methods to obtain costefficient production plans for manufacturing systems that employ remanufacturing. A crucial challenge here is regarding parameter uncertainty. Commonly, decision makers have to tackle uncertainties on aspects such as operational costs, demands and in the case of

more »

... oduction recoveries, the number of items returned by customers. Our work aims to address these challenges through the framework of robust optimization, where we impose uncertainty on demands and/or returns. In Chapter 1, we provide an introduction outlining the methods and formulations related to our study. In Chapter and 2, we review the literature on lot sizing problems, and the implications of remanufacturing. Following this, we introduce a robust lot sizing problem with remanufacturing in Chapter 3, where we consider uncertainties in both customer demand and return and implement a min-max decomposition approach for this problem. In Chapter 4, we propose a novel approach which employs extended reformulations for the master problem, while addressing computational challenges. In Chapter 5, we consider a different setting for the lot sizing problem with remanufacturing, where a two-level multi-component variation is considered. We consider the case where fixed costs are imposed on components and only variable costs are considered for the end-item. This is then followed by a robust formulation for and a min-max decomposition approach. We show that certain optimality properties have to be enforced to derive costs correctly. In Chapter 6, we consider the two-level multi-component problem with fixed costs on the end-item level. We show certain optimality conditions for this problem, and show how these properties can be exploited in the dynamic programming setting introduced in Teunter et al. [2006]. Throughout Chapters 4 and 5, we provide extensive computational experiments. vi Finally, in Chapter 7, we provide future research directions and conclusions, where potential advantages and limitations are discussed.