PhD defence: Mario Guajardo

On Friday 30 August 2013 Mario Guajardo will hold a trial lecture on a prescribed topic, and defend his thesis for the PhD degree at NHH.

The thesis consists of five articles on joint decision making and cost/revenue sharing in supply chains for natural resources, especially oil. A challenge in this type of supply chain is formulating decision models that can cope with the constraints and goals from all the agents.

The first two articles address this challenge through the formulation of several optimization models. It is shown that an integrated decision approach outperforms a traditional decoupled decision approach in terms of a maximum contribution objective function. However, in the integrated approach the sellers may be left with a worse outcome, because the premiums for sales that they receive become lower. The contribution is therefore re-allocated to leave all agents better off.

The third article deals with a case of inventory management of spare parts in a main energy company, producer of oil and gas. Several demand models are tested in a large number of items and decision rules are proposed in order to select a demand model for each item. The article also investigates the impact of correcting inventory inaccuracy records and pooling inventory for different platforms.

The fourth article uses cooperative game theory for answering how the cost must be shared in an inventory pool subject to multiple demand classes. A new cost allocation method is proposed, based on the notion of service level referential costs. The allocation is core guaranteed, in the sense that if the core is not empty, the allocation belongs to the core. In situations where it is inconvenient to form the grand coalition, the relevant question arises of which coalitions should be formed.

The fifth article of this thesis formulates mixed integer linear programming models to answer this question, considering core stability and strong equilibrium. Concerned about the difficulties of implementing large coalitions in practice, the models also incorporate constraints on the maximum cardinality per coalition. The models are tested in two applications, one in collaborative forest transportation and one in pooling of inventory of spare parts for oil operations. The results illustrate the relevance that the stability notions may have when structuring the coalitions and that the marginal contribution of allowing one more player per coalition may significantly decrease when the upper bound approaches the cardinality of the grand coalition.