| This dissertation addresses a number of fundamental risk factors regarding the supply mechanisms in general supply chains. Typically, two principal types of uncertainty pertain to any production or purchase order: (1) (LEADTIME UNCERTAINTY) How long will the order leadtime be, i.e. how long will it take before the order is received? (2) (YIELD UNCERTAINTY) In view of quality risks and potential supply disruptions due to natural causes, labor strikes as well as planned acts of sabotage, what fraction of the order size will become available as useable units?; As to the first supply risk factor, almost all supply chain planning models make one of two assumptions regarding the process which generates the leadtimes of consecutive orders: (a) leadtimes are of a constant, deterministic duration, (b) the leadtime process, while allowing for randomness, precludes order crossing, i.e. with probability one, orders arrive in the same sequence in which they are placed. The first approach (a) ignores any source of randomness, while approach (b) is made for the sake of analytical tractability, but is rarely satisfied per se. Chapter 2 of this dissertation studies inventory systems under a broad class of stochastic leadtime processes allowing for order crossing. To our knowledge, this class includes all previously studied leadtime models.; The remainder of the dissertation addresses the second major supply risk factor, i.e. the complications arising from uncertain yields and their implications for the determination of the "right number " of suppliers, the selection of the best set of suppliers; the optimal diversification of orders among this selected set, as well as models characterizing the competition between suppliers. This part of the dissertation consists of three chapters. |
