Access Type

Open Access Dissertation

Date of Award

January 2013

Degree Type


Degree Name



Industrial and Manufacturing Engineering

First Advisor

Ratna B. Chinnam


This dissertation focuses on demand modeling and capacity planning for innovative short life-cycle products. We first developed a new model in the class of stochastic Bass formulations that addresses the shortcomings of models from the extant literature. The proposed model considers the common fact that the market potential of a product is not fixed and might change during a life-cycle due to exogenous (e.g., economic- or competitors-related) or endogenous (e.g., quality-related) factors. Allowing this parameter (market potential in the Bass model) to follow a geometric random walk, we have showed that the future demand of a product in each period follows a lognormal distribution with specific mean and variance.

We also developed a novel stochastic capacity expansion model that can be used by a make-to-order manufacturer, who faces stochastic stationary/non-stationary demand, in order to optimally determine policies for specifying the sizes of capacity procurement. In addition to the cost of expansion decisions, the proposed risk-neutral expansion model considers procurement lead-times, irreversibility of investments, and the costs associated with lost sales and unutilized capacity. We provide necessary and sufficient conditions for the derived optimal policy. We then present an exact solution method, which is more efficient than classical recursive methods.

Additionally, three extensions of the proposed expansion model that can address more complicated settings are presented. The first extension increases the capability of the model in order to tackle capacity planning for a multi-sourcing scenario. Multi-sourcing is a case in which the manufacturer can procure capacity from two supply modes whose marginal expansion costs and lead-times are complementary. The second extension addresses a scenario in which an installed capacity can be used for producing future generations of a product. The last extension accounts for salvage value of the installed capacity in the model and provides the necessary and sufficient conditions for the optimal policy.

Finally, using the proposed stochastic Bass model, we present the results and managerial insights gathered from numerical experiments that have been conducted for the stochastic capacity expansion models.