Controlled atmosphere delivery decision support system

Abstract

This thesis describes a decision support system that aids in decisions involving the shipment of perishable items. It is currently possible to arrest the aging process of produce by controlling the atmospheric conditions under which they are packaged. This thesis utilizes three mixed integer optimization programs written in General Algebraic Modeling System (GAMS) to determine possible sites for placement of the controlled atmosphere stations, to minimize total costs involved, and to optimally load produce demands in compatible atmospheric vans and, hence, minimize the number of vans required to meet demands. The user of the system has control over the input parameters which are maintained in spreadsheet form and read into the appropriate optimization program. These parameters include: commodity and transportation costs, demand locations and amounts, supply countries, supply amounts, van constraints, van load possibilities (binary), and shipping possibilities (binary). Based on which model is run, possible outputs include: amount to ship from source to demand, backorders needed, shipment routes needed, number of each van type needed, total vans needed, direct costs, total costs, projected number of vans needed in advance, and candidates for possible gassing sites (locations where produce would be loaded into controlled atmospheric vans) The following possible gassing sites were identified: Canada, Guatemala, Mexico, Dominican Republic, and the Bahamas. Further investigation is warranted for these countries.