SUMMARY
The Responsive Supply Chain is defined as the appropriate for innovative products and stable processes. This Supply Chain (SC) has to face a highly uncertain demand forecast for its multiple products (with seasonal and volatile demand). It is important to reduce costs, stock-outs and closings at the end of each selling season.   
The objective of this thesis is to improve the Operations Tactical Planning of Responsive Supply Chains with alternative structure of processes (i.e. the possibility to manufacture the products in several ways applying the concept of postponement), for short life cycle products (with drastic loss of value on the market at the end of the sales season), with alternative suppliers, with stable processes and with common and uncommon components. It seeks to maximize the benefits as difference between the income for sales and the total costs of production, storage and transportation.
For the modeling methodology, two moments has been looked in the decision-making. The first moment of tactical operations planning takes place several months before the start of the sales season and is planned to highly uncertain forecasts of demand. The second moment of tactical operations planning occurs at the beginning of the sales season, when the first sales allow obtaining highly accurate forecasts of demand. Therefore two mathematical models of operations tactical planning have been developed (a stochastic and a deterministic model) with a new approach of strokes, in which the processes are planned instead of the products. This new approach simplifies mathematical modeling of the multiple alternatives of postponement and multiple processes of manufacture and transportation.
In order to implement the consecutive execution of the models, a tool in Java has been developed that: generates a XML file which contains all the data entry to the first model, simulates demand for the second model (using the Bass diffusion method for new products and the simulation of Monte Carlo), calculates the results of the first model in the beginning of the sales season (based on the simulated demand) and enters them in the second model, executes the second model as many times as the decider considers suitable (for equal number of simulated demands), calculates results (average of different executions) and generates antocher XML file with the results. The mathematical models have been programmed in AMPL language and they are solved with the commercial optimizer Gurobi Optimizer 3.0.3.
Based on the information provided by a Valencian company, that manufactures locally and in addition subcontracts (to a suppliers in China) the manufacture of components and articles of plastic and metal, different alternatives of costs have been experimented and also different: demand uncertainty, moments for the execution of the second model, lead times, etc., for different problematic of the SC.
It is possible to affirm that the set of developed models improves the responsive SC operations tactical planning, allows the selection of different alternative suppliers and, allows to anticipate and to analyze different demand scenes and alternatives of processes, capacities, values of clearance sale, etc.