الفهرس 
Introduction
Overview on Product RolloverOnly 14 pages are availabe for public view
 |  | from | 192 |  |  |
| | | from | 192 | | |
Abstract
With the increased pace of innovation where consumer demand patterns are shifting faster than before, firms view new product development as a competitive advantage that leads to market leadership. Firms introduce new products to eliminate and replace the older products in the market, which is known as a product rollover. As the life cycle of the product get shorter, the rollover process is becoming a regular habit in firms.
Planning decisions for a product rollover are made more difficult by a number of variables, including demand uncertainty during the transition phase, production lot sizes, lengthy lead times, and product cannibalization when the new product enters the market. While it is true that a good product rollover requires proper management at all stages of the supply chain, which necessitates advance planning, a poorly planned rollover results in high inventory levels at the end of production, which results in significant waste costs.
In this research, we study a hypothetical case of a firm supply chain consisting of two echelons of distributers and manufacturers and undergoing a shift in demand from the old to the new products. The firm compares between the two rollover strategies (solo and dual) performance on the supply chain throughout the studied planning horizon.
This research aims to find the optimal production plans for the supply chain echelons, as well as the optimal product rollover timings to maximize the firm’s profit over the studied planning horizon. A MILP model is developed to mathematically model a multi-product multi-period multi-echelon supply chain that consists of distributers and manufacturers. The model objective is to maximize the total profit during the studied planning horizon while considering several parameters such as demand from each product, available manufacturing capacity, inventory holing cost, selling price of the products, opportunity cost of lost sales, salvage value for the remaining inventory at the end of the planning horizon, as well as the cost of unutilized capacity.
Moreover, the model provides flexibility to the decision maker to choose a pre-defined production rollover strategy so that the resulting optimal solution results could be compared between the two rollover strategies.
In this thesis, a numerical example is designed to verify the feasibility and applicability of the mathematical model. A small-sized numerical example consisting of only one distributer and one manufacturer is first presented and the model is tested in three different cases of available capacity: insufficient, sufficient, and excess capacity. The small sized problem is only studied over a six-period planning horizon, with only one old product and one new product replacing its sales and production. The results of the experimentation are demonstrated in this thesis. The numerical example is then extended to a mid-sized problem to verify the practicality of the model with larger problems. The problem is extended to 12-periods and four products, as well as two distributers.
Also, this thesis demonstrates the results of the carried-out sensitivity analyses to test the optimal solution sensitivity to changes in parameters of the opportunity cost, the price ratio between the new and old products, the inventory holding cost, the initial inventory of the old product at the beginning of the rollover, as well as the cost of unutilized capacity. The analysis is carried out using the small-sized numerical example.
Comparing between the optimal solutions of solo and dual rollover strategies for the numerical example, this study finds that the dual rollover gives better results of total profit, average fill rate, as well as lower inventory levels at all cases of available capacity. Also, the results of the sensitivity analysis show that the solo rollover production plan is more sensitive to changes in parameters than that resulting from a dual product rollover strategy.
The MILP model presented in this thesis attempts to close the research gap in this field by being the first multi-product, multi-period, and multi-echelon model to provide the optimal production and sales plan according and can compare between the optimal solutions under either solo or dual production rollover strategy.
The research can be extended in the future to optimize the inventory and purchasing decisions of the raw materials that the finished products are composed of. The model can also be extended to include the echelon of the suppliers in the supply chain, and to study the impact of commonality of raw materials on the performance of the product rollover strategy.