| dc.description.abstract | The increasing interest in the market for innovation has become a key characteristic of food supply chains (FSC). New product launches are part of the strategic advantage of a food company; however, new products are characterized by high failure rates and demand uncertainty. To deal with these situations, researchers have been proposing more accurate methods to forecast demand for new products and strategies for selling these, but little attention has been given to the possibility of mitigating the risk related to new product launches from the manufacturing process.
Before a product is developed and launched in the market, several decisions take place along the supply chain. In the food manufacturing stage, the production planning and control (PPC) approach of the company can facilitate or worsen the risk related to new product launches. A key decision area within PPC is lot-sizing, which identifies when and how much of a product to produce such that the setup, holding, and production costs are minimized. Thus, a proper determination of a product’s lot size can reduce its costs, and inventory level, and increase the level of responsiveness to the customer’s demand. However, it is imperative to select the appropriate lot-sizing approach.
Dynamic lot-sizing (DLS) is a more adequate approach for products exhibiting a time-varying demand. For new products, the initial demand is unknown, thus the initial volume of demand estimated needs to be produced efficiently and easily adjusted to the demand behavior. DLS can be applied to new product launches in order to reduce over costs and excess inventory that can end up in spoilage while aiming to fulfill the customer’s demand on time. Therefore, this thesis attempts to improve the lot-sizing decisions for new product launches in a company from the food sector applying a DLS model, aiming to show the potential gains and limitations of this solution. In order to do this contribution, the following research questions (RQ) were answered:
• RQ1: What are the characteristics of dynamic lot-sizing models for food products?
• RQ2: How dynamic lot-sizing can be applied to new products during their launch period?
• RQ3: What are the potential gains and limitations of applying dynamic lot-sizing for new product launches?
The project was conducted through a combination of a literature study and a case study. The literature study helped to identify the characteristics of DLS models and the characteristics of FSC. With these characteristics, it was possible to map the main features of DLS for food products and thus answer RQ1. The first RQ is the foundation to answer RQ2 and RQ3, which focus on a practical problem from the case company.
The case company had several introductions of new products in the period from 2016 to 2020, showing a high level of innovation. The life cycle of the products introduced during this period varied significantly. A family of new products was selected and further analyzed in terms of demand behavior, lot-sizing approach, and production process. The findings from the analysis and RQ1 were used to develop the mathematical formulation of a DLS model for new product launches for the selected family of products. Furthermore, the mathematical model was implemented in a non-dominated Sorting Genetic Algorithm (NSGA II) to propose a solution. With the proposed model, the RQ2 was answered.
To answer the third question, the findings from RQ1 and RQ2 were used. Through the literature study, several benefits of DLS models were identified amidst lower setup and inventory holding cost, better inventory management, and efficient response to the demand, as well as limitations such as high complexity to apply or insufficient inventory to buffer drastic changes in the demand.
Overall, the multi-objective DLS model applied in this study has proven to contribute to both efficiency and responsiveness through four important aspects namely service level, total cost, inventory level, number of periods of production in advance of the launch date, and flexibility. The objective related to cost reduction ensures that the number of setups is optimal such its related cost is minimized, this also contributes to reducing downtime and improving the utilization of capacity, benefits that enhance the efficiency of the company. Furthermore, the minimization of inventory holding costs contributes to a reduction in the number of units held in inventory, which in turn allows for a streamlined production process, where problems can easily be identified and corrected. Regarding responsiveness, the proposed DLS model has service level as a core objective. Guaranteeing a high service level led to meeting customer demand on time and in the quantities ordered. | |