Management of Supply Chain for Perishable Products: Packing Optimization and Production Scheduling

Tran Thi Ngoc Mai

Abstract

Globalization, intensive competition and an increase in demand for perishable products of high quality and safety have put manufacturing companies in such industries in tremendous pressures. It is imperative that companies recognize the need to change and adopt a more effective supply chain model to stay competitive in the market and also cost-efficient. It is, therefore, necessary to explore and investigate the possibilities towards an optimal supply chain for perishable products.

The focus of this thesis is on perishable products with short shelf-life, which have low values and are usually packed in large volume Two major issues with regards to the supply chain of this type of perishable products are explored: production scheduling and distribution planning, and packing optimization. Different from other works on packing optimization which study risks in packing configurations, this paper focuses on space optimization since cost of space is relatively higher than that of spoilage and the consequence of product perishing is low when products are bulky and of low values. Both issues presented are solved using various Optimization models.

For production scheduling and distribution planning, a Pure Integer Linear Programming (PILP) model which integrates product quality deterioration is proposed. The purpose of this model is to aid the company on daily decision making on production and shipment amount to meet customers’ demands and to maximize the profit in a short time frame. The results from the sensitivity analysis correspond with the intuitive approach, validating that the model’s performance is reflective of various conditions and is consistent with expectation. An algorithm is developed, allowing re-planning upon new order arrivals. Also, the algorithm could also be used to decompose the long planning horizon into overlapping micro-periods. Comparison of the results across the different scenarios demonstrates that the proposed approach outperforms the scenario with non-overlapping periods.

For Packing Optimization, the thesis explores an approach to pack three-dimensional non-identical cartons into bins of same dimensions to optimize space utilization of the latter. The Three-Stage Approach, in which cartons are being packed into strips, strips are then packed into layers, and layers are packed into bins, is proposed. Different Optimization models are formulated at the different stages of the approach and solved using LINGO software to achieve the optimal solutions. Application of the Three-Stage Approach on the data of real settings shows promising results.