Datasets for lot sizing and scheduling problems in the fruit-based beverage production process.

Area of Science:

• Operations Research
• Industrial Engineering
• Supply Chain Management

Background:

• Production planning and scheduling in fruit-based beverages are complex due to two-stage processes (preparation tanks, production lines) and temporal cleaning requirements.
• Existing research lacks standardized datasets, hindering the validation and benchmarking of proposed optimization methods.
• Previous studies often used small, unrepresentative datasets, limiting the assessment of method performance across diverse production scenarios.

Purpose of the Study:

• To introduce a comprehensive set of datasets for the lot sizing and scheduling problem in fruit-based beverage manufacturing.
• To provide researchers with a benchmark for evaluating and comparing production planning and scheduling algorithms.
• To facilitate the assessment of optimization methods under realistic conditions, including seasonal demand and restricted capacity.

Main Methods:

• Datasets were constructed using real-world data from five beverage companies.
• Four distinct datasets were generated, focusing on scenarios with restricted capacity and balanced costs.
• The datasets are designed to represent a wide range of production situations, accounting for seasonal demand patterns.

Main Results:

• The presented datasets offer a valuable resource for the scientific community.
• They enable rigorous testing of lot sizing and scheduling heuristics and exact methods.
• The datasets facilitate the evaluation of method effectiveness in solving real-world production challenges.

Conclusions:

• The lack of standard datasets has been a significant limitation in advancing lot sizing and scheduling research for beverage production.
• The introduced datasets provide a crucial foundation for future research and development in this area.
• These datasets will promote more robust and reliable evaluation of optimization techniques for the industry.