A Holistic Framework for Production Scheduling in Industry 4.0

• Published in: 2023 19th International Conference on Distributed Computing in Smart Systems and the Internet of Things (DCOSS-IoT) pp. 269 - 276
• Main Authors: Dimoudis, Dimitris, Vafeiadis, Thanasis, Nizamis, Alexandros, Musiari, Emma, Ziliotti, Luca, Ioannidis, Dimosthenis, Tzovaras, Dimitrios
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01.06.2023
• Subjects: Companies; Costs; deep learning; forecasting; Industries; Industry 4.0; Job shop scheduling; Machine learning algorithms; Pipelines; Production; safety stock; scheduling; Wagner-Whitin
• ISSN: 2325-2944
• DOI: 10.1109/DCOSS-IoT58021.2023.00053

In today's highly competitive retail industry, companies must constantly find ways to optimize their processes and improve efficiency in order to stay ahead. This often involves incorporating advanced technology and data-driven approaches to forecasting, inventory management, and production planning to reach the Industry 4.0 standards. In this work, a comprehensive approach is presented that integrates multiple tools and methods to optimize retail processes, including forecasting, safety stock calculation, and batch item production optimization. The proposed pipeline is evaluated using a real-world sales dataset from a dairy company, where a 30-day forecast is generated with low MAPE and sMAPE errors. The performance of the solution is assessed using two safety stock formulas, integrated into the forecast, and optimized using the Wagner-Whitin algorithm. The results showed a significant cost reduction, representing 38.7% of the total cost, compared to a benchmark approach. In conclusion, the proposed methodology is an effective and practical solution for retail companies seeking to improve their operations and gain a competitive advantage.