Analysis of techniques for sales forecasting and inventory optimisation for the company
This project aims to enhance supply chain management for companies by integrating a forecasting functionality into the Odoo Enterprise Resource Planning (ERP) software and testing it using a Business Game (BG) environment. Specifically, the project leverages the Neural Prophet forecasting package to...
| Autor: | |
|---|---|
| Tipo de recurso: | tesis de maestría |
| Fecha de publicación: | 2024 |
| País: | España |
| Institución: | Universitat Politècnica de Catalunya (UPC) |
| Repositorio: | UPCommons. Portal del coneixement obert de la UPC |
| Idioma: | inglés |
| OAI Identifier: | oai:upcommons.upc.edu:2117/413998 |
| Acceso en línea: | https://hdl.handle.net/2117/413998 |
| Access Level: | acceso abierto |
| Palabra clave: | Industrial management Sales forecasting Empreses--Direcció i administració Vendes--Previsió Àrees temàtiques de la UPC::Economia i organització d'empreses |
| Sumario: | This project aims to enhance supply chain management for companies by integrating a forecasting functionality into the Odoo Enterprise Resource Planning (ERP) software and testing it using a Business Game (BG) environment. Specifically, the project leverages the Neural Prophet forecasting package to improve the calculation of Average Daily Usage (ADU), a critical parameter in Demand Driven Material Requirements Planning (DDMRP). This tool is designed to enable enterprise and inventory managers, regardless of their forecasting expertise, to predict demand more accurately and efficiently size stock buffers. Key objectives include a comprehensive literature review of DDMRP and ADU, preparation of an ERP simulation environment, adaptation of the demand function, and modification of various demand-influencing parameters to serve as regressors in the forecasting module. Additionally, the project involves developing functionalities for utilising ERP variables and external xlsx data as regressors, generating forecasts, and evaluating their accuracy using metrics such as Mean Absolute Error (MAE), Mean Squared Error (MSE), and Root Mean Squared Error (RMSE). The result is a coherent system with a newly created demand dashboard and modified marketing investment planning view, allowing users to simulate future demand based on configurable parameters. The project demonstrates the forecasting tool's effectiveness in predicting demand by training models with various regressor combinations and the BG's usefulness for introducing individuals into the ERP functionalities. Evaluations indicate the model's robustness, with consistent MSE, RMSE, and MAE values converging around optimal points, suggesting minimal overfitting and reliable performance. Overall, the developed tool provides the base of a user-friendly solution for non-data scientists to enhance supply chain management through accurate demand forecasting within the Odoo ERP framework. |
|---|