Loading...
Publication
Desenvolvimento de um modelo de gestão de inventários para cadeias de abastecimento em ciclo fechado
| Name: | Description: | Size: | Format: | |
|---|---|---|---|---|
| 6.42 MB | Adobe PDF |
Authors
Advisor(s)
Abstract(s)
A crescente consciencialização dos consumidores sobre práticas ambientalmente responsáveis,
juntamente com regulações impostas pelos governos, tem incentivado as empresas a incorporarem
desenvolvimento sustentável na generalidade dos seus processos e operações.
Consequentemente, sistemas como cadeias de abastecimento em ciclo fechado são de elevada
relevância, uma vez que diminuem os recursos necessários, através da reincorporação de produtos
no sistema após a venda. Contudo, estas cadeias de abastecimento apresentam desafios e
dificuldades acrescidas, nomeadamente a nível de gestão de inventários.
Neste contexto, a presente dissertação teve como objetivo desenvolver um modelo de gestão de
inventários adaptado a cadeias de abastecimento em ciclo fechado, aplicável ao contexto real de
uma empresa portuguesa da indústria de gás de petróleo liquefeito.
Para atender ao desafio, foram, primeiramente, desenvolvidas previsões relativas a cinco
variáveis-chave: Vendas, Retornos, Enchimentos, Recondicionamentos e Abates, tendo sido
utilizados os dados reais da empresa. Na fase preditiva, foram comparados dois modelos aplicáveis
a séries temporais: Prophet e Theta. O modelo Prophet apresentou melhor desempenho na
previsão dos Retornos, Enchimentos e Vendas, com erro percentual médio absoluto de 8,92%,
8,83% e 10,81%, respetivamente. Para a previsão dos Recondicionamentos e Abates o método
Theta revelou maior precisão, com erros de 15,79% e 23,28%, respetivamente.
As previsões obtidas alimentaram o modelo de otimização, cuja função objetivo pretendia
minimizar os custos totais, compostos pelos custos económicos e ambientais, através da
incorporação da métrica Cap-and-Trade. A aplicação do modelo resultou numa redução de 27,17%
nos custos totais, de 24,8% nos custos económicos e de 29,25% nas emissões de CO₂, face ao
cenário inicial da empresa.
Os resultados demonstram a aplicabilidade do modelo formulado, evidenciando o potencial de
aplicação a outros contextos industriais, contribuindo para cadeias logísticas mais eficientes e
ambientalmente responsáveis.
The increasing consumer awareness regarding environmentally responsible practices, alongside government-imposed regulations, has encouraged companies to incorporate sustainable development across their processes and operations. Consequently, systems such as Closed-Loop Supply Chains have become highly relevant, as they reduce the need for resources by reintegrating products into the system after sale. However, these supply chains present additional challenges and complexities, particularly in inventory management. In this context, the present dissertation aimed to develop an inventory management model tailored to closed-loop supply chains, with application to the real-world context of a Portuguese company operating in the liquefied petroleum gas industry. To address this challenge, forecasts were first developed for five key variables: Sales, Returns, Refills, Refurbishments and Disposals, using the company’s historical data. In the forecasting phase, two time series models were applied and compared: Prophet and Theta. The Prophet model outperformed in forecasting Returns, Refills and Sales, achieving mean absolute percentage errors of 8.92%, 8.83%, and 10.81%, respectively. For forecasting Refurbishments and Disposals, the Theta method performed better results, with MAPE values of 15.79% and 23.28%, respectively. The forecasts generated served as input for the optimization model, whose objective function aimed to minimize the total cost, consisting of both economic and environmental costs, through the incorporation of the Cap-and-Trade metric. The implementation of the model led to a 27.17% reduction in total costs, a 24.8% decrease in economic costs and a 29.25% reduction in carbon emissions compared to the company’s initial situation. The results demonstrate the applicability of the formulated model, highlighting its potential for implementation in other industrial contexts and contributing to the development of more efficient and environmentally responsible supply chains.
The increasing consumer awareness regarding environmentally responsible practices, alongside government-imposed regulations, has encouraged companies to incorporate sustainable development across their processes and operations. Consequently, systems such as Closed-Loop Supply Chains have become highly relevant, as they reduce the need for resources by reintegrating products into the system after sale. However, these supply chains present additional challenges and complexities, particularly in inventory management. In this context, the present dissertation aimed to develop an inventory management model tailored to closed-loop supply chains, with application to the real-world context of a Portuguese company operating in the liquefied petroleum gas industry. To address this challenge, forecasts were first developed for five key variables: Sales, Returns, Refills, Refurbishments and Disposals, using the company’s historical data. In the forecasting phase, two time series models were applied and compared: Prophet and Theta. The Prophet model outperformed in forecasting Returns, Refills and Sales, achieving mean absolute percentage errors of 8.92%, 8.83%, and 10.81%, respectively. For forecasting Refurbishments and Disposals, the Theta method performed better results, with MAPE values of 15.79% and 23.28%, respectively. The forecasts generated served as input for the optimization model, whose objective function aimed to minimize the total cost, consisting of both economic and environmental costs, through the incorporation of the Cap-and-Trade metric. The implementation of the model led to a 27.17% reduction in total costs, a 24.8% decrease in economic costs and a 29.25% reduction in carbon emissions compared to the company’s initial situation. The results demonstrate the applicability of the formulated model, highlighting its potential for implementation in other industrial contexts and contributing to the development of more efficient and environmentally responsible supply chains.
Description
Keywords
Closed-Loop Supply Chain Inventory Management Carbon Emissions Time Series Forecasting Prophet Theta Cadeias de abastecimento em ciclo fechado Gestão de inventários Emissões de carbono Previsão de séries temporais