A carregar...
Publicação
Desenvolvimento de um sistema de apoio à decisão para o planeamento de rotas de entrega no setor do mobiliário
| Nome: | Descrição: | Tamanho: | Formato: | |
|---|---|---|---|---|
| 4.77 MB | Adobe PDF |
Autores
Orientador(es)
Resumo(s)
Este projeto surge no âmbito da conclusão do Mestrado em Engenharia e Gestão Industrial pelo
Instituto Superior de Engenharia do Porto (ISEP), desenvolvido em contexto de estágio na
empresa Laskasas, fabricante e distribuidora de mobiliário sediada em Rebordosa.
O trabalho centra-se no desenvolvimento de um sistema de apoio à decisão para o
planeamento de rotas de entrega, respondendo a um processo manual que se revelava
ineficiente, dependente do conhecimento tácito de um número reduzido de colaboradores e
sem qualquer suporte estruturado de apoio à decisão.
No que respeita à revisão da literatura, foram abordados os conceitos fundamentais da logística
de distribuição, da qualidade do serviço logístico e da informatização de processos. O
enquadramento teórico centrou-se no Vehicle Routing Problem (VRP) e na sua variante com
janelas temporais (VRPTW), bem como nos principais algoritmos heurísticos utilizados na sua
resolução, nomeadamente o Vizinho Mais Próximo, o Clarke & Wright e o Guided Local Search.
O diagnóstico do processo existente permitiu identificar as principais ineficiências e definir os
requisitos funcionais da solução. Foram desenvolvidos dois módulos complementares: um
sistema de planeamento de rotas baseado em Python e Flask, integrando geocodificação via
Nominatim, cálculo de matrizes de tempo e distância via OSRM e otimização através do solver
VRPTW da biblioteca OR-Tools; e um sistema de notificação SMS autónomo para automatização
do envio de lembretes aos clientes.
A avaliação dos resultados demonstrou uma redução de 87% no tempo de planeamento
semanal, de 85% no tempo dedicado ao envio de lembretes, uma redução média de 12,3% na
distância percorrida e de 16,6% no tempo de condução nas rotas comparadas, com uma
poupança financeira anual estimada de 45.646 euros.
This project was developed as part of the completion of the Master's degree in Industrial Engineering and Management at the Instituto Superior de Engenharia do Porto (ISEP), carried out in the context of a placement at Laskasas, a furniture manufacturer and distributor based in Rebordosa, Portugal. The work focuses on the development of a decision support system for delivery route planning, addressing a manual process that has proved inefficient, heavily reliant on the tacit knowledge of a small number of staff, and lacking any structured decision-making support. The literature review covered the fundamental concepts of distribution logistics, logistics service quality and process digitalization. The theoretical framework focused on the Vehicle Routing Problem (VRP) and its time-windowed variant (VRPTW), as well as the principal heuristic algorithms employed in its resolution, namely the Nearest Neighbor heuristic, the Clarke & Wright Savings Algorithm and Guided Local Search. The diagnosis of the existing process enabled the identification of the main inefficiencies and the definition of the functional requirements for the solution. Two complementary modules were developed: a route planning system built in Python and Flask, integrating address geocoding via Nominatim, time and distance matrix calculation via OSRM, and optimization through the OR-Tools VRPTW solver; and a standalone SMS notification system to automate delivery reminders to customers. The evaluation of results demonstrated an 87% reduction in weekly planning time, an 85% reduction in the time spent sending reminders, an average reduction of 12.3% in distance travelled and 16.6% in driving time across the routes assessed, with an estimated annual financial saving of €45,646.
This project was developed as part of the completion of the Master's degree in Industrial Engineering and Management at the Instituto Superior de Engenharia do Porto (ISEP), carried out in the context of a placement at Laskasas, a furniture manufacturer and distributor based in Rebordosa, Portugal. The work focuses on the development of a decision support system for delivery route planning, addressing a manual process that has proved inefficient, heavily reliant on the tacit knowledge of a small number of staff, and lacking any structured decision-making support. The literature review covered the fundamental concepts of distribution logistics, logistics service quality and process digitalization. The theoretical framework focused on the Vehicle Routing Problem (VRP) and its time-windowed variant (VRPTW), as well as the principal heuristic algorithms employed in its resolution, namely the Nearest Neighbor heuristic, the Clarke & Wright Savings Algorithm and Guided Local Search. The diagnosis of the existing process enabled the identification of the main inefficiencies and the definition of the functional requirements for the solution. Two complementary modules were developed: a route planning system built in Python and Flask, integrating address geocoding via Nominatim, time and distance matrix calculation via OSRM, and optimization through the OR-Tools VRPTW solver; and a standalone SMS notification system to automate delivery reminders to customers. The evaluation of results demonstrated an 87% reduction in weekly planning time, an 85% reduction in the time spent sending reminders, an average reduction of 12.3% in distance travelled and 16.6% in driving time across the routes assessed, with an estimated annual financial saving of €45,646.
Descrição
Palavras-chave
Distribution Logistics Route Optimization Vehicle Routing Problem VRPTW Decision Support System Automation Logística de distribuição Otimização de rotas Sistema de apoio à decisão Automatização