A GA Based Heuristic for the Vehicle Routing Problem with Multiple Trips

A Nearest Neighbor Algorithm to Optimize Recycling Networks

This article analyses the processes of collecting used non-returnable packaging to improve the recycling of material. A collection system is proposed by applying a profitable visit algorithm based on the widely-known Nearest Neighbor Algorithm. A comparative study is performed to achieve a higher volume of recycled material while decreasing the cost of collection. The proposed algorithm shows a much better performance than the reference. The developed algorithm was evaluated in a real scenario and confirmed by a simulation runs. Savings in material sourcing processes can be achieved in real operations. The proposed algorithm shows some advantage.

An Improved Genetic Algorithm for Solving Multi Depot Vehicle Routing Problems

The classical Vehicle Routing Problem (VRP) tries to minimise the cost of dispatching goods from depots to customers using vehicles with limited carrying capacity. As a generalisation of the TSP, the problem is known to be NP-hard and several authors have proposed heuristics and meta-heuristics for obtaining good solutions. The authors present genetic algorithm-based approaches for solving the problem and compare the results with available results from other papers, in particular, the hybrid clustering based genetic algorithm. The authors find that the proposed methods give encouraging results on all these instances. The approach can be extended to solve multi depot VRPs with heterogeneous fleet of vehicles.

Emergency evacuation planning in natural disasters under diverse population and fleet characteristics

Purpose

The purpose of this paper is to formulate and solve a new emergency evacuation planning problem. This problem addresses the needs of both able and disabled persons who are evacuated from multiple pick-up locations and transported using a heterogeneous fleet of vehicles.

Design/methodology/approach

The problem is formulated using a mixed integer linear programming model and solved using a heuristic algorithm. The authors analyze the selected heuristic with respect to key parameters and use it to address theoretical and practical case studies.

Findings

Evacuating people with disabilities has a significant impact on total evacuation time, due to increased loading/unloading times. Additionally, increasing the number of large capacity vehicles adapted to transport individuals with disabilities benefits total evacuation time.

Research limitations/implications

The mathematical model is of high complexity and it is not possible to obtain exact solutions in reasonable computational times. The efficiency of the heuristic has not been analyzed with respect to optimality.

Practical implications

Solving the problem by a heuristic provides a fast solution, a requirement in emergency evacuation cases, especially when the state of the theater of the emergency changes dynamically. The parametric analysis of the heuristic provides valuable insights in improving an emergency evacuation system.

Social implications

Efficient population evacuation studied in this work may save lives. This is especially critical for disabled evacuees, the evacuation of whom requires longer operational times.

Originality/value

The authors consider a population that comprises able and disabled individuals, the latter with varying degrees of disability. The authors also consider a heterogeneous fleet of vehicles, which perform multiple trips during the evacuation process.

Reverse logistics of recovery and recycling of non-returnable beverage containers in the brewery industry

Purpose

– The purpose of this paper is to analyse the processes of collecting used non-returnable packaging. The objective is to increase the quality and quantity of recycling material used in the production of new packages.

Design/methodology/approach

– A reverse logistics network is proposed, nevertheless the focus of the study is addressing the collection routing problem. Specifically a new “profitable visit algorithm” (PVA) based on the well-known Nearest Neighbor is proposed and tested, in both real and simulated scenarios, in order to achieve higher volume of collected material while lowering the cost of collection.

Findings

– The proposed algorithm is set to compete against the well-known Nearest Neighbor. The “PVA” showed a far better performance than the Nearest Neighbor. The latter was assessed with real data in a real scenario and was also confirmed by a simulation.

Research limitations/implications

– Further research should envisage stochastic models and tighter time constraints.

Practical implications

– Managers can use the algorithm in similar scenarios to address different routing needs. New business activities could be foreseen by starting a logistics scheme of this type. Additionally, new savings in material procurement processes could be achieved in current operations.

Social implications

– Social responsibility and environmental awareness can be approached by this type of endeavors. While reducing waste disposal, the recovering of these materials yields savings and offer new labor opportunities for people.

Originality/value

– The “PVA” introduced in this paper, showed substantial advantages. First it automatically determines the visit frequency for each client. Also, it evaluates whether a visit should be granted or not based on its “profitability.” The latter ensures that even if the routing circuit is broken at any moment, the company will not lose “profit.” Additionally, this research provides performance evaluation in real and simulated scenarios, which is hard to obtain in a single study.

Keywords

 Reverse logistics, Recycling, Routing algorithm, Non-returnable packages, Real scenarios

Paper type

 Research type

Resumen

Objetivo

Objetivo

– Este artículo analiza los procesos de recolección deenvases no retornable usados. El objetivo es aumentar la calidad y cantidad de material reciclable usado en la producción de nuevos envasados.

Diseño/metodología/enfoque

– Se propone una red de logística reversa, no obstante el enfoque del estudio es abordar el problema de la ruta de recolección. Específicamente, se propone un nuevo “algoritmo de visita rentable” basado en el bien conocido Vecino Más Próximo y se testea en ambos escenarios reales y simulados, para poder lograr un volumen más alto de material recolectado y a su vez disminuir los costo de recolección.

Hallazgos

– El algoritmo propuesto está definido para competir contra el bien conocido Vecino Más próximo. El “algoritmo de visita rentable” mostro un rendimiento mucho mejor que el de Vecino Más próximo. Este último se evaluó con datos reales en un escenario real y también se confirmo a través de una simulación.

Limitaciones/implicaciones del estudio

– Investigaciones futuras deberían visualizar modelos estocásticos y restricciones de tiempo más limitadas.

Implicaciones practices

– Administradores pueden usar el algoritmo en escenarios similares para dirigir las diferentes necesidades de la ruta. Nuevas actividades de negocios podrían ser previstas iniciando un esquema logístico de este tipo. Adicionalmente, nuevos ahorros en procesos de material de adquisición podría ser logrado en las operaciones actuales.

Implicaciones sociales

– La responsabilidad social y conciencia ambiental puede ser abarcado con este tipo de actividad. No sólo se reducen los residuos, pero también la recuperación de estos materiales genera ahorros y ofrece nuevas oportunidades de trabajo para las personas.

Originalidad/valor

– El “algoritmo visita rentable” introducido en este articulo, mostró ventajas substanciales. Primero, determina automáticamente la frecuencia de visitas para cada cliente. Además, evalúa si una visita basada en su “rentabilidad” debería ser dada o no. Esta última asegura que aunque el circuito de la ruta se quiebra en cualquier momento, la empresa no perderá “ganancias.” Adicionalmente, esta investigación provee evaluaciones de desempeño en escenarios reales y simulados, lo cual es difícil de obtener en un sólo estudio.

Tipo de papel

Trabajo de investigación

Resumo

Objetivo

Objetivo

– Este artigo analisa o processo de coletade embalagens não retornáveis usadas. O objetivo é aumentar a qualidade e as quantidades de materiais recicláveis utilizados na produção de novas embalagens.

Desenhos/metodologia/enfoque

– Uma rede de logística reversa é proposta, embora o enfoque da pesquisa é abordar o problema da roteamento da coleta. Especificamente, um novo algoritmo de “Profitable Visit” baseado no conhecido Nearest Neighboré proposto Dois cenários são testados, um real e um simulado, com objetivos deatingimento de um volume maior de material coletado e também de diminuir os custos decoleta.

Achados

– O algoritmo foi proposto para competir com o conhecido Nearest Neighbor. O algoritmo de “Profitable Visit” mostra um rendimento melhor que o Nearest Neighbor. Este último foi avaliado com dadosreais em um cenário real e também confirmado através de simulação.

Limitações/implicações da pesquisa

– Pesquisas futuras deveriam abordar modelos estocásticos e maiores restrições de tempo.

Implicações práticas

– Gerentes podem utilizar o algoritmo em cenários similares para gerenciar diferentes necessidades de roteamento de coleta. Novos processos de negócios poderiam ser previstos com base em projetos logísticos deste tipo. Adicionalmente, ganhos adicionais em processos de compra de materiais poderiam ser obtidosem operações presentes.

Implicações sociais

– A responsabilidade social e a consciência ambiental podem ser abordadas com este tipo de projeto. Junto com a eliminação de resíduos, a recuperação desses materiaistambém gera renda e oferece novas oportunidades de trabalho para as pessoas.

Originalidade/valor

– O algoritmo Profitable Visit” apresentadoneste artigo mostra vantagens substanciais. Primeiro, determina automaticamente a freqüência de visitas para cada cliente. Também, avalia se uma visita com base na “rentabilidade” deveria ocorrer ou não. Este aspecto assegura que quando o circuito da rota é modificado em qualquer momento, a empresa não perderá seus “ganhos.” Adicionalmente, esta pesquisa fornece avaliações de desempenho em cenários reais e simulados, o que é difícil de ocorrer em um único estudo.

Tipo de papel

Trabathos de pesquisa