Parcels delivery in city centers is a topical issue in major cities and metropolis. To meet this problem, alternatives delivery modes have been proposed such as for instance drones, robots, bikes as well a combination of these with trucks. In this paper, we compare two delivery modes: trucks alone and drone-equipped trucks. Our comparison is based on an exact approach using an improved Branch-Cut-and-Price algorithm to solve the two-echelon vehicle routing problem with drones (VRPD) and without drones (VRP). Experiments are carried out on instances of literature and real data from the city of Bordeaux. The performance indicators used for comparison are the duration of the tour and the number of vehicles used; in addition, we carry out a preliminary analysis of tour costs and emissions.

Currently, the production, distribution and consumption modes disrupt ecosystems. To preserve the environment, circular economy (CE) proposes to reduce the environmental footprint in a social approach. Among the principles of CE, industrial ecology seeks to limit the impact of industries on the environment by involving them in collaboration schemes through industrial symbioses (IS). IS enables the upcycling of a company’s waste, called by-products, by another company as raw material. In this context, the goal of this study is to optimize the production planning of supply chains involved in an IS. We formulate this problem as a multi-level collaborative lot-sizing problem applied to industrial symbiosis. Thus, we introduce and define a new problem class: Centralized Collaborative Planning of an Industrial Symbiosis with N actors (CCP-IS-N). First, a new mixed-integer linear (MILP) model is presented, and then the complexity of the problem is studied. Finally, the proposed model is analyzed to understand its behavior and draw some key properties allowing to identify situation(s) in which IS is economically interesting or not.

The delivery of parcels in city centers is nowadays a topical issue and a major challenge for the local authorities. In this paper, we tackle the parcels distribution network in the city of Bordeaux. This city is at the very beginning of the process of moving from direct deliveries towards city center, to a more rational organization based on joint deliveries. We focus on the potential contributions of a two-tier distribution system made up of warehouses, proximity logistics spaces and intermediate depots so-called urban consolidation centers (UCCs). We seek to optimize the costs of the distribution network (i.e. transport and UCCs) and the selection of UCCs, while assessing the impact in terms of sustainable development by measuring CO2 emissions. We analyze and compare different scenarios using different deliveries' variants (with or without UCCs), different total UCCs capacities and different truck speeds. The findings demonstrate the benefits of using UCCs for city center distribution. The reductions in total cost and CO2 are most significant when the size of the UCCs is balanced with the total deliveries demands.

The Physical Internet (PI) is an innovative concept that has the potential to significantly improve the efficiency, cost-effectiveness, and sustainability of the global supply chain industry, particularly in cross-docking operations. This paper addresses the truck-scheduling problem in rail–road PI-Hubs, taking into account simultaneously both uncertainty and multi-objective decision-making, which has not been fully explored in the literature, particularly for PI-structures. Our proposed approach defines a Multi-Objective Mixed-Integer Programming model (FMO-MIP) that incorporates fuzzy chance-constrained programming and ϵ-constraint to minimize both the total delay and the sum of PI-containers traveled distances, while considering the uncertainty on truck arrival times. This work takes into account the particularities of the Physical Internet and presents a novel decision-making solution to generate a robust Pareto front that aligns with decision-makers' attitudes towards risk (optimistic/pessimistic) while balancing trade-offs between conflicting objectives.

Thanks to the globalization, and the large worldwide offer for supply, making the right choice would make a huge difference in terms of client satisfaction and cost minimization. This paper presents an optimized Demand Driven Material Requirements Planning (DDMRP) implementation scenario in a multi-echelon supply chain with guaranteed service, for a single product manufacturing process resourced by multiple suppliers. As required by the DDMRP method, a scenario where the demand is known over a planning horizon is considered. The lead times are deterministic. The raw materials and outsourced parts are delivered from a set of approved suppliers. The supplier selection process is coupled with the strategic buffer positioning and the buffer profiling in one model. This allows the decision maker to set the optimal inventory positioning with the best selected suppliers. Two non-linear MILP models are introduced and solved using CPLEX solver for multiple instances. The results provide an insight of the impact of the supplier selection on the DDMRP buffer positioning.

To operate a cross-dock successfully, an efficient assignment of trucks to docks is one of the key decisions. In this paper, we are interested in the cross-dock assignment of trucks to docks problem, where the number of trucks exceeds the number of docks. The objective is to minimize the cost of transferring goods within the cross-dock while avoiding delivery penalties. This problem being NP-hard, we use Variable Neighborhood Search metaheurisitc (VNS) to solve it approximately. More specifically, we conduct a structured empirical study to compare several VNS configurations and to find which is/are the most effective for this cross-dock problem. In this work, first we analyze the way the search strategy and the neighborhood operators can be combined in a VNS framework according to their efficiency within a local search. Then the best configurations are tested within three VNS variants, namely Basic VNS (BVNS), General VNS (GVNS) using Basic VND (B-VND) and GVNS using Union VND (U-VND) according to the number of used operators and the order of applying these operators. Finally we evaluate the influence of the stopping criterion within these variants. Some significant differences among these configurations are shown and illustrated by conducting the Friedman test.

Firms are more and more interested in reducing the carbon footprint related to their activity. Their supply chain remains one of the main sources of carbon emissions. Better operational routines and planning adjustments have proven to be an effective way to reduce the carbon emissions but not enough. In this paper, the carbon footprint is taken into consideration in the Demand Driven Material Requirements Planning (DDMRP) strategic buffer positioning problem. The focus is put on the storing activities and transportation, for which a function was proposed to quantify the associated emissions. Two environmental regulations are simulated: carbon emissions tax and carbon emissions cap. These approaches have been implemented in a buffer positioning model with a cost minimization objective function. The resulting models were solved using CPLEX solver for multiple instances. The numerical results provide a better measurement of the buffer positioning impact on the carbon emissions. The observed sensibility of the problem to the environment parameters would give insights for further research work.

This paper discusses the buffer positioning problem in a hybrid Make-To-Order/Make-To-Stock (MTO/MTS) setting with deterministic lead times. With a more demanding market, delivery times have become more competitive. A dynamic amount of inventory is made immediately available at certain points of the manufacturing process, including the final product, allowing to improve the flow of materials and meet more competitive deadlines. This work is particularly interested in optimizing buffer positioning in the context of Demand Driven Material Requirement Planning (DDMRP), where the decision maker can process the location and the sizing of the inventory buffers in complex Bills of materials (BOMs) in the most efficient strategy all while minimizing inventory costs under service time constraint. To do so, an improvement of the existing model of the two first steps of DDMRP: the strategic inventory positioning and calculating the buffer levels, is proposed. A linearization of the model and better solutions in shorter computational times compared to the literature are given by the enhanced model which is proven through various simulations. A cost objective function was evaluated in this model, and was solved using CPLEX and its CP Optimizer. These simulations would prove a better and more efficient approach for an enhanced DDMRP implementation.

To make a product, many components: made inside a supply chain or provided by external suppliers, may be needed as specified in the bill of material (BOM). This variety exposes the production process to various risks as stock ruptures for instance. Due to companies efforts to face the volatile demand and market pressure, Work-in-process (WIP) inventories become redundant in flow production lines. Uncertainty became a major flaw of the system, to which many methods were developed and adopted to manage and optimize the use of resources. One of the most recent methods is Demand Driven Material Requirement Planning (DDMRP) that uses buffers: visual indicators of the net flow integrating the real stock, the actual demand and the expected provisions; This paper describes an approach to solve the optimization problem of minimizing the storing costs, under service time constraint for uncapacitated buffer positioning alongside with the buffer allocation following the DDMRP planning methodology in a hybrid MTO/MTS manufacturing system. Finally, numerical examples are applied to test the efficiency and the performance of the developed algorithm.

For decades, the industry is a huge consumer of resources, generates waste and has serious impacts on our planet. In order to preserve the environment, new economic models are emerging including the circular economy. Circular economy is a system that fits with sustainability and aims to reduce both resources consumption and waste production. Among the components of circular economy, industrial ecology is the collaboration of industries on a territory in order to exchange materials, water or energy through industrial symbiosis. Industrial ecology and industrial symbiosis are more and more present in the literature. However, most of articles in the literature focused on the strategic decision level. In this paper, we focus on tactical decisions, and investigate production planning as a lot sizing problem in a collaborative context involving two actors belong an industrial symbiosis.

Demand forecasting plays an important role for supply chains decision making. It also represents a basis step for activity planning in response to customer demand. In this paper, recent advances in times series allow a new robust and easy approach for demand forecasting. Such technique which is based on algebraic methods of estimation seems to be more adequate for such supply chains task. Several computer simulations and comparative studies demonstrate the relevance of the proposed approach.

We propose a method that combines the paradigms of Column Generation (CG) and Iterated Local Search (ILS) to solve the Capacitated Arc-Routing problem. One of the goals is to integrate into the ILS (some of) the duality information that underpins the CG. We consider a space of permutations and sub-permutations (sequences) of the set of required edges [ 1..m ] . This space is explored by an ILS process and a CG process that run in parallel and that can repeatedly exchange sub-permutations. The ILS uses an exact decoder that maps any permutation s : [ 1..m ] → [ 1..m ] to a list of sequences (routes) of minimum cost that services [ 1..m ] in the order s(1), s(2), . . . , s(m). The first use of the CG paradigm in ILS is the following: all sequences discovered by the CG process are sent to the ILS process that can inject them into the current ILS solution. The second application of CG in ILS consists of a “CG improver” that starts from the current ILS solution and tries to improve it by running several CG iterations. The first half of the paper describes the proposed method in a general framework based on sequences, permutations and set covers. The second part is devoted to more specialized Arc-Routing techniques. For instance, the CG convergence could be accelerated by factors of tens or even hundreds by exploiting two ideas in the Dynamic Programming (DP) pricing: (i) avoid as much as possible to traverse edges without service before the end of the CG process, and (ii) detect and prune dominated DP states by recording them in a fast data structure that relies on an array and a red-black tree. Regarding the ILS, we show that the permutation-level search can be substantially improved if the exact decoder is reinforced with a deterministic post-decoding operator that acts on explicit routes. The overall results are competitive (reducing the best-known gap of five instances) and certain ideas could be potentially useful for other set-covering or permutation search problems.

We propose a method that combines Column Generation (CG) and Iterated Local Search (ILS) to solve the Capacitated Arc -Routing Problem (CARP). One of the goals is to integrate into the ILS (some of) the duality information that underpins the CG paradigm. The CARP is expressed in a space of permutations and sub-permutations (sequences, routes) of the set of required edges. For this, the ILS uses an exact decoder that maps any permutation s to a list of sequences (routes) of minimum cost servicing all edges in the order s(1), s(2), s(3), etc. This permutation space is explored both by an ILS process and a CG process that run in parallel and that communicate by exchanging sequences. The first use of the CG paradigm in ILS is the following: all sequences discovered by CG are sent to the ILS process that can inject them into the current ILS solution. The second application of CG in ILS is a "CG improver" operator that acts on the current ILS solution, so as to (try to) improve it by running several CG iterations. The first half of the paper describes the method in a general framework based on sequences, permutations and set covering. The second part is devoted to more specialized Arc -Routing techniques. For instance, the CG convergence could be accelerated by factors of tens or even hundreds by exploiting two ideas in the Dynamic Programming (DP) pricing: (i) avoid as much as possible to traverse edges without service, and (ii) record only non-dominated DP states using a fast-access data structure mixing an array and a red-black tree. Regarding the ILS, we show that the permutation-level search can be substantially improved if the exact decoder is reinforced by a deterministic post-decoder acting on explicit routes. The general results are competitive (reducing the best-known gap of five instances) and certain ideas could be potentially useful for other set-covering or permutation problems. (C) 2016 Elsevier B.V. All rights reserved.

Dynamic supply chain control is tackled in this paper using the concept of differential flatness where we focus on operational activities and algebraic estimation techniques for the uncertain parameters. The main objective is to maintain the inventory level at a desired one according to the customer demand while synchronizing the flows within a set of physical boundaries and constraints. In addition, we consider uncertainties on delay and adaptation times and we use the new setting of the algebraic estimation techniques to adapt and update the controller when any change occurs. These two methods lead to a simple design of the control algorithm without integration of any differential equation and without a need to any model to estimate, which yield to an efficient control action. Convincing numerical simulations using a case study of a crude oil blending and distribution system demonstrate the effectiveness of the proposed approach. Copyright 2017 IFAC.

Dynamic supply chain management is tackled, in this paper, using the concept of differential flatness. Such approach seems to be more adequate with supply chains, which are complex and meshed systems. The paper focuses on the operational activities of supply-chains, the main objective is to maintain the inventory level at a desired one according to the customer demand while synchronizing the flows within a set of physical boundaries and constraints. The main advantage of differential flatness is characterized by the fact that the whole system behavior is described by the trajectory of a so-called “flat output” and a number of its successive time derivatives. This leads to a simple design of the control algorithm, without integration of any differential equation. Convincing numerical simulations, using a case study of a multi-product polyethylene petrochemical plant, demonstrate the effectiveness of the proposed approach.

We propose a method that combines the paradigms of Column Generation (CG) and Iterated Local Search (ILS) to solve the Capacitated Arc-Routing problem. One of the goals is to integrate into the ILS (some of) the duality information that underpins the CG. We consider a space of permutations and sub-permutations (sequences) of the set of required edges [ 1..m ] . This space is explored by an ILS process and a CG process that run in parallel and that can repeatedly exchange sub-permutations. The ILS uses an exact decoder that maps any permutation s : [ 1..m ] → [ 1..m ] to a list of sequences (routes) of minimum cost that services [ 1..m ] in the order s(1), s(2), . . . , s(m). The first use of the CG paradigm in ILS is the following: all sequences discovered by the CG process are sent to the ILS process that can inject them into the current ILS solution. The second application of CG in ILS consists of a “CG improver” that starts from the current ILS solution and tries to improve it by running several CG iterations. The first half of the paper describes the proposed method in a general framework based on sequences, permutations and set covers. The second part is devoted to more specialized Arc-Routing techniques. For instance, the CG convergence could be accelerated by factors of tens or even hundreds by exploiting two ideas in the Dynamic Programming (DP) pricing: (i) avoid as much as possible to traverse edges without service before the end of the CG process, and (ii) detect and prune dominated DP states by recording them in a fast data structure that relies on an array and a red-black tree. Regarding the ILS, we show that the permutation-level search can be substantially improved if the exact decoder is reinforced with a deterministic post-decoding operator that acts on explicit routes. The overall results are competitive (reducing the best-known gap of five instances) and certain ideas could be potentially useful for other set-covering or permutation search problems.

Rising vehicles number and increased use of private cars have caused significant traffic congestion, noise and energy waste. Public transport cannot always be set up in the non-urban areas. Car pooling, which is based on the idea that sets of car owners having the same travel destination share their vehicles has emerged to be a viable possibility to reduce private car usage around the world. In this paper, we present a multi-agent based self-adaptive genetic algorithm to solve long-term car pooling problem. The system is a combination of multi-agent system and genetic paradigm, and guided by a hyper-heuristic dynamically adapted by a collective learning process. The aim of our research is to solve the long-term car pooling problem efficiently with limited exploration of the search space. The proposed algorithm is tested using large scale instance data sets. The computational results show that the proposed method is competitive with other known approaches for solving long-term car pooling problem.

In this paper, we are interested in the truck dock assignment problem in cross-docks with several constraints. A cross-dock is a warehouse with limited storage capacity. Arriving trucks are assigned to docks and have to deliver goods to a warehouse. These goods are sorted out and loaded into the outbound trucks immediately. The objective is to find an optimal dock assignment for each truck in order to minimize the cost of transferring shipments within the cross-dock while avoiding delivery problems. To solve this problem, we propose several meta-heuristics. The first one is a local search that is characterized by different neighboring operators. These operators are tested and compared one by one on various instances of the cross-docking problem. The second metaheuristic is a Variable Neighborhood Search (VNS). According to the efficiency of the local search operators, we test the way they can be combined in a VNS structure. Our experimental study shows that VNS makes use of the operators depending on their properties.