Publications
Technical reports (submitted/under revision)
Abstract: The Vehicle Routing Problem with Synchronisation (VRPSync) aims to minimise the total routing costs while considering synchronisation requirements that must be fulfilled between tasks of different routes. These synchronisation requirements are especially relevant when it is necessary to have tasks being performed by vehicles within given temporal offsets, a frequent requirement in applications where multiple vehicles, crews, materials or other resources are involved in certain operations. Although several works in the literature have addressed this problem, mainly the deterministic version has been tackled so far. This paper presents a robust optimisation approach for the VRP with Synchronisation taking into consideration the uncertainty in vehicle travel times between customers. This work builds on existing approaches in the literature to develop mathematical models for the Robust VRP with Synchronisation, as well as a branch-and-cut algorithm to solve more difficult problem instances. A set of computational experiments is also devised and presented to obtain insights regarding key performance parameters of the mathematical models and the solution algorithm. The results suggest that solution strategies where certain standard problem constraints are only introduced if a candidate solution violates any of those constraints provide more consistent improvements than approaches that rely on tailor-made cutting planes, added through separation routines. Furthermore, the analysis of the Price of Robustness indicators shows that the adoption of robust solutions can have a significant increase in the total costs, however, this increase quickly plateaus as budgets of uncertainty increase.
Authors: Ricardo Soares · Sophie N. Parragh · Pedro Amorim · Alexandra Marques
Journal Articles
Abstract: The practical relevance and challenging nature of the Vehicle Routing Problem (VRP) have motivated the Operations Research community to consider different practical requirements and problem variants throughout the years. However, businesses still face increasingly specific and complex transportation requirements that need to be tackled, one of them being synchronisation. No literature contextualises synchronisation among other types of problem aspects of the VRP, increasing ambiguity in the nomenclature used by the community. The contributions of this paper originate from a literature review and are threefold. First, new conceptual and classification schemas are proposed to analyse literature and re-organise different interdependencies that arise in routing decisions. Secondly, a modelling framework is presented based on the proposed schemas. Finally, an extensive literature review identifies future research gaps and opportunities in the field of VRPs with synchronisation.
Authors: Ricardo Soares · Alexandra Marques · Pedro Amorim · Sophie N. Parragh
Abstract: This paper addresses the integration of the planning decisions concerning inbound logistics in an industrial setting (from the suppliers to the mill) and outbound logistics (from the mill to customers). The goal is to find the minimum cost routing plan, which includes the cost-effective outbound and inbound daily routes (OIRs), consisting of a sequence of deliveries of customer orders, pickup of a full truck-load at a supplier, and its delivery to the mill. This study distinguishes between three planning strategies: opportunistic backhauling planning (OBP), integrated inbound and outbound planning (IIOP) and decoupled planning (DIOP), the latter being the commonly used, particularly in the case of the wood-based panel industry under study. From the point of view of process integration, OBP can be considered as an intermediate stage from DIOP to IIOP. The problem is modelled as a Vehicle Routing Problem with Backhauls, enriched with case-specific rules for visiting the backhaul, split deliveries to customers and the use of a heterogeneous fleet. A new fix-and-optimise matheuristic is proposed for this problem, seeking to obtain good quality solutions within a reasonable computational time. The results from its application to the wood-based panel industry in Portugal show that IIOP can help to reduce total costs in about 2.7%, when compared with DIOP, due to better use of the delivery truck and a reduction of the number of dedicated inbound routes. Regarding OBP, fostering the use of OIRs does not necessarily lead to better routing plans than DIOP, as it depends upon a favourable geographical configuration of the set of customers to be visited in a day, specifically, the relative distance between a linehaul that can be visited last in a route, a neighboring backhaul, and a mill. The paper further provides valuable managerial insights on how the routing plan is impacted by the values of business-related model parameters which are set by the planner with some degree of uncertainty. Results suggest that increasing the maximum length of the route will likely have the largest impact in reducing transportation costs. Moreover, increasing the value of a reward paid for visiting a backhaul can foster the percentage of OIR in the optimal routing plan.
Authors: Alexandra Marques · Ricardo Soares · Maria João Santos · Pedro Amorim
Abstract: The search for higher efficiency in transportation planning processes in real life applications is challenging. The synchronisation of different vehicles performing interrelated operations can enforce a better use of vehicle fleets and decrease travelled distances and non-productive times, leading to a reduction of logistics costs. In this work, the full truck-load pickup and delivery problem with multiple vehicle synchronisation (FT-PDP-mVS) is presented. This problem is motivated by a real-life application in the biomass supply chain “hot-system”, where it is necessary to simultaneously perform chipping and transportation operations at the forest roadside. The FT-PDP-mVS consists in determining the integrated routes for three distinct types of vehicles, which need to perform interrelated operations with minimum logistics costs. We extend existing studies in synchronisation of multiple routes by acknowledging several synchronisation aspects, such as operations and movement synchronisation. A novel mixed integer programming model (MIP) is presented, along with valid inequalities to tighten the formulation. A solution method approach is developed based on the fix-and-optimise principles under a variable neighbourhood decomposition search. Results of its application to 19 instances based on a real-world case-study demonstrate its performance. For a baseline instance, the synchronisation aspects tackled in this problem allowed for significant gains when compared to the company’s current planning approach. Furthermore, the proposed approach can enhance planning and decision making processes by providing valuable insights about the impact of key parameters of biomass logistics over the routing results.
Authors: Ricardo Soares · Alexandra Marques · Pedro Amorim · Jussi Rasinmäki
Abstract: The growing economic importance of the biomass-for-bioenergy in Europe motivates research on biomass supply chain design and planning. The temporally and geographically fragmented availability of woody biomass makes it particularly relevant to find cost-effective solutions for biomass production, storage and transportation up to the consumption facility. This paper addresses tactical decisions related with optimal allocation of wood chips from forest residues at forest sites to terminals and power plants. The emphasis is on a “hot-system” with synchronized chipping and chips transportation at the roadside. Thus, decisions related with the assignment of chippers to forest sites are also considered. We extend existing studies by considering the impact of the wood chips energy content variation in the logistics planning. This is a key issue in biomass-for-bioenergy supply chains. The higher the moisture content of wood chips, the lower its net caloric value and therefore, a larger amount of chips is needed to meet the contracted demand. We propose a Mixed Integer Programming (MIP) model to solve this problem to optimality. Results of applying the model in a biomass supply chain case in Finland are presented. Results suggest that a 20% improvement in the supplier profit can be obtained with the proposed approach when compared with a baseline situation that relies on empirical estimates for a fixed and known moisture content in the end of an obliged storage age.
Authors: Alexandra Marques · Jussi Rasinmäki · Ricardo Soares · Pedro Amorim
Conference Papers
Abstract: The potential of the Internet of Things (IoT) and other technologies in the realm of Industry 4.0 to generate valuable data for monitoring the performance of the production processes and the whole supply chain is well established. However, these large volumes of data can be used within planning and control systems (PCSs) to enhance real-time planning and decision-making. This paper conducts a literature review to envisage an overall system architecture that combines IoT and PCS for planning, monitoring and control of operations at the level of an industrial production process or at the level of its supply chain. Despite the extensive literature on IoT implementations, few studies explain the interactions between IoT and the components of PCS. It is expected that, with the increasing digitization of business processes, approaches with PCS and IoT become ubiquitous in the near future.
Authors: Ricardo Soares · Alexandra Marques · Reinaldo Gomes · Luís Guardão · Elder Hernández · Rui Rebelo
PhD Dissertations
Abstract: Across different value chains, transportation and logistics are at the core of research agendas, motivated by their impact on economic, social and environmental terms. In particular, the Vehicle Routing Problem (VRP) has been an effervescent field of research for the past 60 years. This can be attributed not only to the NP-hard combinatorial nature of the problem, thus constituting a challenge when solving it, but also to its practical relevance to organisations when planning their logistics and transportation activities. Despite the obtained breakthroughs by the Operations Research (OR) community, businesses still face increasingly specific and complex transportation requirements that need to be tackled. In this context, one of the challenges that have been emerging is synchronisation. The VRP with Synchronisation is a generalisation of the VRP where the feasibility of a route is dependent on the feasibility of another route. Examples of these problems typically include requirements where different vehicles must temporally synchronise their arrival to a customer (e.g., scheduling patient homecare) – either at the same time or within given time intervals – or when different vehicles need to be travelling simultaneously – as it happens with problems using tractors and trailers, for example. This thesis advances the state of the art on the topic of vehicle routing with synchronisation. Initially, this thesis systematises knowledge on the topic. This research was grounded on qualitative analyses based on an integrative literature review, which is also used to identify relevant research gaps. This thesis also tackles specific and practical challenges identified in the literature. One of these challenges consists of modelling and solving complex routing problems with multiple types of synchronisation constraints. To that effect, a systematic approach is proposed, motivated by a real-life case-study. The obtained results from this research provide valuable managerial insights that can be implemented in the context of the case-study. Furthermore, this thesis also tackles VRPs with Synchronisation acknowledging sources of uncertainty. Based on consolidated knowledge from the literature, modelling and solution approaches are developed and tested in a set of computational experiments. This thesis is complemented by a supplementary material section that presents relevant research work performed by the PhD candidate in the course of his training.