Search results

The purpose of this paper is to solve large-scale many-to-many hub location-routing problem (MMHLRP) using variable neighborhood search (VNS). The MMHLRP is a combination of a single allocation hub location and traveling salesman problems that are known as one of the new fields in routing problems. MMHLRP is considered NP-hard since the two sub-problems are NP-hard. To date, only the Benders decomposition (BD) algorithm and the variable neighborhood particle swarm optimization (VNPSO) algorithm have been applied to solve the MMHLRP model with ten nodes and more (up to 300 nodes), respectively. In this research, the VNS method is suggested to solve large-scale MMHLRP (up to 1,000 nodes).

Generated MMHLRP sample tests in the previous work were considered and were added to them. In total, 35 sample tests of MMHLRP models between 10 and 1,000 nodes were applied. Three methods (BD, VNPSO and VNS algorithms) were run by a computer to solve the generated sample tests of MMHLRP. The maximum available time for solving the sample tests was 6 h. Accuracy (value of objective function solution) and speed (CPU time consumption) were considered as two major criteria for comparing the mentioned methods.

Based on the results, the VNS algorithm was more efficient than VNPSO for solving the MMHLRP sample tests with 10–440 nodes. It had many similarities with the exact BD algorithm with ten nodes. In large-scale MMHLRP (sample tests with more than 440 nodes (up to 1,000 nodes)), the previously suggested methods were disabled to solve the problem and the VNS was the only method for solving samples after 6 h.

The computational results indicated that the VNS algorithm has a notable efficiency in comparison to the rival algorithm (VNPSO) in order to solve large-scale MMHLRP. According to the computational results, in the situation that the problems were solved for 6 h using both VNS and VNPSO, VNS solved the problems with more accuracy and speed. Additionally, VNS can only solve large-scale MMHLRPs with more than 440 nodes (up to 1,000 nodes) during 6 h.

This paper aims to build a novel model and approach that assist an aircraft MRO procurement and overhaul management problems from the perspective of aircraft maintenance service provider, in order to ensure its smoothness maintenance activities implementation. The mathematical model utilizes the data related to warehouse inventory management, incoming customer service planning as well as risk forecast and control management at the decision-making stage, which facilitates to alleviate the negative impact of the uncertain maintenance demands on the MRO spare parts inventory management operations.

A stochastic model is proposed to formulate the problem to minimize the impact of uncertain maintenance demands, which provides flexible procurement and overhaul strategies. A Benders decomposition algorithm is proposed to solve large-scale problem instances given the structure of the mathematical model.

Compared with the default branch-and-bound algorithm, the computational results suggest that the proposed Benders decomposition algorithm increases convergence speed.

The results among the same group of problem instances suggest the robustness of Benders decomposition in tackling instances with different number of stochastic scenarios involved.

Extending the proposed model and algorithm to a decision support system is possible, which utilizes the databases from enterprise's service planning and management information systems.

A novel decision-making model for the integrated rotable and expendable MRO spare parts planning problem under uncertain environment is developed, which is formulated as a two-stage stochastic programming model.

To solve the problem, a mathematical model is proposed; it relies on a directed acyclic graph (DAG), in which arcs are used to indicate whether a pair of appointments can be assigned to the same route or not (and so to the same care worker). The proposed model aims at minimizing the personnel required to meet daily demand and balancing workloads among the workers while considering the varying traffic patterns derived from traffic congestion.

This paper aims at providing solution approaches for addressing the problem of assigning care workers to deliver home health-care (HHC) services, demanding different skills each. First, a capacity planning problem is considered, where it is necessary to define the number of workers required to satisfy patients' requests and then, patients are assigned to the care workers along with the sequence followed to visit them, thus solving a scheduling problem. The benefits obtained by permitting patients to propose multiple time slots where they can be served are also explored.

The results indicate that the problem can be efficiently solved for medium-sized instances, that is, up to 100 daily patient requests. It is also indicated that asking patients to propose several moments when they can receive services helps to minimize the need for care workers through more efficient route allocations without affecting significantly the balance of the workloads.

This article provides a new framework for modeling and solving a HHC routing problem with multiskilled personnel. The proposed model can be used to identify efficient daily plans and can handle realistic characteristics such as time-dependent travel times or be extended to other real-life applications such as maintenance scheduling problems.

The execution of product recovery strategies and the definition of an adequate system to manage its performance are crucial to move toward the employment of a successful circular economy (CE) concept. Defining strategies for the efficient management of product recovery requires product data that is difficult to obtain, making it harder to handle. However, efficient product recovery management can play a key role in shifting companies from a linear economy model to a more sustainable CE model, providing economic benefits and increasing customer satisfaction by recovering and adding value to the discarded product. Therefore, this study aims to provide better models to support decision-making and to evaluate product recovery performance.

The present study highlights a comprehensive two-stage decision approach to identify and examine the relevant key performance indicators (KPIs) for performance improvement of an information facilitated product recovery system (IFPRS) in a CE context. In the first phase, a structural equation modeling (SEM) methodology is adopted to categorize the KPIs by employing exploratory factor analysis and measurement of the model fit is obtained using the confirmatory factor analysis. Further, in the second phase, the KPIs are ranked and prioritized on the basis of expert’s recommendations adopting fuzzy-technique for order of preference by similarity to ideal solution (FTOPSIS).

Empirical investigation is conducted by compiling data from an association of six decision-makers (DMs) and two DMs from a respective prospect. The results highlight that “Technology Capacity” is ranked as the highest and is the most prominent KPI for successful employment of IFPRS practices. The results of the study would benefit policy makers and company directors in the selection of KPIs based on their importance in a context of high competition and greater pressure to adopt sustainable practices in the management of their companies.

As far as the authors know, no study has been performed till date to identify and construct a structural KPIs model for IFPRS performance improvement in the context of CE. The paper, therefore, proposes a two-phase SEM-TOPSIS technique to measure the impact of KPIs which is a new integration in the existing literature. The results of the study would benefit policy makers and company directors in the selection of KPIs based on their importance in a context of high competition and greater pressure to adopt sustainable practices in managing their organizations.