Papers by Angelo Sifaleras
Variable Neighborhood Search (ICVNS 2025), 2026
The Job Shop Scheduling Problem is a classic combinatorial optimization problem and one of the mo... more The Job Shop Scheduling Problem is a classic combinatorial optimization problem and one of the most well-studied scheduling problems. Several methodologies, both exact and metaheuristic, have already been proposed for the solution of this computationally difficult problem. This work presents for the first time a solution approach based on Variable Neighborhood Programming for the Job Shop Scheduling Problem. Variable Neighborhood Programming is a recent methodology which constitutes a combination of Genetic Programming and Variable Neighborhood Search. In addition, some encouraging comparative computational results are also shown against the state-of-the-art Gurobi optimization solver using medium-and large-scale benchmark instances. The findings of this work have a plethora of modern applications in Manufacturing-asa-Service online platforms. All experimental evaluations were performed on the Google Cloud Platform.
International Journal of Systems Science: Operations & Logistics, Dec 2025
This paper advances the Manufacturing-as-a-Service paradigm through the Tec4MaaSEs (T4M) project,... more This paper advances the Manufacturing-as-a-Service paradigm through the Tec4MaaSEs (T4M) project, where production and manufacturing processes are delivered as on-demand services using advanced Industry 4.0 and Industry 5.0 technologies, in order to create a resilient ecosystem of distributed value networks. This idea is based on a highly configurable Digital Twin architecture that dynamically adapts to fluctuations in supply and demand, enabling collaboration and optimization across diverse manufacturing scenarios and various stakeholders. Although Manufacturing-as-a-Service (MaaS) platforms promise to enable dynamic configuration of distributed production systems, most existing implementations exhibit limited ability to handle multi-actor processes involving diverse service types, heterogeneous data, and coordination needs. This work presents T4M, a MaaS framework that combines production planning, semantic interoperability, and service modeling to support the flexible composition of manufacturing value chains. A key innovation lies in its iterative feedback structure, allowing analytics and planning functions to co-evolve with service configurations. To structure the design space, we introduce a three-dimensional framework encompassing product-process variety, granularity, and Functional Integration Level (FILe). These dimensions guide the functional specification of platform services and determine where analytics and automation provide tangible value. The framework is instantiated through three representative value networks (VN1-VN3), each illustrating distinct demands in terms of information flows, coordination intensity, and decision complexity. Our analysis shows that effective MaaS ecosystems must align digital mechanisms not only with physical production resources, but also with the informational structure and functional logic of each setting. In particular, the volume-variety concept, and more specifically the notions of granularity and FILe, emerge as key enablers in identifying the level of platform integration and the appropriate scope in shaping MaaS ecosystems. These insights support the development of collaborative, resilient, and circular industry practices.
Swarm and Evolutionary Computation, 2025
Trained Reward-based Action Classification Engine (TRACE) is a general process for capturing oper... more Trained Reward-based Action Classification Engine (TRACE) is a general process for capturing operator outcome data during metaheuristic search, training classifiers to predict whether an operator will yield an improved solution, and deploying those models to guide neighborhood selection during future search runs. This study introduces TRACE-VNS, a modular extension of General Variable Neighborhood Search (GVNS) applied to the Capacitated Vehicle Routing Problem (CVRP), where neighborhood selection is driven by these offline-trained models. Classifiers are trained on features extracted from GVNS traces, including action history, graph metrics, temporal state, and Upper Confidence Bound (UCB) indicators. Twelve classifiers, including tree ensembles, neural networks, and kernel-based models, are benchmarked using the Precision-Recall Area Under the Curve (PR-AUC) to evaluate predictive quality. Empirical results show that TRACE-VNS improves convergence speed and final solution quality over conventional GVNS across 84 CVRP instances. A detailed feature importance analysis identifies strong contributors, offering insights into the effective selection of operators. TRACE requires no runtime exploration or feedback loops and can generalize to other metaheuristics through minimal structural adaptation.
Swarm and Evolutionary Computation, Feb 24, 2025
This study presents the design and integration of novel adaptive components within the Double-Ada... more This study presents the design and integration of novel adaptive components within the Double-Adaptive General Variable Neighborhood Search (DA-GVNS) algorithm, aimed at improving its overall efficiency. These adaptations utilize iteration-based data to refine the search process, with enhancements such as an adaptive reordering mechanism in the refinement phase and a knowledge-guided approach to adjust the search strategy. Additionally, an adaptive mechanism for dynamically controlling the shaking intensity was introduced. The proposed knowledge-guided adaptations demonstrated superior performance over the original DA-GVNS framework, with the most effective scheme selected for further evaluation. Initially, the symmetric Traveling Salesman Problem (TSP) was used as a benchmark to quantify the impact of these mechanisms, showing significant improvements through rigorous statistical analysis. A comparative study was then conducted against six advanced heuristics from the literature. Finally, the most promising knowledge-guided GVNS (KG-GVNS) was tested against the original DA-GVNS on selected instances of the Quadratic Assignment Problem (QAP), where detailed statistical analysis highlighted its competitive advantage and robustness in addressing complex combinatorial optimization problems.
Supply Chain Finance Modelling and Optimization, 2024
This work addresses a new Inventory-Routing Problem with realistic characteristics, such as stric... more This work addresses a new Inventory-Routing Problem with realistic characteristics, such as strict driving hours’ regulations and driving speed limits. According to these realistic assumptions, the time needed to perform a route will be strictly bounded by an upper time limit based on the driving hours’ regulations provided by the European Commission. Moreover, the decisions about the speed level selection, for traveling between two network points, will be subject to specific driving speed limits. Moreover, the impact of two inventory policies, the classic (R,Q) replenishment policy and the flexible replenishment policy, on the total supply chain network cost is studied. The consideration of two inventory policies leads to the development of two mixed-integer linear programming models. The computational experiments were conducted on random, small-sized, problem cases, using the state-of-the-art solver, Gurobi. The impact of the two inventory policies on the supply chain network is investigated through a numerical analysis.
Disruptive technologies and optimization towards Industry 4.0 logistics, 2024
Industry 4.0 era is characterized by several technological advances, such as the technology of Au... more Industry 4.0 era is characterized by several technological advances, such as the technology of Autonomous Vehicles, which is expected to increase mobility efficiency. A critical component in the autonomous vehicle navigation process is path planning. To this end, the present survey focused on the investigation of research contributions on the optimal path/routes scheduling of autonomous vehicles. The main objective of the conducted review is the classification of published relative research papers, according to their optimization criteria, optimization models, and optimization methods.
Metaheuristics (MIC 2024), 2024
Scheduling of reclaimers activities in dry bulk terminals significantly impact terminal throughpu... more Scheduling of reclaimers activities in dry bulk terminals significantly impact terminal throughput, a crucial performance indicator for such facilities. This study addresses the Reclaimer Scheduling Problem (RSP) while considering periodic preventive maintenance activities for reclaimers. These machines are integral for reclaiming dry bulk materials stored in stockyards, facilitating their loading onto vessels via ship-loaders. The primary aim of the objective function entails the minimization of the overall completion time, commonly referred to as the makespan. Since this problem is N P-hard, we propose a novel greedy constructive heuristic. The solutions obtained from this heuristic serve as the starting point for an efficient General Variable Neighborhood Search (GVNS) algorithm to handle medium-scale instances resembling real stockyard configurations. Computational experiments are conducted by comparing the proposed methods across various problem instances. The results demonstrate that the developed GVNS, coupled with the constructive heuristic for initial solution finding, efficiently improves scheduling efficacy. Thus, it emerges as a new state-of-the-art algorithm for this problem.
Metaheuristics (MIC 2024), 2024
This study explores the use of Autoregressive Integrated Moving Average (ARIMA) and Long Short-Te... more This study explores the use of Autoregressive Integrated Moving Average (ARIMA) and Long Short-Term Memory (LSTM) machine learning models in metaheuristic algorithms, with a focus on a modified General Variable Neighborhood Search (GVNS) for the Capacitated Vehicle Routing Problem (CVRP). We analyze the historical chain of actions in GVNS to demonstrate the predictive potential of these models for guiding future heuristic applications or parameter settings in metaheuristics such as Genetic Algorithms (GA) or Simulated Annealing (SA). This "optimizing the optimizer" approach reveals that, the history of actions in metaheuristics provides valuable insights for predicting and enhancing heuristic selections. Our preliminary findings suggest that machine learning models, using historical data, offer a pathway to more intelligent and data-driven optimization strategies in complex scenarios, marking a significant advancement in the field of combinatorial optimization.
Central European Journal of Operations Research, 2024
This work introduces a multi-period, multi-commodity, inventory-routing problem with strategic fl... more This work introduces a multi-period, multi-commodity, inventory-routing problem with strategic fleet scheduling decisions, under the consideration of speed limits, as well as strict European Union regulations on truck drivers' working and driving time. To address the new problem, a mixed integer linear programming model was developed. Several artificial but realistic problem instances were randomly generated following relative guidelines from the open literature, to validate and assess the performance of the novel mathematical model. Furthermore, in an effort to produce useful managerial insights, several sensitivity analyses were performed considering different fluctuation rates on key model parameters.
Computers & Operations Research, 2024
This study presents the Pollution Traveling Salesman Problem with Refueling, a novel optimization... more This study presents the Pollution Traveling Salesman Problem with Refueling, a novel optimization problem which integrates two recently proposed variants of the Traveling Salesman Problem: the Pollution Traveling Salesman Problem and the Traveling Salesman Problem with Refueling. The proposed problem captures the operational dynamics of a real-world routing scenario involving a single vehicle originating from a central depot and delivering products to end customers. When considering the vehicle’s fuel tank capacity and fuel consumption during the routing process, the need to visit fuel stations for refueling arises. To address this complex problem, a new mixed integer linear programming model was developed, and the Gurobi solver was employed to solve smaller instances. For the effective resolution of larger practical problem cases, a two-stage double adaptive general variable neighborhood search method was proposed. The proposed methodology exhibits comparable efficiency to a commercial solver, demonstrating notably low execution time requirements. To further assess its performance, a comparative study was conducted on TSPLib instances. In comparison to various solution approaches documented in the open literature, encompassing both VNS-based and alternative methods, our proposed approach consistently yields highly competitive results within low execution times.
This paper studies the Hierarchical Multi-Switch Multi-Echelon VRP (HMSME-VRP), a newly introduce... more This paper studies the Hierarchical Multi-Switch Multi-Echelon VRP (HMSME-VRP), a newly introduced VRP variant based on a real-world case involving High Capacity Vehicles (HCV). The problem originates from the policies of a distribution company in the Nordic countries where HCVs of up to 34.5 meters and up to 76 tons are allowed. The HMSME-VRP offer a new way to model distribution problems to cover large geographical areas without substantial costs in infrastructure. Furthermore, it adds complexity to the standard VRP and, as such, remains N P-hard and difficult to solve to optimality. Indeed, it has been demonstrated that only very small instances can be solved to optimality by a commercial solver. Thus, in order to handle instances of real-world size, we propose two General Variable Neighborhood Search (GVNS) procedures, the second of which is adaptive, utilizing an intelligent reordering mechanism. In order to evaluate the proposed procedures, 48 benchmark instances of various sizes and characteristics are generated and made publicly available, comprising of clustered, random, and semi-clustered customers. The computational results show that both GVNS procedures outperform the exact solver. Additionally, the adaptive version outperforms the conventional version based on both average and best solutions. Furthermore, we present a statistical analysis to verify the superiority of the adaptive version.
Optimization Letters, 2023
This special issue contains 15 papers submitted by the participants of the 8th International Conf... more This special issue contains 15 papers submitted by the participants of the 8th International Conference on Variable Neighborhood Search (ICVNS 2021), which was held in Abu Dhabi, U.A.E., online due to COVID-19 restrictions, on March 22–24, 2021.
Computational Management Science, 2023
In the big data era which we have entered, the development of smart scheduler has become a necess... more In the big data era which we have entered, the development of smart scheduler has become a necessity. A Distributed Stream Processing System (DSPS) has the role of assigning processing tasks to the available resources (dynamically or not) and route streaming data between them. Smart and efficient task scheduling can reduce latencies and eliminate network congestions. The most commonly used scheduler is the default Storm scheduler, which has proven to have certain disadvantages, like the inability to handle system changes in a dynamic environment. In such cases, rescheduling is necessary. This paper is an extension of a previous work on dynamic task scheduling. In such a scenario, some type of rescheduling is necessary to have the system working in the most efficient way. In this paper, we extend our previous works Souravlas and Anastasiadou (Appl Sci 10(14):4796, 2020); Souravlas et al. (Appl Sci 11(1):61, 2021) and present a mathematical model that offers better balance and produces fewer communication steps. The scheduler is based on the idea of generating larger sets of communication steps among the system nodes, which we call superclasses. Our experiments have shown that this scheme achieves better balancing and reduces the overall latency.
Operations Research Forum, 2023
This article is dedicated to the memory of Professor Nenad Mladenović. It provides a short biogra... more This article is dedicated to the memory of Professor Nenad Mladenović. It provides a short biography, a description of his main scientific achievements, and also testimonials from scholars all over the world who were collaborators for several years. This peer recognition highlights not only his exceptional academic career but also the integrity of his character.
Handbook of Smart Energy Systems, 2023
Supply chain sustainability refers to the optimal balance between economic, environmental, and so... more Supply chain sustainability refers to the optimal balance between economic, environmental, and societal criteria during decision-making processes. This chapter presents a comprehensive review of recent research contributions on sustainable supply chain optimization. The selected articles have been classified according to several criteria. More specifically, an initial classification has been performed based on the sustainability concerns taken under consideration in each study. Next, a classification has been provided based on decision levels, and finally, the reviewed works have been categorized according to the available solution methods. The chapter is closing with some main insights and suggestions for future research.
Expert Systems with Applications, 2023
Finding the best sequence of local search operators that yields the optimal performance of Variab... more Finding the best sequence of local search operators that yields the optimal performance of Variable Neighborhood Search is an important open research question in the field of metaheuristics. This paper proposes a Reinforcement Learning method to address this question. We introduce a new hyperheuristic scheme, termed Bandit VNS, inspired by the Multi-armed Bandit, a particular type of a single state reinforcement learning problem. In Bandit VNS, we utilize the General Variable Neighborhood Search metaheuristic and enhance it by a hyperheuristic strategy. We examine several variations of the Upper Confidence Bound algorithm to create a reliable strategy for adaptive neighborhood selection. Furthermore, we utilize Adaptive Windowing, a state of the art algorithm to estimate and detect changes in the data stream. Bandit VNS is designed for effective parallelization and encourages cooperation between agents to produce the best solution quality. We demonstrate this concept's advantages in accuracy and speed by extensive experimentation using the Capacitated Vehicle Routing Problem. We compare the novel scheme's performance against the conventional General Variable Neighborhood Search metaheuristic in terms of the CPU time and solution quality. The Bandit VNS method shows excellent results and reaches significantly higher performance metrics when applied to well-known benchmark instances. Our experiments show that, our approach achieves an improvement of more than 25% in solution quality when compared to the General Variable Neighborhood Search method using standard library instances of medium and large size.
Optimization Letters, 2023
We address in this paper the parallel machine scheduling problem with a shared loading server and... more We address in this paper the parallel machine scheduling problem with a shared loading server and a shared unloading server. Each job has to be loaded by the loading server before being processed on one of the available machines and unloaded immediately by the unloading server after its processing. The objective function involves the minimization of the overall completion time, known as the makespan. This important problem raises in flexible manufacturing systems, automated material handling, healthcare, and many other industrial fields, and has been little studied up to now. To date, research on it has focused on the case of two machines. The regular case of this problem is considered. A mixed integer programming formulation based on completion time variables is suggested to solve small-sized instances of the problem. Due to its NP-hardness, we propose two greedy heuristics based on the minimization of the loading, respectively unloading, server waiting time, and an efficient General Variable Neighborhood Search (GVNS) algorithm. In the computational experiments, the proposed methods are compared using 120 new and publicly available instances. It turns out that, the proposed GVNS with an initial solution-finding mechanism based on the unloading server waiting time minimization significantly outperforms the other approaches.
Applied Soft Computing, 2022
This work addresses a novel General Variable Neighborhood Search (GVNS) solution method, which in... more This work addresses a novel General Variable Neighborhood Search (GVNS) solution method, which integrates intelligent adaptive mechanisms to re-order the search operators during the intensification and diversification phases, in an effort to enhance its overall efficiency. To evaluate the performance of the new GVNS scheme, asymmetric and symmetric instances of the classic Traveling Salesman Problem (TSP) from the TSPLib were solved. The obtained results of the Double-Adaptive GVNS were compared with those achieved by two single-adaptive GVNS, which use an adaptive mechanism either for the intensification or the diversification phase and with a conventional GVNS. For a fair comparison, all GVNS schemes were structured using the same local search and shaking operators. Moreover, the novel GVNS algorithm was compared with some recent solution methods for the TSP, found in the open literature. The comparative studies revealed the high efficiency of the novel VNS scheme and underlined the significant impact of intelligent mechanisms on the performance of classic metaheuristic frameworks.
An Empirical Study on Factors Influencing the Effectiveness of Algorithm Visualization
The determination of the factors influencing the effectiveness of algorithm visualization poses a... more The determination of the factors influencing the effectiveness of algorithm visualization poses an interesting research question. In this paper, we present the results of a longitude empirical study regarding this question. The study was based on an evaluation of the Visual LinProg educational tool inside classrooms. Visual LinProg is a web-based educational tool, which solves linear programming problems using animation and visualization techniques. Visual LinProg was developed to be used in linear programming courses to supplement the teaching. Our empirical study is based on questionnaires that include quantitative and qualitative topics. This evaluation first indicates that Visual LinProg facilitates the learning of the Revised Simplex algorithm and second presents more results on factors influencing the understanding of this algorithm by the students/users of the Visual LinProg.
This special issue of Optimization Letters presents selected, peer-reviewed, papers that were acc... more This special issue of Optimization Letters presents selected, peer-reviewed, papers that were accepted for presentation in the two recent International Conferences on Variable Neighborhood Search.
Uploads
Papers by Angelo Sifaleras