Following decades of sustained improvement, metaheuristics are one of the great success stories of optimization research. However, in order for research in metaheuristics to avoid fragmentation and a lack of reproducibility, there is a pressing need for stronger scientific and computational infrastructure to support the development, analysis and comparison of new approaches. To this end, we present the vision and progress of the Metaheuristics In the Large project. The conceptual underpinnings of the project are: truly extensible algorithm templates that support reuse without modification, white box problem descriptions that provide generic support for the injection of domain specific knowledge, and remotely accessible frameworks, components and problems that will enhance reproducibility and accelerate the field’s progress. We argue that, via such principled choice of infrastructure support, the field can pursue a higher level of scientific enquiry. We describe our vision and report on progress, showing how the adoption of common protocols for all metaheuristics can help liberate the potential of the field, easing the exploration of the design space of metaheuristics.
Jerry Swan, Steven Adrænsen, Colin G. Johnson, Ahmed Kheiri, Faustyna Krawiec, J.J. Merelo, Leandro L. Minku, Ender Özcan, Gisele L. Pappa, Pablo García-Sánchez, Kenneth Sörensen, Stefan Voß, Markus Wagner, David R. White, Metaheuristics “In the Large”, European Journal of Operational Research, 2021,ISSN 0377-2217, DOI: 10.1016/j.ejor.2021.05.042
Crowdsourced logistics: The pickup and delivery problem with transshipments and occasional drivers
This article considers a setting in which a courier, express, and parcel service provider operates a fleet of vehicles with regular drivers (RDs) to ship parcels from pickup to delivery points. Additionally, the company uses a platform where occasional drivers (ODs) offer their willingness to take on requests that are on or near the route they had originally planned. There exist transshipment points (TPs) to better integrate these ODs. ODs or RDs may transfer load at these predetermined TPs. The problem is modeled as a mixed-integer programming model and called pickup and delivery problem with transshipments and occasional drivers (PDPTOD). We develop a solution approach based on an adaptive large neighborhood search. The article provides insights on how the number and location of TPs impact the cost advantages achieved by integrating ODs. It also shows that the cost savings are highly sensitive to the assumed flexibility and compensation scheme.
Voigt, S, Kuhn, H. Crowdsourced logistics: The pickup and delivery problem with transshipments and occasional drivers Networks 2021; 1– 24. DOI: 10.1002/net.22045
An exact approach to the restricted block relocation problem based on a new integer programming formulation
This study addresses the block(s) relocation problem (BRP), also known as the container relocation problem. This problem considers individually retrieving blocks piled up in tiers according to a predetermined order. When the block to be retrieved next is not at the top, we have to relocate the blocks above it because we can access only the topmost blocks. The objective is to retrieve all the blocks with the smallest number of relocations. In this study, a novel exact algorithm is proposed for the restricted BRP, a class of the problem where relocatable blocks are restricted. The proposed algorithm computes lower and upper bounds iteratively by solving the corresponding integer programming problems until the optimality gap is reduced to zero. The novelty of the algorithm lies in the formulations based on complete and truncated relocation sequences of the individual blocks. We examine the effectiveness of the proposed algorithm through computational experiments for benchmark instances from the literature. In particular, we report that, for the first time, all the instances with up to 100 blocks are solved to proven optimality.
Shunji Tanaka, Stefan Voß, An exact approach to the restricted block relocation problem based on a new integer programming formulation, European Journal of Operational Research, 2021, ISSN 0377-2217, DOI: 10.1016/j.ejor.2021.03.062
Maximization of Open Hospital Capacity under Shortage of SARS-CoV-2 Vaccines—An Open Access, Stochastic Simulation Tool
The Covid-19 pandemic has led to the novel situation that hospitals must prioritize staff for a vaccine rollout while there is acute shortage of the vaccine. In spite of the availability of guidelines from state agencies, there is partial confusion about what an optimal rollout plan is. This study investigates effects in a hospital model under different rollout schemes. Methods. A simulation model is implemented in VBA, and is studied for parameter variation in a predefined hospital setting. The implemented code is available as open access supplement. A rollout scheme assigning vaccine doses to staff primarily by staff’s pathogen exposure maximizes the predicted open hospital capacity when compared to a rollout based on a purely hierarchical prioritization. The effect increases under resource scarcity and greater disease activity. Nursing staff benefits most from an exposure focused rollout. The model employs SARS-CoV-2 parameters; nonetheless, effects observable in the model are transferable to other infectious diseases. Necessary future prioritization plans need to consider pathogen characteristics and social factors.
Bosbach, Wolfram A.; Heinrich, Martin; Kolisch, Rainer; Heiss, Christian. 2021. Maximization of Open Hospital Capacity under Shortage of SARS-CoV-2 Vaccines—An Open Access, Stochastic Simulation Tool, Vaccines 9, no. 6: 546, DOI: 10.3390/vaccines9060546
Fixed set search application for minimizing the makespan on unrelated parallel machines with sequence-dependent setup times
This paper addresses the problem of minimizing the makespan on unrelated parallel machines with sequence-dependent setup times. The term unrelated machines is used in the sense that there is no correlation between the processing times for jobs between different machines. Due to the NP-hardness of this problem, a wide range of metaheuristics has been developed to find near-optimal solutions. Out of such methods, the ones based on constructive greedy algorithms like the Greedy Randomized Adaptive Search Procedure (GRASP), Ant Colony Optimization (ACO) and Worm Optimization (WO) proved to be most efficient. The Fixed Set Search (FSS) is a novel population-based metaheuristic of this type that adds a learning mechanism to the GRASP. The basic concept of FSS is to avoid focusing on specific high quality solutions but on parts or elements that such solutions have. This is done through fixing a set of elements that exist in such solutions and dedicating computational effort to finding near-optimal solutions for the underlying subproblem. In this work, the FSS is applied to the problem of interest. Computational experiments show that the FSS manages to significantly outperform the GRASP, ACO and WO on the standard benchmark instances when the quality of found solutions is considered without an increase in computational cost. This application of the FSS is significant as it shows that it can also be applied to scheduling type problems, in addition to covering and routing ones.
Raka Jovanovic, Stefan Voß, Fixed set search application for minimizing the makespan on unrelated parallel machines with sequence-dependent setup times, Applied Soft Computing,2021,107521,ISSN 1568-4946, DOI: 10.1016/j.asoc.2021.107521