Over time, the focus on supportive and geriatric care has shifted from being predominantly provided in institutional settings like nursing or rest homes to be delivered within the homes of the patients. Trained caregivers now provide home healthcare services by visiting patients in their own homes and carrying out specific services based on each patient's individual needs before moving on to the next patient. Planning such a service involves considering the routing aspect and ensuring synchronization between services and designated time windows for patients. To solve the problem, we propose a local search approach that combines different neighborhood operators guided by the simulated annealing metaheuristic. Additionally, we introduce a realistic and diverse dataset and a robust and flexible file format based on JSON. This dataset and format have the potential to facilitate future comparisons and analyses. Our study shows that by appropriately tuning our algorithm in a statistically rigorous manner, it outperforms existing methods on all benchmarks.

Curriculum-Based Course Timetabling (CB-CTT) is a well-established benchmark problem in educational timetabling, with numerous real and synthetic datasets. Despite its long history of algorithmic developments, designing search methods that work effectively across diverse datasets under time limits remains challenging.In this paper, we focus on local search and, after a thorough analysis of all neighborhood structures reported in the literature, we propose an improved Simulated Annealing (SA) algorithm driven by an extended version of the classic LectureMove neighborhood, called LectureKick, that allows flexible lecture relocation while preserving feasibility. The neighborhood is further enriched with several sampling biases specifically designed to address the soft constraints of CB-CTT, while the overall behavior of the SA is regulated by a cut-off mechanism embedded in the cooling schedule.The method involves numerous parameters, both for the SA schedule and for the neighborhood biases, which are rigorously tuned on a large artificial dataset. The final algorithm is subsequently evaluated on the available benchmarks, including the “standard” ITC 2007 comp dataset, under the original competition timeout conditions.The results show that our approach clearly outperforms all previously proposed local search methods and is competitive with the best methods reported for CB-CTT overall, while improving upon them on several instances.

Home Healthcare is an essential component of healthcare systems, where caregivers visit patients' homes to deliver services. While presenting advantages with respect to institutional care, such as being cost-effective and alleviating family burdens, it presents challenges in scheduling and routing caregivers efficiently. While various formulations of such a problem-the home healthcare routing and scheduling problem-have been proposed, many fail to include key real-world features, limiting their applicability in practice. This paper consolidates relevant features from existing models and significantly extends the formulation by Mankowska et al. (2014) to develop a unified framework incorporating as many real-world aspects as possible. We introduce a new instance generator, a set of real-world features to describe these instances, and a comprehensive infrastructure for validating and comparing solutions. Furthermore, we extend two state-of-the-art solution methods-simulated annealing and one based on a compact mixed integer linear programming model solved by a state-of-the-art software - to integrate all these features and provide a comparative analysis of their performance. Results show that our general solution methods are able to outperform methods specialized for a specific formulation on many instances.

We consider a recently proposed integrated healthcare problem that deals with the assignment of patients to suitable rooms in wards (Patient-to-Room) and the assignment of nurses to patients to balance their workload and reduce patient disturbance (Nurse-to-Patient), in one single stage. For this problem, we designed an integrated local search method that uses four distinct neighborhoods and is guided by a Simulated Annealing metaheuristic. We configured and tuned our search method, ran it on the available datasets with different running times, and validated it using the available solution checker. Finally, we compared the results with the ones available in the literature and with a sequential version of our method. The outcome is that our search method outperforms the previous ones, in particular for large and real-world instances. Finally, we observe that the integrated approach is particularly effective when the planning horizon is short.

We consider a complex real-world problem in slab yard management that deals with block relocation and slab stack shuffling in a single problem. For this integrated problem, we designed a two-level metaheuristic approach with a local search procedure that works at the level of the slab selections and uses an inner greedy heuristic for deciding relocations and retrievals. We tested our solution approach on a novel dataset using an actual yard layout and a simulation tool to set its initial configuration. Our search method successfully obtained high-quality solutions, even for large instances involving over 2000 slabs. To assess the performance of our search method compared to state-of-the-art ones, we adapted it to a simpler formulation from recent literature and tested it on public benchmarks. Our method proved as good as the best existing ones in most available instances and consistently outperformed them on the largest ones.

In a dynamic global economic landscape, logistics companies have to be able to respond quickly and flexibly to changes in demand. This is where the concept of On-Demand Warehousing (ODW) comes in; an emerging approach that promises to revolutionize the way companies manage their warehouse space. This approach allows companies with temporary excess capacity to offer their space to others, who want to cover short-term demand peaks. By this, this concept provides advantages over traditional models, such as dedicated storage facilities or long-term leasing. However, the dynamic nature of this system presents unique challenges, especially in terms of matching customer requests with available storage in real time. Unlike offline models, where future demands are known or estimated, the Online ODWP requires decisions to be made without prior knowledge of upcoming requests. Our work addresses online ODWP by proposing an innovative methodology that integrates Machine Learning methods with sequential stochastic optimization to enhance decision making processes in real time. In an extensive computational study, we show that the newly proposed approach outperforms state-of-the-art heuristics and yields near optimal solutions within very short run times. Detailed algorithmic analyses as well as managerial insights are derived. We, for instance, provide decision guidelines for platform providers facing acceptance or rejection decisions on dynamically arriving storage requests.

Real-world timetabling problems in healthcare usually encompass several interdependent optimization problems. Despite their practical importance and the existence of algorithms for solving subproblems that compose these ‘integrated’ problems, there are no methodologies or guidelines for addressing them adequately.In the interest of stimulating research on these relevant integrated optimization problems, the authors prepared a computational challenge focusing on the Integrated Healthcare Timetabling Problem. In 2024, they launched an international competition associated with this original problem: the Integrated Healthcare Timetabling Competition.The scientific initiative can be called a success with respect to several criteria. The competition attracted a large number of participants from both industry and academia, leading to 32 teams submitting their computational results. Interestingly, the participants developed a diverse set of complementary algorithmic approaches, with several smart ideas to restrict the huge solution space or speed up the computations.This paper introduces the Integrated Healthcare Timetabling Competition’s statement, along with its regulations and datasets. It additionally reports the progress, participants’ submissions and results.All generated knowledge remains available to boost future scientific investigation, algorithm development and applications.

Home healthcare has become more and more central in the last decades, due to the advantages it can bring to both healthcare institutions and patients. Planning activities in this context, however, presents significant challenges related to route planning and mutual synchronization of caregivers.In this paper we propose a new compact model for the combined optimization of scheduling (of the activities) and routing (of the caregivers) characterized by fewer variables and constraints when compared with the models previously available in the literature. The new model is solved by a constraint programming solver and compared experimentally with the exact and metaheuristic approaches available in the literature on the common datasets adopted by the community. The results show that the new model provides improved lower bounds for the vast majority of the instances, while producing at the same time high quality heuristic solutions, comparable to those of tailored metaheuristics, for small/medium size instances.

In On-Demand Warehousing, an online platform acts as a central mechanism to match unused storage space and related services offered by suppliers to customers. Storage requests can be for small capacities and very short commitment periods if compared to traditional leasing models. The objective of the On-Demand Warehousing Problem (ODWP) is to maximize the number of successful transactions among the collected offers and requests, considering the satisfaction of both the supply and demand side to preserve future participation to the platform. The Online ODWP can be modeled as a stochastic reservation and assignment problem, where dynamically arriving requests of customers must be rapidly assigned to suppliers. Firstly, an online stochastic combinatorial optimization framework is adapted to the Online ODWP. The key idea of this approach is to generate samples of future requests by evaluating possible allocations for the current request against these samples. In addition, expectation, consensus, and regret, and two greedy algorithms are implemented. All solution methods are compared on a dataset of realistic instances of different sizes and features, demonstrating their effectiveness compared to the oracle solutions, which are based on the assumption of perfect information about future request arrivals. A newly proposed approach of risk approximation is shown to outperform alternative algorithms on large instances. Managerial insights regarding acceptance and rejection strategies for the platform are derived. It is shown how requests with large demand, long time frame, not very long spanning time, and average compatibility degree, are very likely to be rejected in the optimal solution.

EasyLocal++ is a white-box C++ framework for designing local search algorithms. Over the years, it has been successfully used across various domains, such as timetabling, rostering, scheduling, and logistics, and has produced state-of-the-art results in benchmark datasets and competitions. Beyond research, EasyLocal++ has found practical use in real-world and industrial settings, demonstrating the flexibility and adaptability of the framework for different applications. In this paper, we position EasyLocal++ within the existing literature by comparing its capabilities with those of available alternative/similar tools. We then trace its history from its initial design 25 years ago to the current version. Furthermore, we describe its architecture, highlighting its design principles and functionalities. We also discuss the features developed to simplify the design of local search methods and enhance their performance. Lastly, we explore potential future perspectives and developments.

EasyLocal++ is a white-box C++ framework for designing local search algorithms. Over the years, it has been successfully used across various domains, such as timetabling, rostering, scheduling, and logistics, and has produced state-of-the-art results in benchmark datasets and competitions. Beyond research, EasyLocal++ has found practical use in real-world and industrial settings, demonstrating the flexibility and adaptability of the framework for different applications. In this paper, we position EasyLocal++ within the existing literature by comparing its capabilities with those of available alternative/similar tools. We then trace its history from its initial design 25 years ago to the current version. Furthermore, we describe its architecture, highlighting its design principles and functionalities. We also discuss the features developed to simplify the design of local search methods and enhance their performance. Lastly, we explore potential future perspectives and developments.

This study investigates the application of reinforcement learning for the adaptive tuning of neighborhood probabilities in stochastic multi-neighborhood search. The aim is to provide a more flexible and robust tuning method for heterogeneous scenarios than traditional offline tuning. We propose a novel mix of learning components for multi-neighborhood Simulated Annealing, which considers both cost- and time-effectiveness of moves. To assess the performance of our approach we employ two real-world case studies in timetabling, namely examination timetabling and sports timetabling, for which multi-neighborhood Simulated Annealing has already obtained remarkable results using offline tuning techniques. Experimental data show that our approach obtains better results than the analogous algorithm that uses state-of-the-art offline tuning on benchmarking datasets while requiring less tuning effort.

We consider the Capacitated Facility Location Problem with Customer Incompatibilities, which is a recently -proposed variant of the classic facility location problem whose distinctive feature is to take into account incompatibilities between customers. We tackle this problem using local search and we propose a combination of neighborhoods and ad hoc techniques to reduce the size of the search space, in order to effectively deal with large instances. The resulting multi-neighborhood approach is guided by a simulated annealing procedure. Our method, suitably tuned in a statistically-principled way, has been able to outperform all previous techniques on the publicly available dataset, on both short and long running times.

We consider the medical student scheduling (MSS) problem, which consists of assigning medical students to internships of different disciplines in various hospitals during the academic year to fulfill their educational and clinical training. The MSS problem takes into account, among other constraints and objectives, precedences between disciplines, student preferences, waiting periods, and hospital changes. We developed a local search technique, based on a combination of two different neighborhood relations and guided by a simulated annealing procedure. Our search method has been able to find the optimal solution for all instances of the dataset proposed by Akbarzadeh and Maenhout (Comput Oper Res 129: 105209, 2021b), in a much shorter runtime than their technique. In addition, we propose a novel dataset in order to test our technique on a more challenging ground. For this new dataset, which is publicly available along with our source code for inspection and future comparisons, we report the experimental results and a sensitivity analysis.

We consider a recently-proposed integrated healthcare problem that deals with the assignment of patients to suitable rooms in wards (Patient-to-Room) and the assignment of nurses to patients to balance their workload (Nurse-to-Patient), in one single stage. For this problem, we designed a local search approach that uses the union of two distinct neighborhoods and is guided by a Simulated Annealing metaheuristic. We tuned our search method, ran it on the available dataset, and validated it using the available solution checker. Finally, we report our results for different running times, to show how the scores evolve based on the granted time.

We consider the home healthcare scheduling and routing problem, and we extend the classic formulation introduced by Mankowska et al, by adding several real-world features. For this novel problem, we created a new realistic dataset, and we developed a metaheuristic approach based on a combination of neighborhoods guided by a Simulated Annealing procedure. Our solver, properly engineered and tuned, is able to solve all instances in a short time. Our experimental results highlight the relative importance of the various (original and new) cost components.

We address the Permutation Flow-Shop Scheduling Problem with Predictive Maintenance presented by Varnier and Zerhouni (2012), that consists in finding the integrated schedule for production and maintenance tasks such that the total production time and the advance of maintenance services are minimized. Predictive maintenance services are scheduled based on a prognostics system that is able to provide the remaining useful life of a machine. To solve this problem, we propose a local search method with neighborhoods specifically tailored for maintenance interventions. Computational experiments performed on generated benchmarks demonstrate the effectiveness and scalability of our method with respect to an exact technique based on the mathematical model proposed by Varnier and Zerhouni (2012).

Home healthcare services are carried out by trained caregivers who visit the patient’s home, perform their service operations that depend on the patient’s need (e.g., medical care or just instrumental activities of daily living), and then move to the next patient. We consider the home healthcare scheduling and routing problem, in the formulation proposed by Mankowska et al. (2014), which includes synchronization among services and time windows for patients. For this problem, we propose a local search approach based on a novel neighborhood operator and guided by the Simulated Annealing metaheuristic. We show that our approach, properly tuned in a statistically-principled way, is able to outperform state-of-the-art methods on most of the original instances made available by Mankowska et al.

We propose a survey of the research contributions on the field of Educational Timetabling with a specific focus on "standard" formulations and the corresponding benchmark instances. We identify six of such formulations and we discuss their features, pointing out their relevance and usability. Other available formulations and datasets are also reviewed and briefly discussed. Subsequently, we report the main state-of-the-art results on the selected benchmarks, in terms of solution quality (upper and lower bounds), search techniques, running times, statistical distributions, and other side settings.

We consider a mobile network in which traditional static terrestrial base stations are not capable of completely serving the existing user demand, due to the huge number of connected devices. In this setting, an equipped Unmanned Aerial Vehicle (UAV) can be employed to provide network connection where needed in a flexible way, thereby acting as an unmanned aerial base station. The goal is to determine the best UAV trajectory in order to serve as many users as possible. The UAV can move at different speeds and can serve users within its communication range, although the data rate depends on the positions of UAV and users. In addition, each user has a demand (e.g., the number of bits the user wants to download/upload from/to the network) and a time window during which requires the service. We propose a Biased Random-Key Genetic Algorithm (BRKGA) and a Simulated Annealing Algorithm (SAA), and compare them on realistic instances with more than 500 users in different settings.