This paper addresses the critical challenge of optimal allocation and scheduling of Internet of Things (IoT) tracking devices for patient monitoring in healthcare facilities, where limited resources must be efficiently distributed to minimize cost and maximize care quality. We formulate this healthcare management problem as a specialized variant of the Resource-Constrained Scheduling Problem that incorporates patient-specific factors such as duration of stay and priority. After establishing the computational complexity of the problem, we propose a Non-dominated Sorting Genetic Algorithm II (NSGA-II) to address the complex problem of balancing multiple objectives: cost minimization and quality of care maximization. Our approach offers a set of optimal trade-offs, enabling informed decision-making to select the best final solution. Computational experiments conducted on both simulated scenarios and real-world healthcare facility datasets demonstrate that our approach outperforms existing methods, achieving between 1.2 and 3.0 times more solutions than the state of the art. Moreover, in comparison to manual scheduling by medical center managers, our method achieves cost savings of up to 12% (with an average of 6.3%) and quality improvements of up to 20% (with an average of 10%) across the tested experiments. The proposed method scales effectively to realistic healthcare settings with varying numbers of patients and tracking devices, maintaining solution quality while keeping computational time within practical limits for daily operational use. Our findings contribute to both healthcare operations research and clinical practice by providing an efficient methodology for optimizing the use of limited monitoring resources while prioritizing patient safety.

This study presents an optimisation model for scheduling homebound vaccination in a more efficient way to address the existing workforce management challenge. We consider a home healthcare routeing challenge for people to be vaccinated at home based on limited resources. There are different types of patients that are categorised based on the services they require and should be served by appropriate workforce teams or a single medical staff, where teams are transported by rental vehicles. In this context, our goal is to minimise the total cost of transportation while considering patient requirements and workforce qualifications, as well as resource constraints and the time limit within which the vaccine must be administered. To pursue this goal, a mathematical formulation, based on the vehicle routeing dynamics is proposed, along with an algorithm to address the challenge. A case study with a Physician who administers vaccinations at home in southeastern Italy is analysed. Driving and working times are subject to uncertainty and are defined by empirical data. Our approach allows the physician to identify the most promising solutions and thus the best one in terms of reducing work time and risk. The resulting schedule maximises the vaccine delivery rate.

The proliferation of Internet of Things (IoT) devices demands efficient resource management in fifth-generation (5G) networks, particularly through network slicing mechanisms supporting massive machine-type communications (mMTCs). This paper addresses IoT connectivity in 5G network slicing through a bi-objective optimization framework balancing operational costs with quality-of-service. We formulate a bi-objective optimization problem that balances operational costs with quality-of-service (QoS) requirements across heterogeneous 5G network slices. The proposed approach employs a tailored Non-dominated Sorting Genetic Algorithm II (NSGA-II) incorporating domain-specific constraints, including device priorities, slicing isolation requirements, radio resource limitations, and battery capacity. Through extensive simulations on scenarios with up to 5000 devices, our method generates diverse Pareto-optimal solutions achieving hypervolume improvements of 8–13% over multi-objective DRL, 15–28% over single-objective DRL baselines, and 22–41% over heuristic approaches while maintaining computational scalability suitable for real-time network management (sub-2 min execution). Validation with real-world traffic traces from operational deployments confirms algorithm robustness under realistic burstiness and temporal patterns, with 7% performance degradation vs. synthetic traffic—within expected simulation–reality gaps. This work provides a practical framework for IoT resource scheduling in current 5G and future Beyond-5G (B5G) telecommunications infrastructures, validated in scenarios of up to 5000 devices.

We study the predictive maintenance scheduling for IoT-enabled medical equipment in multi-facility healthcare networks. The problem involves skill matching, time windows, and risk-aware priorities. We model a multi-skill Technician Routing and Scheduling Problem with IoT-predicted failure intervals and minimize a composite cost for technician activation and labor, travel/time, risk exposure within the failure window, and lateness beyond it. We propose a hybrid solver coupling a Genetic Algorithm (GA) for rapid exploration and feasible schedule generation with a Proximal Policy Optimization (PPO) agent warm-started via behavior cloning on GA elites and refined online in a receding-horizon manner. An optional, permissioned blockchain records tamper-evident maintenance events off the control loop for auditability. Across four case studies (10–30 facilities), the hybrid approach reduces total cost by 2.09–10.31% versus pure GA, by 0.57–2.65% versus pure Deep Reinforcement Learning (DRL), and by 0.93–2.86% versus OR-Tools VRP heuristic baseline. In controlled early-stopping runs guided by admissible GA/DRL time splits, we realized average wall-time savings up to 47.5% while keeping solution costs within 0.5% of full-budget runs and maintaining low or zero lateness and risk exposure. These results indicate that GA seeding improves sample efficiency and stability for DRL in complex, data-driven maintenance settings, yielding a practical, adaptive, and auditable scheduler for healthcare operations.

Background: This work studies the single vehicle routing problem (VRP) with multishift and fuzzy uncertainty. In this case, a company perpetually exploits a vehicle to accomplish demand over a scheduling period of several work shifts. In our problem, a crew performs maintenance jobs at different locations. The working team operates in different shifts with a maximum duration but recurrently returns to the depot by the end of the shift to avoid overtime. Methods: The objective is to minimize the number of shifts and the completion time (makespan). In addition, we analyze the influence of uncertainty in driving and processing times on the overtime avoidance constraint in shift duration. We develop an Artificial Immune Heuristic to determine optimal solutions considering both makespan and overtime avoidance. We implement a Pareto-based framework to evaluate the impact of uncertainty. Results: We present several numerical case studies to examine the problem. In particular, we analyze different case study scenarios inferred from the environmental changes in travel and processing times observed in the Apulia region (SE Italy) during the COVID-19 lockdown periods that occurred in spring (started on March 9, 2020) and autumn (after November 6, 2020) of the year 2020. Conclusion: The work program was revised as soon as the Italian COVID-19 restrictions were implemented in the spring and autumn of 2020 due to the changing environment. Our approach allowed for the rapid release of new robust maintenance programs. Results show significant improvements with the presented approach.