In this paper, we introduce, model, and solve a clustered resource allocation and routing problem for humanitarian aid distribution in the event of an earthquake and subsequent tsunami. First, for the preparedness stage, we build a set of clusters to identify, classify, sort, focus, and prioritize the aid distribution. The clusters are built with k-means method and a modified version of the capacitated p-median model. Each cluster has a set of beneficiaries and candidate delivery aid points. Second, vehicle routes are strategically determined to visit the clusters for the response stage. A mixed integer linear programming model is presented to determine efficient vehicle routes, minimizing the aid distribution times. A vulnerability index is added to our model to prioritize aid distribution. A case study is solved for the city of Iquique, Chile.

In this paper, we present the Time-Dependent Vehicle Routing Problem with Fleet and Crew Assignment (TD-VRPFCA). This problem involves finding efficient routes for vehicles that deliver goods to customers located within buildings or on street level. It also considers the assignment of different crews to these vehicles. Moreover, the problem assumes Time-Dependent Travel Times (TD-TT) for the road network and Time-Dependent Service Times (TD-ST) for in-building deliveries. To solve this problem, we propose a Mixed Integer Linear Programming (MILP) formulation for solving small-size instances and a Black-Widow Optimization (BWO) metaheuristic for real-world-size instances. Extensive computational experiments on test instances, as well as a real-world case study from a Chilean delivery company, demonstrate the effectiveness of our approach. The findings suggest that our proposed methods can provide efficient delivery routes, times, crews, and fleet assignments, taking into account the complexities of urban logistics.

Background: Early unplanned readmission to critical care units is associated with adverse outcomes and consumes scarce capacity twice, yet the decision to transfer a patient out of the unit is still made without objective criteria. Machine learning may support this decision, but metaheuristic hyperparameter search remains largely unexplored here. Methods: We analyzed 371 patients discharged from the critical care unit of a Chilean public hospital, of whom 41 were readmitted within seven days of transfer to a general ward. Five models (logistic regression, support vector classification, decision tree, random forest, and XGBoost) were trained on clinical and nursing variables available at transfer, using nested stratified five-fold cross-validation with preprocessing fitted within the training folds. Bayesian optimization and a genetic algorithm were compared under identical folds. Results: Under nested cross-validation, discrimination was moderate (AUC 53.3–75.2%), and no single model performed best across all metrics. Logistic regression achieved the highest AUC (75.2%) with the lowest variability across folds, followed by XGBoost (75.0%), whose results varied markedly across folds. The genetic algorithm yielded higher AUC than Bayesian optimization for most tree-based models. Days in the critical care unit, age, and unit of origin ranked highest in impurity-based importance. Conclusions: This proof-of-concept study suggests that routinely collected clinical and nursing information available at transfer may inform early readmission prediction. As the sample was case-enriched and derived from a single center, recalibration to the natural readmission prevalence and prospective external validation are required before clinical implementation.