Power grids are increasingly exposed to high-impact disturbances driven by extreme weather, cyber–physical threats, and the growing penetration of converter-based renewable resources. In this context, Machine Learning (ML) has emerged as a key enabler for resilience-oriented monitoring, prediction, control, and restoration. This paper presents a structured review of ML strategies for power-grid resilience applications using a four-phase resilience lens (Prevention and Improvement, Control and Mitigation, Restoration, and Cyber Resilience). The literature is organized through a functional taxonomy that includes fault diagnosis, event prediction, control and stability support, restoration, and cyber resilience. In addition to the qualitative synthesis, a quantitative analysis of a dataset of 13,647 peer-reviewed publications (2015–2026) is conducted to characterize research activity across resilience functions and implementation contexts. This analysis is used to illustrate the increasing adoption of machine learning approaches and to distinguish between simulation-based and real-world applications. The results indicate a methodological shift toward Deep Learning and Reinforcement Learning for complex tasks, while federated and edge-based approaches are gaining attention for privacy preserving and real-time applications. These findings provide a structured view of current research directions and support the growing relevance of machine learning in resilience-oriented power system applications, offering a foundation for the development of intelligent and scalable cyber–physical energy systems.

In the dynamic and complex environment of highly automated vehicles, ensuring driver safety is the most critical task. While automation promises to reduce human error, the driver’s role is shifting to that of a teammate who must remain vigilant and ready to intervene, making it essential to monitor their attention level. However, a significant challenge in this domain is the considerable inter-individual variability in how people physiologically respond to cognitive states, such as distraction. This study addresses this by developing a methodology that first groups drivers into distinct physiology-based clusters before training a predictive model. The study was conducted in a high-fidelity driving simulator, where multimodal data streams, including heart rate variability and electrodermal activity, were collected from 30 participants during conditional-automated driving experiments. Using a time-series k-means clustering algorithm, the researchers successfully partitioned the drivers into clusters based on their physiological and behavioral patterns, which did not correlate with demographic factors. Then, a Long Short-Term Memory model was trained for each cluster, which achieved similar predictive performance compared to a single, generalized model. This finding demonstrates that a personalized, cluster-based approach is feasible for physiology-based driver monitoring, providing a robust and replicable solution for developing accurate and reliable attention estimation systems.

Transportation–electrification pathways are increasingly used as inputs for long-term infrastructure and energy-planning decisions. In Puerto Rico, legislative, regulatory, and planning instruments provide context for transportation electrification and decarbonization through mid-century; however, they do not prescribe a unique temporal pathway for electrification of the reference vehicle fleet. This paper develops a consistency-assessment framework that compares planning-reference electrification envelopes, informed by Puerto Rico’s policy and decarbonization context, with a generalized-logistic trajectory calibrated from historical registered EV-stock observations. Historical registered EV-stock data from $$2014$$ through $$2023$$ are used to calibrate the generalized-logistic model, while front-loaded, linear, and delayed electrification envelopes are constructed around a common reference planning endpoint. The resulting pathways are compared using an adoption-level discrepancy metric, $$\Delta P(t)$$ , and a transition-time discrepancy metric, $$\Delta T_x$$ , which quantify differences in electrification levels and milestone-attainment timing, respectively. The calibrated model reproduces the observed series with $$R^{2}=0.996$$ on cumulative stock and $$R^{2}=0.959$$ on annual increments, and parameter uncertainty is propagated to both metrics by profile likelihood and residual bootstrap. Under the $$100\%$$ reference endpoint, the calibrated trajectory remains below the three primary planning-reference envelopes at the representative planning years evaluated. Relative to the common $$2050$$ planning benchmark, the benchmark-to-model adoption-level discrepancy reaches approximately $$25$$ percentage points ( $$95\%$$ interval $$22$$ to $$30$$ ). Representative electrification milestones are attained approximately $$2$$ to $$19$$ years later than under the corresponding envelopes, with delays of approximately $$11$$ to $$16$$ years at the $$90\%$$ electrification milestone; under the $$100\%$$ reference endpoint, the base-case discrepancies for the three primary envelopes remain positive across the bootstrap replicates evaluated. These results illustrate how differences between planning assumptions and a calibrated diffusion trajectory may affect charging-demand projections, infrastructure-deployment schedules, and long-term power-system planning; they are not deterministic forecasts or assessments of policy compliance. Sensitivity analyses across the reference endpoint, the saturation reference, the power-law envelope shape, and a sigmoid-family robustness check show that the reported discrepancy is governed primarily by endpoint and temporal-shape assumptions, with a break-even endpoint near $$75\%$$ , whereas the saturation reference has a comparatively small effect on normalized adoption-level and transition-time metrics but materially affects absolute plug-in-equivalent charging-energy estimates. The framework supports periodic reassessment as adoption evidence evolves.

Planning fast-charging infrastructure for electric vehicles (EVs) supports sustainable transport electrification. This study identifies when a corridor-wide plan along Puerto Rico Highway 2 (PR-2) becomes infeasible and which modeled constraint limits it first. A mixed-integer linear programming (MILP) model with 229 candidate sites across 27 municipalities, integer 150 kW port decisions, a 3 MW per-site ceiling, corridor coverage, and a shared upstream-capacity screen is solved for a February 2024 baseline and six electrification scenarios under an upper-bound demand assumption. The baseline is feasible with 30 stations, 178 ports, and 26.7 MW, whereas all six future scenarios are infeasible. San Juan and Guaynabo, which hold one-third of the island’s registered EV stock but four candidate sites each, exhaust their candidate-set capacity first, at closed-form thresholds of 1.78% and 3.38% electrification, versus 8.99% in the calibrated-2030 case. This threshold reproduces the MILP shortfall to four decimal places in all six infeasible scenarios. At a screening fraction of 0.20, removing the upstream screen changes neither feasibility nor shortfall. Restoring 2030 feasibility would require at least 21 and 11 represented 3 MW sites, versus four each. Under the modeled assumptions, these findings support examining candidate-site limitations before upstream grid reinforcement in sustainable corridor planning.