Last-mile delivery (LMD) is an important aspect of contemporary logistics that directly affects operational cost, efficiency, and customer satisfaction. In this paper, we provide a review of the optimization techniques of LMD, focusing on Artificial Intelligence (AI) driven decision-making, IoT-supported real-time monitoring, and hybrid delivery networks. The combination of AI and IoT improves predictive analytics, dynamic routing, and fleet management, but scalability and regulatory issues are still major concerns. Hybrid frameworks that integrate drones or Unmanned Aerial Vehicles (UAVs), ground robots, and conventional vehicles reduce energy expenditure and increase delivery range, especially in urban contexts. Furthermore, sustainable logistics approaches, including electric vehicle fleets and shared delivery infrastructures, provide promise for minimizing environmental impact. However, economic viability, legal frameworks, and infrastructure readiness still influence the feasibility of large-scale adoption. This review offers a perspective on the changing patterns in LMD, calling for regulatory evolution, technological advancement, as well as interdisciplinary approaches toward cost-effective, durable, and environmentally friendly logistics systems.

This paper presents a modular and intelligent smart energy metering system that integrates real-time sensing, decentralized communication, and deep learning-based analytics. The proposed architecture captures high-resolution electrical parameters and transmits them to a remote server using eProsima Fast Data Distribution Service (Fast DDS), a broker-less, low-latency middleware based on a publish-subscribe communication model. On the server side, a multi-task neural network is employed to simultaneously classify appliance types and detect operational anomalies using short time-series windows of power data. The system achieves over 93% accuracy in appliance classification and more than 99% in anomaly detection. Communication performance evaluation demonstrates stable throughput and an average latency of 25.44 ms, confirming the system’s suitability for real-time applications. The complete pipeline from data acquisition to semantic inference has been validated on real-world datasets and lab measurements, demonstrating its effectiveness for edge-level monitoring and intelligent energy management in smart grid environments.

The burgeoning demand for power across the globe and the disastrous effect of climate change that results from power generation through depleting unclean fuel sources have necessitated the use of renewable and clean fuels. Hydrogen is a good alternative but only when produced from low‐carbon energy resources. However, hydrogen production via this pathway has a high cost and low system efficiency. Therefore, a comparative cost analysis has been performed for hydrogen production through solar power tower (SPT) and parabolic trough collector (PTC) with proton exchange membrane (PEM) electrolyzer. This is carried out using different cycle configurations comprising a modified steam Rankine cycle (MSRC), supercritical carbon dioxide Brayton cycle (SCBC), and organic Rankine cycle (ORC). Optimization of the solar tower power configurations is also performed for improved system performance. The result of the analysis showed that all SPT’s configurations outperformed the PTC’s configurations except for the configuration comprising the steam Rankine cycle (SRC) and ORC due to how well it performed under the low operating temperature of the PTC system. The levelized cost of hydrogen (LCOH) ranged from 7.78 $/kg for the PTC‐MSRC‐ORC configuration to 17.76 $/kg for the PTC‐MSRC‐SCBC configuration. The parametric analysis of the system also showed that the capacity factor had the greatest influence on system performance. Furthermore, a multiobjective optimization of the system using a brute‐force algorithm for minimizing the LCOH and maximizing efficiency improved the system performance in all configurations.

We propose a Reinforcement Learning (RL) framework for hybrid truck-drone Last-Mile Delivery (LMD), where drones launch from a truck and either rendezvous with it or return to the depot after serving customers, subject to vehicle range limits and customer-defined deadlines. Deliveries completed after the latest allowable service time are treated as deadline violations (lateness) and are penalized in the objective. To provide an exact reference, a Mixed Integer Linear Programming (MILP) model is formulated. It serves both as a benchmark for evaluating the RL approach and as a practical method when the delivery set is small or when minimizing lateness is the primary objective, since the exact MILP solution is guaranteed to be optimal but become computationally expensive to solve as problem size grows due to the combinatorial increase in routing and coordination possibilities. The RL method is introduced as a scalable alternative for larger hybrid truck-drone deployments because it maintains practical computation times through iterative learning and avoids the exponential growth in solving effort associated with exact optimization. The RL framework employs an $\epsilon $ -greedy policy with decay and a reward structure that mirrors the MILP objective. Traffic congestion is incorporated through a scaling matrix that adjusts truck travel times on road segments, while drones remain unaffected due to aerial operation. RL is further benchmarked against Genetic Algorithm (GA) and Particle Swarm Optimiztion (PSO), which are also formulated to align with the same MILP objective to ensure a consistent basis for comparison. Results show that the RL component scales more efficiently with increasing delivery size while maintaining competitive solution quality.