The widespread adoption of Electric Vehicles (EVs) presents new challenges for efficient and timely access to Charging Stations (CSs), particularly under constraints of limited availability and variable demand. The current investigation addresses the EV charging station allocation problem, aiming to guide EVs to optimal CSs based on real-time and forecasted system dynamics. An integrated framework that combines load profile forecasting, optimal path planning, and drone-assisted edge computing is proposed to support decision-making. Specifically, a Nonlinear Auto-Regressive with Exogenous inputs (NARX) model is used to predict future load profiles at CSs, enabling proactive management of charging demand. To determine the most accessible stations, Dijkstra’s algorithm for shortest-path computation based on the EV’s current location and the locations of the CSs around is applied. Furthermore, drones with lightweight edge computing algorithms enabled real-time data exchange between CSs and EVs, providing up-to-date information on slot availability and local crowd conditions. For the forecasting component, the NARX model has provided a correlation coefficient of 90% for the CS real data collection. Dijkstra’s algorithm was employed to effectively optimize the routing of EVs to their nearest charging stations by determining optimal shortest paths. The simulation results demonstrate that the proposed approach significantly enhances EV allocation efficiency while reducing both waiting times and travel distances. Further research is needed to address regulatory and logistical challenges associated with drone deployment in real-time applications.
Electric vehicles (EVs) have emerged as a transformative solution in the global shift toward sustainable and environmentally friendly transportation. Unlike conventional internal combustion engine vehicles, EVs rely on electric power stored in batteries, resulting in significantly lower greenhouse gas emissions, reduced dependence on fossil fuels, and quieter operation. In addition, EVs are known for their reduced maintenance requirements. In this perspective, EVs can significantly mitigate climate change impacts, especially if powered by renewable energy sources such as solar panels. This advantage, combined with technological advancements, supportive government policies, and increased environmental awareness, has driven widespread EV adoption. However, the global transition to EVs faces challenges, particularly in developing charging infrastructure and managing charging operations.
The rapid development and the growing adoption of EVs have posed new concerns for the almost stakeholders including the EVs’ owners, the charging infrastructure managers, the aggregators, the electric grid operators, and the relevant decision-makers. As EVs become increasingly favorite car user option, their success depends not only on manufacturing technology but also on the effectiveness and accessibility to their supporting logistical infrastructure, including charging stations (CSs). EVs charging infrastructure plays a critical role in enabling widespread use, yet it faces several pressing challenges. Firstly, the spatial distribution of CSs is often uneven, with urban areas better served than rural or suburban regions. Secondly, the current infrastructure struggles to keep pace with the rapidly growing EV fleet, leading to congestion and long waiting times at CSs. Additionally, many CSs lack efficient energy management systems, resulting in inefficient utilization and potential overloading of local power grids. Technical issues such as varying charging standards, insufficient real-time information on CSs availability, and the absence of predictive analytics for demand forecasting further complicate the user experience. These limitations underscore the need for intelligent, scalable, and future-proof solutions to ensure that EV charging infrastructure can support the next phase of sustainable mobility.
The problem of EVs allocation to charging stations (CSs) is inherently multi-dimensional, involving temporal, spatial, and computational complexities. Indeed, EVs must be routed to suitable CSs based on geographical proximity, the dynamic availability of charging slots and the forecasting of energy demand at each station. A simplistic shortest-path strategy may lead to congestion or excessive wait times if the nearest station is already overloaded. Key challenges include: (1) Real-time availability uncertainty since charging station occupancy can change rapidly, requiring real-time information sharing; (2) Forecasting energy load. Indeed, accurate load forecasting is crucial, as unplanned EV arrivals during peak demand periods can result in significant charging delays; (3) Path optimization under uncertainty for which deciding on a charging station based purely on distance ignores critical factors like expected waiting time or crowd density; and (4) Scalability of data processing: centralized systems face significant performance limitations when processing large-scale data in real time. To address these challenges, the present framework integrates load profile forecasting using NARX models, optimal path planning with Dijkstra’s algorithm, and drone-based edge computing for real-time, decentralized decision support. This holistic approach aims to provide EVs with actionable information to make informed charging decisions in dynamic environments with space-time dependencies.
While the proposed framework incorporates drone-based edge computing for real-time EV-CS communication, this component remains theoretical and was not fully implemented in the current evaluation. Key barriers to full implementation include regulatory and logistical challenges, specifically airspace authorization requirements, public safety policies, and operational constraints associated with UAV deployment in urban environments. This study’s scope emphasizes solutions compatible with current infrastructure limitations.
As EV adoption rises, new challenges emerge, particularly around establishing efficient, feasible/optimal, and intelligent allocation of EVs to CSs. Therefore, EVs’ owners should efficiently participate in any effort aiming to improve the charging ecosystem and contribute to the optimal scheduling. Optimal allocation of CSs to moving EVs has become a cornerstone of EV infrastructure planning and management. This ensures users’ needs are met while maximizing network efficiency and reliability. For instance, a recent survey paper investigated the problems related to control and forecasting aiming to improve the charging operations based on emerging technologies such as data-driven approaches and machine learning (ML). In the context of a deregulated energy market, the price of electricity is also important and impactful. To reduce the cost, save time, and increase the life-time of the EV battery, efficiently forecasting the price of charging is also critical for all EVs charging ecosystem components. To meet the rising charging needs, effective EV to CS allocation requires a multifaceted approach involving load profile forecasting at the CS level, optimal path planning for EVs in low State of Charge (SoC), and real-time computational support at the various devices/equipment and enablers level. Traditional approaches for station allocation have primarily relied on static parameters, such as population density, proximity to main highways, and estimates of average daily usage. However, the rapid evolution of data analytics, machine learning, and the Internet of Things (IoT) has created opportunities to enhance these methods by using predictive models, real-time data analysis, and path optimization. In fact, as opposed to the classical charging problem where vehicles are to be charged during their parking time (usually during night hours), dynamic EV charging is requested in real-time when EVs are on the road while facing a shortage of charge. Under such circumstances, EVs may need assistance from other parties playing the role of enablers. The EV drivers should therefore be alerted about their vehicle battery SoCs in real-time. Once the SoC reaches a minimum threshold, an optimal (not necessarily the nearest) CS as well as the optimal path he should follow to reach the allocated CS should be assigned. By leveraging data on electricity demand at the CSs level, road patterns, and vehicle energy consumption, advanced forecasting models and path planning algorithms can assist in determining future demand and establish optimal station allocation that aligns with both current EV SoC and its short-term need for charging. Additionally, EVs’ relatively short driving range and dependence on accessible CSs necessitate an approach that accounts for route planning and energy consumption patterns. This may enable drivers to access charging facilities with minimal deviation from their planned paths. To ensure efficient charging operations, several attributes related to load and energy storage should be considered carefully. The EV charging behavior and load have randomness and uncertainty, making them affected by many factors. Road network structure (rectilinear and curvilinear paths), traffic congestion, CSs distribution, driving path, travel destination, initial SoC and even the driver(s) psychology and level of anxiety are the most relevant attributes.
In the current landscape, integrating emerging technologies such as drones, Internet-of-Things (IoT), servers and Mobile Edge Computing (MEC) technologies are important to process and analyse data related to EVs charging locally and globally. Meanwhile, urban and high-traffic areas exhibit data latency that could be detrimental for the efficient operation. Therefore, integrating those contributors may help in improving the EVs ecosystem. Via Edge computing devices, computational power is brought closer to EVs and CSs, providing real-time processing capabilities that reduce the need for centralized data centres and minimize data transfer delays. Drone-enabled edge computing can support a wide range of functionalities in EV charging infrastructure, from monitoring CSs availability and energy demand to providing real-time updates on route crowds by accurately surveying the urban area traffic around the CSs. By deploying a fleet of drones, it becomes possible to dynamically manage charging station resources and optimize power distribution based on real-time data. Additionally, using drones with high-resolution cameras and lightweight computer vision algorithms may assist the CS allocation by efficiently surveying traffic congestion, mainly around CSs. To contribute to the effort of improving the EVs’ charging operations, the current paper introduces a comprehensive framework to efficiently assign any EV needing charge to the “best” charging station by involving new technologies and algorithms. The proposed solution integrates path planning optimization at the vehicle level with energy management strategies at the CS level.
