Autonomous Multirotor UAV Docking and Charging: A Comprehensive Review of Systems, Mechanisms, and Emerging Technologies

Abstract

Multirotor Unmanned Aerial Vehicles (UAVs), characterized by their inherently symmetrical propulsion configurations, are increasingly applied across diverse domains, yet their endurance remains fundamentally constrained by the high energy demand of flight. Autonomous docking and charging systems have emerged as practical solutions, enabling UAVs to recharge or replace batteries without human intervention. This paper provides a structured review of current approaches, offering a systematic categorization of UAV docking platforms into fixed and mobile systems, followed by an analysis of positioning and landing strategies, charging mechanisms, and modular docking concepts. Advances in vision-based guidance and sensor fusion are highlighted as key enablers of precise and reliable autonomous recovery. Contact-based charging and wireless power transfer are compared, with their benefits and limitations outlined. In addition to charging solutions, the paper presents a dedicated review of mechanisms that enable automated battery swapping, increasingly recognized as a complementary pathway to extend mission duration. By synthesizing state-of-the-art research and implementations, this study identifies key technological trends, persisting challenges, and future directions toward scalable, fully autonomous ecosystems capable of long-duration operations.

1. Introduction

The evolution of unmanned aerial vehicle (UAV) technologies has been strongly shaped by advances in sensor miniaturization, computational capabilities, and propulsion systems. Early UAV applications were predominantly realized using fixed-wing platforms, valued for their relative simplicity of control and energy-efficient flight dynamics. Over the past two decades, however, the rapid development of microelectromechanical systems (MEMS) sensors and onboard processing power has significantly improved the ability of UAVs to integrate data from multiple sources to correct errors and improve precision, as shown in [1]. Characterized by their mechanical simplicity, versatility, and the ability to perform precise hovering and maneuvering, multirotors have become the dominant class of UAVs for missions ranging from surveillance and inspection in fields such as the security industry [2] to logistics and precision agriculture. Their inherent instability necessitates advanced control algorithms, but their design flexibility allows for omnidirectional flight modes not achievable with fixed-wing aircraft or helicopters. Despite these advantages, multirotors suffer from high energy consumption, which imposes strict limitations on endurance where novel approaches to bypass the limitations begin to surface, as in [3].

In general, multirotor platforms comprise an even number of rotors arranged in a symmetric geometric configuration. This symmetry is not only a design choice but also a fundamental requirement for achieving equilibrium in flight. The total thrust generated by the propulsion units must counteract gravity, while the generated moments must cancel out around each axis of the aircraft. By symmetrically distributing the thrust and torque, the multirotor platform can maintain a stationary position and orientation. Conventional propulsion systems are based on electric motor units that convert the energy stored in LiPo batteries into the mechanical work of propellers. Electric propulsion configurations consist of symmetrically arranged rotors, typically in pairs with one clockwise (CW) and one counterclockwise (CCW) rotating unit, each fitted with a propeller of matching design to balance opposing torques. Figure 1 illustrates the most common symmetric multirotor configurations and the performance characteristics of a wide range of electric propulsion units.

To address the endurance limitations inherent to multirotor UAVs, multiple approaches have been investigated. At the system level, mission planning, trajectory optimization, and aerodynamic improvements can extend flight duration, while advancements in propulsion and battery management systems contribute to more efficient energy utilization [4]. In parallel, alternative power sources have been explored, such as hybrid systems with internal combustion engines running on renewable or unrenewable sources [5,6,7], systems with fuel cells and supercapacitors, each offering distinct advantages in terms of energy density, power delivery, and recharging speed. Multi-modal power systems can manage and convert energy between different forms, such as electricity, heat and gas, to meet various demands and ultimately increase the efficiency of the system. Recent developments in UAV technologies have shown an increase in the use of hydrogen to provide an additional power source to multirotors. Hydrogen is known to boost a much higher energy density per unit of mass compared to traditional LiPo batteries, offering 100 times more energy by weight than standard batteries. The downside of hydrogen is that it has a very low energy density per unit of volume, meaning it takes up a lot of space at a low weight. As [7] shows, with a properly sized hydrogen system, a 300% longer flight time can be achieved compared to battery-only systems. Reference [6] demonstrates how hydrogen fuel cells can be the primary power source for the UAV, with the battery system handling peak power demands of the motors in rapidly changing or extreme conditions. However, the integration of such technologies is often constrained by added weight, system complexity, noise, and environmental impact. As a result, autonomous docking stations, which allow UAVs to land, recharge, or swap batteries without human intervention, have emerged as one of the most practical and scalable solutions for enabling persistent and reliable UAV operations as shown in [8]. Existing surveys do not explicitly address the limitations of combining hybrid power systems with docking-based approaches, since endurance is achieved in fundamentally different ways. This difference explains why the two approaches are rarely integrated.

These docking systems can be broadly categorized into fixed ground-based platforms and mobile platforms. Fixed stations are typically deployed at predefined locations and offer stable infrastructure for precise landing, battery charging, and/or swapping, therefore increasing productivity and autonomy of the systems [9]. Mobile docking systems can be integrated with different classes of unmanned platforms, distinguished by the domain in which they operate. These include unmanned ground vehicles (UGVs), surface vessels, and larger aerial platforms, providing increased flexibility by enabling multirotor UAVs to rendezvous with moving bases and thereby extend operational range and autonomy even further, as shown in [10]. The development and integration of such systems relies on a combination of mechanical design, sensing technologies, guidance algorithms, and power transfer methods, all of which are examined in the subsequent sections of this paper.

The ability of a multirotor UAV to land at a designated location for energy replenishment provides a high degree of versatility for automated missions without the need for operator intervention. This is particularly valuable for repetitive missions where the UAV performs recurring tasks, such as inspecting wind turbines across a wind farm. The demand for precise, reliable, and automated landing capabilities has driven advancements in the area of precision landing, fostering extensive research into various approaches that support docking systems, including vision-based guidance solutions. In parallel, significant progress has been achieved in the development of conventional charging methods and mechanisms, followed by advancements in wireless charging mechanisms [11,12,13].

In this paper, a comprehensive review of UAV docking systems is presented, focusing on their role in extending multirotor UAV endurance and enabling autonomous long-duration missions. The first contribution is a structured categorization of docking solutions, distinguishing between fixed and mobile platforms. This is followed by an in-depth examination of enabling technologies, including positioning and landing mechanisms, charging methods, and automated battery swapping approaches. Relevant papers were selected using key words specific to the field of study. These keywords included common terms such as “docking”, “charging”, and “landing”, along with “uav docking”, “uav charging”, and “uav landing”; additionally, keywords such as “battery swapping” or “mobile platform” were used to go into specific subfields to gain valuable insight. As the keywords that were used are quite common and encompass a large set of technologies, the same were utilized for some of the keywords of this very paper. The literature search was conducted primarily through scientific databases such as Web of Science and Scopus. Most of the reviewed works are journal articles, with a smaller portion coming from conference proceedings. The majority of the publications were published within the past five years, ensuring that recent technological developments are well represented. Publications lacking sufficient technical detail were excluded. Using this methodology, recent advances in technology and implementation were compared and synthesized to review the current state of docking systems used by UAVs. The main focus of the research was to make a comprehensive overview of the state of docking systems from the hardware perspective, as the software side of navigation and landing is well documented. Overall, over a hundred papers were analyzed using major keywords, with around seventy of them being used in the scope of this paper. The paper outlines persistent challenges and highlights future research directions required to achieve scalable and fully autonomous UAV operations.

The remainder of this paper is organized as follows. Section 2 introduces a categorization of UAV docking systems, distinguishing between fixed ground-based platforms and mobile solutions. Section 3 reviews positioning and landing mechanisms, covering mechanical, passive, and sensor-based approaches. Section 4 examines charging technologies, including both physical contact methods and wireless power transfer. Section 5 discusses automated battery swapping concepts. Section 6 provides a discussion of open challenges and future perspectives, while Section 7 concludes the paper.

2. Categorization of Multirotor UAV Docking Systems

To complement the qualitative categorization of UAV docking systems, a quantitative overview of the surveyed literature is provided in Figure 2. In total, 69 references regarding docking systems were grouped into fixed ground-based docking systems and mobile docking systems, the latter further subdivided into docking platforms implemented on UGVs, UAV Carrier Platforms, and Unmanned Surface Vehicles (USVs). This breakdown highlights the relative distribution of research efforts across different implementation domains, providing insight into which approaches have received more attention in recent studies.

2.1. Fixed Docking Systems

Fixed docking systems represent stationary platforms for multirotor landings with various capabilities. These systems often incorporate weather protection, security features, automated charging or swapping of batteries and storage management [2,14] for loading or offloading of cargo from the multirotor. Fixed docking stations are typically deployed in fixed, static, locations such as near substations or along power lines. They are often referred to as “drone in a box” systems as they completely cover up the multirotor when sitting idle. Their primary functions include charging or replacing the UAV battery after landing. Having the ability for the UAV to have its battery charged or swapped autonomously makes the system more efficient and cheaper to operate as there is no need for an additional operator to stand nearby to initiate charging of the batteries. The docking station also must compensate for variations in the landing position which can vary due to errors in navigation or external forces acting upon the UAV such as wind. Various weather factors also require the docking station to manage and safely store the UAV to protect it from rain, dust and similar substances that could damage the UAV or lead to further issues. Docking platform hardware symmetry plays a role in increasing reliable landings. Symmetrical layouts of landing surfaces, protective covers, or access points allow the UAV to approach from multiple angles without requiring strict alignment, improving flexibility and operational efficiency.

In recent studies, the most utilized landing mechanisms use machine vision systems to position the multirotor correctly to land on the platform. Around 51.4% of the 35 analyzed papers utilize this method of landing navigation. These systems utilize onboard cameras and computer vision algorithms to detect and track landing targets through various approaches. Markers are commonly used to aid machine vision as they provide orientation and position data through high contrasts shapes. As far as positioning mechanisms are concerned, most utilize a passive landing mechanism, namely 60% of the analyzed papers. Mechanical docking systems account for the remainder of the analyzed papers, specifically 40% of them, as shown in Figure 3. The most utilized combination of landing and docking systems are vision-passive combinations, accounting for a total of 40% of the 35 papers on fixed docking systems [4,9,15,16,17,18,19,20,21,22,23,24]. Vision-mechanical combinations [12,25,26,27,28] on the other hand account for 11.4% of the analyzed papers. With an overview on the distribution of different technologies utilized in docking stations, a clear picture can be painted on the current heading of the research community. As is obvious from the percentages, the current focus is on the machine vision and software side of docking systems with passive systems taking the lead. This can be attributed to the lower hardware complexity of such systems, therefore being more available to a wider range of researchers unlike complex hardware-driven systems that require lot of prototyping. The overall distribution and percentages are given in

Table 1. Docking, navigation and charging technology overview in accordance with platform mobility.

Platform		Fixed	UGV	UAV	USV
Docking	Mechanical	14 (40%)	5 (25%)	7 (87.5%)	3 (37.5%)
	Passive	21 (60%)	12 (60%)	0	5 (62.5%)
Landing navigation	Machine vision	18 (51.4%)	10 (50%)	5 (62.5%)	2 (25%)
	Sensor fusion	7 (20%)	6 (30%)	0	3 (37.5%)
	Other	2 (5.7%)	1 (5%)	0	2 (25%)
Charging	Direct contact	3 (8.6%)	1 (5%)	0	1 (12.5%)
	Wireless	5 (14.3%)	3 (15%)	1 (12.5%)	0
	Other	10 (28.6%)	3 (15%)	3 (37.5%)	1 (12.5%)

for corresponding categories.

Docking systems with passive positioning rely on geometric guidance rather than active mechanical positioning. Pure passive positioning systems feature positioning with common global navigation satellite system (GNSS) or direct methods with passive positioning on the docking platform [14,29]. Passive systems with enhanced approach strategies are presented in [4,16,30]. Mechanical docking systems with a focus on battery swapping are shown in [8,31]. References [32,33,34] demonstrate advanced mechanical docking systems such as cooperation between multiple ground station systems, advanced Stewart platform compensation for landing and in-flight docking between two multirotors.

Figure 3. Treemap visualization of references categorized by UAV docking system aspects. References concerning Landing navigation: Machine vision [2,4,9,11,12,13,15,16,17,18,19,20,21,22,23,24,25,26,27,28,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50], Sensor fusion [1,10,13,27,30,32,33,34,46,47,48,51,52,53,54,55,56], Other [9,29,57,58,59]. Concerning Docking: Passive [4,9,10,13,14,15,16,17,18,19,20,21,22,23,24,29,30,35,39,40,44,45,46,48,50,51,52,53,54,55,58,60,61,62,63,64,65,66], Mechanical [2,3,8,11,12,25,26,27,28,31,32,33,34,36,37,38,41,42,43,49,56,57,59,65,67,68,69,70,71,72,73,74]. Concerning Platform mobility: Fixed ground [4,8,9,12,14,15,16,17,18,19,20,21,22,23,24,25,27,28,29,30,31,32,33,34,48,50,54,60,63,64,69,70,71,73,74,75,76], Mobile UGV [10,13,16,26,37,38,39,44,45,46,47,52,53,59,60,61,66,72,77,78], Mobile UAV [3,11,36,41,42,43,67,79], Mobile USV [35,40,51,55,56,57,58,68]. Concerning Charging: Other [3,4,5,6,7,8,11,31,49,57,61,63,64,65,67,69,70,71,72,75,76,78,80,81], Wireless [4,9,12,32,37,59,62,65,74,77,79,82,83], Direct contact [2,14,28,29,35,38,65].

Figure 3. Treemap visualization of references categorized by UAV docking system aspects. References concerning Landing navigation: Machine vision [2,4,9,11,12,13,15,16,17,18,19,20,21,22,23,24,25,26,27,28,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50], Sensor fusion [1,10,13,27,30,32,33,34,46,47,48,51,52,53,54,55,56], Other [9,29,57,58,59]. Concerning Docking: Passive [4,9,10,13,14,15,16,17,18,19,20,21,22,23,24,29,30,35,39,40,44,45,46,48,50,51,52,53,54,55,58,60,61,62,63,64,65,66], Mechanical [2,3,8,11,12,25,26,27,28,31,32,33,34,36,37,38,41,42,43,49,56,57,59,65,67,68,69,70,71,72,73,74]. Concerning Platform mobility: Fixed ground [4,8,9,12,14,15,16,17,18,19,20,21,22,23,24,25,27,28,29,30,31,32,33,34,48,50,54,60,63,64,69,70,71,73,74,75,76], Mobile UGV [10,13,16,26,37,38,39,44,45,46,47,52,53,59,60,61,66,72,77,78], Mobile UAV [3,11,36,41,42,43,67,79], Mobile USV [35,40,51,55,56,57,58,68]. Concerning Charging: Other [3,4,5,6,7,8,11,31,49,57,61,63,64,65,67,69,70,71,72,75,76,78,80,81], Wireless [4,9,12,32,37,59,62,65,74,77,79,82,83], Direct contact [2,14,28,29,35,38,65].

2.2. Mobile Docking Systems

Mobile docking systems represent docking platforms mounted on or incorporated into various vehicles that are inherently mobile. These types of systems are mainly mounted on various UGV-s [29], USV-s [35] or on UAV-s [36]. In principle, mobile systems represent a kind of extension of fixed docking systems, but this also increases the complexity of the system, especially from the aspect of mutual integration into a heterogeneous robotic system. The main purpose of mobile systems is to build upon the foundations of fixed docking systems, with the ability to move the docking platform to better adapt to the flow of the operations and further reduce downtime. By moving closer to the work area, mobile docking stations enable shorter flights to and from the work area designated by the operator therefore reducing downtime and increasing effective work time. Such systems depend on technologies used in fixed systems, whether they are contact and wireless charging technologies or battery swapping systems, to facilitate switching between operations. Recently, tethered systems have been making a comeback due to the developments in mobile docking systems providing a way to move the docking platform and therefore allow the tether to operate in many different areas. Tethered systems, such as the one presented in [57], do not suffer from the issues of signal interference or battery time, as the power is mainly transmitted over the tether along with the data line, which is commonly implemented as an optical cable.

According to recent studies, ground-based mobile docking systems account for 28.9% of the 69 references analyzed regarding docking station categories [10,13,16,26,37,38,39]. This category encompasses diverse ground-based mobile docking systems for various applications such as inspection platforms, mobile bases and general-purpose systems. From the analyzed papers, it is evident that passive positioning is the leading method, accounting for 60%, which is expected because mobile docking systems often need to be lightweight, so passive systems are the preferred method there. Vision guidance covers 50% of the cases and has the benefit of reducing the hardware complexity of the docking station which can further aid in reducing weight.

Offshore operations often require surveillance from above; therefore, easy monitoring and deployment of multirotor UAVs is critical. Systems for landing UAVs on ships or USVs [27,35,40,51,57,58] account for 11.6% of the 69 analyzed papers, and many of them utilized passive positioning systems for landing purposes, around 62.5%. One of the specificities of this category of mobile docking systems is the need for motion compensation, which is critical for surface vessels due to external sea interference. Therefore, robust and sufficiently simple to use mechanisms for landing on ship and USVs are crucial. Around 37.5% of the papers utilize sensor fusion to combat ship dynamics with 25% of the papers utilize machine vision for successful multirotor navigation.

Combining UAV docking systems to connect with other UAVs is a great feat, as is shown in [3,11,34,36,41,42,43]. These papers account for 11.6% of the analyzed papers, and almost all of them utilize mechanical positioning methods to achieve docking, namely 87.5% of them. This method is preferred in UAV systems because there are large forces that can move the landed UAV off the docking aircraft, so rigid coupling is mandatory. Although UAV-to-UAV docking and battery swapping remain technically challenging, recent studies have demonstrated their practical feasibility under controlled experimental conditions. A notable example is the AirDock system proposed in [67], which presents a multirotor platform that can be docked in mid-air and swapped batteries in mid-air. The system uses a fully actuated multirotor design and a rail-car-based transmission mechanism. Experimental validation has confirmed stable in-air docking, safe in-flight battery swapping, and extended flight endurance. Although such architectures confirm the technical viability of in-flight swapping, their practical application remains limited to specialized applications due to the high control complexity, synchronization requirements, and custom mechanical interfaces. Consequently, UAV-UAV docking is currently considered an experimental but promising research direction for endurance-critical missions.

UAV to UGV cooperation is shown in [10,13,16,26,37,38,39,44,45,52,84,85]. From the analyzed papers, it is shown that 60% of the systems utilize passive docking mechanisms to position the UAV on the ground vehicle, while also 50% of the papers employ machine vision to guide the UAV to landing. Considering environments where GNSS positioning is not available, the most used alternative is sensor fusion between various types of sensors. A vision-based approach is shown in [45] which allows a UAV to autonomously detect, track, and land on a moving UAV by processing RGB images from an onboard monocular camera. This concept was validated in simulation using a digital twin in ROS (https://www.ros.org/, accessed on 22 September 2025), Gazebo (https://gazebosim.org/, accessed on 22 September 2025), and ArduPilot Software-in-the-Loop (https://ardupilot.org/dev/docs/sitl-simulator-software-in-the-loop.html, accessed on 22 September 2025), and in field experiments with a DJ550-based hexacopter and a modified Jackal ground robot. Figure 4 provides a schematic overview of the main types of UAV docking mechanisms to better visualize the classification and abbreviations used throughout this paper.

3. Positioning and Landing Mechanisms

To facilitate safe landing and transport of UAV systems, docking systems are implemented to provide a precise and secure location for the landing of UAVs. Before reviewing the specific positioning and landing mechanisms, it is important to note that although multirotor UAVs are inherently nonlinear and unstable open-loop systems, this instability is significantly mitigated by their geometric symmetry, especially in conventional configurations such as quadrotors and hexacopters, where the symmetrical position of the rotors ensures a balanced distribution of aerodynamic forces and moments. The geometric and functional symmetry of multirotor platforms contributes to improved dynamic stability and enables predictable position responses, minimizes coupling between rotary axes, and improves controller performance during fine maneuvers near docking stations. Recent research has shown that symmetrical structures contribute to improved control precision and robustness. For example, multistage PID controllers developed for symmetric UAVs have shown improved transient response and reduced energy consumption [84], while adaptive sliding control approaches based on symmetric position dynamics have achieved superior disturbance rejection [85]. Such findings emphasize that symmetry is not just a geometric property, but one of the functional prerequisites for safe autonomous landing in docking applications. Landing is not the only benefit provided by the docking system, as it is often used to facilitate recharging or refueling of the UAV systems, thus allowing the system to operate autonomously, without the need for operator intervention. By eliminating the need for an operator to overlook the entire process, the efficiency and productivity of the system is increased many times, as the cost needed to operate the system is reduced. To achieve the required level of autonomy, a precise method for positioning and landing the vehicle is required. This section will provide an overview of positioning and landing technologies and mechanisms. These landing and positioning mechanisms are independent of the docking station type and, as such, can be used on any type.

3.1. Mechanical Positioning

Mechanical positioning is an approach in which the multirotor UAV lands on a platform and is afterwards positioned at the desired location using mechanical hardware. Such hardware could be pushing or pulling devices, magnets [37] or similar apparatuses which move the whole multirotor assembly to the desired position. The main advantage of such mechanisms is the fact that the multirotor has a wider area where it can safely land and can therefore have simpler software positioning systems in place, thus reducing the overall cost of the multirotor UAV system. On the contrary, the negatives are in the increased complexity and size of the landing platform required to accommodate the positioning mechanism and the corresponding landing plate itself, which needs to be larger as the multirotor UAV has a wider area to land accordingly. This can be an important limitation when placing the positioning device on mobile docking platforms, such as on UAVs where weight plays an important role. The solution presented in [36] shows a manner of docking where the docking mechanism is integrated into the functional mechanism of in-air refueling, thus providing a quite efficient solution to the weight problem.

Multirotor UAV positioning can also be achieved in the form of a robotic arm system capable of “grabbing” the multirotor in the air and thereby guiding it safely towards the desired location. Such systems can be used on marine surface vehicles where waves and hull swaying are very common and therefore pose a problem that disrupts the precise landing procedure. Such a system was demonstrated in [27,68] and was shown to greatly improve the landing ability of the multirotor systems used. Symmetrical designs of landing plates and guiding mechanisms are often used to evenly distribute forces and moments during positioning.

3.2. Passive Positioning

Unlike active methods with mechanical positioning that use actuators, passive systems typically feature no moving mechanical parts or actuators, which simplifies their design, reduces energy consumption, and enhances robustness. These systems often utilize gravity and the aircraft’s weight to properly guide the landing, relying on the aircraft’s own motion.

Conical funnels are the most commonly used implementation of passive positioning because the multirotor has a bigger area to approach and still manages to land accurately. To name a few, the funnel can be implemented as a single large funnel, a large funnel to guide the landing gear legs or individual funnels for guiding each of the landing legs. By modifying the design of the funnel, more complex shapes can be implemented that can act as a landing guide in only one orientation, further increasing the precision of the final landing position and orientation. The symmetrical design of funnels or other types of passive positioning elements is even more important than the symmetrical design of mechanical systems, as it aligns with the inherent symmetry of the multirotor UAV.

Ski-Type landing gear legs implemented with slopes provide a variation on the funnel mechanism by utilizing flat sloping surfaces or lateral rails on the landing platform to guide UAVs equipped with ski-type landing gear as shown in patent [60].

3.3. Vision-Based Approaches

Although proper positioning is fundamental for ensuring a safe aircraft landing on a landing platform, the structured sequence of the landing approach procedures is arguably of even greater importance. The need for precise landing has driven research into the development and use of many types of navigation and positioning systems. For most of the navigation flights, GNSS is used as it is widely available and easy to implement. Since GNSS is a global system, it has been utilized extensively in recent years and has been the standard for landing aircraft at centimeter-level precision. To further increase accuracy, a machine learning system has been developed to analyze video data in real time to recognize certain shapes and objects in the frame. In Refs. [14,25,26,27,28,29,51,53], Fiducial markers (ArUco, AprilTag, H-shaped, T-shaped markers) are placed on the platforms and provide a way for the aircraft to analyze and orient itself according to the position of the markers in the image. In some cases, markers are intentionally placed asymmetrically to provide a unique orientation reference, since asymmetrical placement of markers helps to avoid ambiguity in orientation. Additionally, machine vision for tracking objects and predetermined points in the frame without the use of markers can be utilized, but at a lower level.

The main downside of machine vision is the need for onboard image processing. While this is not a significant issue for larger UAVs, as they have the capacity to carry compact onboard computers it can become a limiting factor for small-scale platforms, potentially offsetting the benefits of vision-based navigation. Nevertheless, vision-based approaches are widely used in areas without GNSS access, such as indoor areas, to provide relative positioning [46]. Recent studies have shown that combining machine vision with complementary sensors greatly improves autonomous landing, especially in complex or GNSS-denied environments. For instance, in [54], a robust control and estimation approach that enables a quadcopter to autonomously and precisely land on a dynamically moving platform is presented. The UWB-IMU-vision framework [47] integrates ultra-wideband sensors and an IMU to estimate relative position during approach, while visual feedback improves guidance during final descent. Outdoor tests using a Pixhawk 4 Mini, GNSS, a monocular camera, a UWB tag, a Jetson Xavier NX onboard computer, and a TFmini rangefinder demonstrated accurate landings on a platform equipped with a 0.8 × 0.8 m landmark and four UWB base stations. The coordinate setup for this multi-sensor landing scenario is shown in Figure 5.

3.4. Sensor Fusion Methods

Many systems nowadays work in tandem with various others to complement each other and improve overall functionality. Multirotor sensors are often very limited and confined in their working area and only work well in a specific set of operating conditions. Utilizing other types of sensors that complement those shortcomings of other sensors is precisely what is today called sensor fusion. Sensor fusion is a process of combining data from multiple onboard sensors to create more comprehensive sensor feedback for the multirotor UAV state. By relying on the strengths and advantages of each individual sensor, the system can mitigate the weaknesses and shortcomings of each individual sensor. These systems use various filters, such as the Kalman Filter or Extended Kalman filter, to process the data at speeds high enough for multirotor operation. Such filters are often heavy on the processing side because they use complex operations to perform the filtering. An interesting system has been proposed in [53] which demonstrates how combining radio frequency positioning with inertial sensing creates redundant and complementary positioning systems that maintain accuracy when individual sensors fail or become unreliable. A similar solution has also been presented in [47] by combining UWB, inertial measurements and computer vision to precisely land UAVs, as previously shown.

Landing on platforms that are inherently unstable, such as on ships or USVs, also provides opportunities for sensor fusion to shine. Motion compensation techniques specifically tackle these kinds of landing challenges by using sensor fusion to predict and counteract platform movement in real-time. In [39], a method has been implemented that combines visual tracking with inertial measurement to enable predictive control algorithms to compensate for the movement of the ship deck caused by waves, allowing UAVs to maintain stable approach trajectories despite platform instability. Similarly, [33,55,56] also showcase advanced systems with mechanical compensation of platform motion to improve the landing accuracy and robustness of the landing system for UAV–USV cooperation.

3.5. Precision and Modular Docking Solutions

Precision landings are a key part of improving the capabilities of UAVs. Conventional GNSS positioning systems usually land the UAVs within a range of a few centimeters to tens of centimeters around the target, while with precise positioning, pinpoint accuracy is achieved. By combining various technologies such as computer vision, GNSS, and specialized markers (fiducials) on the landing platform, the flight software automatically calculates the location of the UAV relative to the landing platform. With continuous monitoring of the position, the UAV can perform a precise landing on to the platform. Generally, these systems are utilized in highly automated systems to reduce the possibility of inaccurate positioning on the landing or charging platforms.

In [15], a computer vision-based control loop has been developed to reduce the touchdown error to less than 5 cm by fusing onboard camera data with inertial measurements. Visual markers can also be used, such as in [18] to calibrate the UAVs position in real-time to within 2 cm, enabling high repeatability in GNSS-denied environments. Such systems often need to incorporate additional lighting in the area so that the cameras can see the markers correctly. Reference [21] demonstrates the use of a machine learning model taught on a large dataset describing various situations of landing in low-contrast or low-light conditions to provide a simple and precise landing mechanism. Further extensions to these ideas have been implemented in [30] for infrastructure inspection, which is often a highly repeatable set of tasks where an automated system highly increases efficiency. A similar system has also been implemented in [48] to provide indoor micro-UAV landing with a high accuracy.

3.6. Summary and Trade-Offs in Positioning and Landing Mechanisms

Table 2. Evaluation table for landing methods.

	Mechanical Positioning	Passive Positioning	Machine Vision	Sensor Fusion
Accuracy	High	Medium	High	Very high
Cost	High	Low	Very high	High
Complexity	High	Low	High	High
Adaptability	Medium	Low	Medium	High

provides a comparative evaluation of the main landing methods, highlighting differences in accuracy, cost, computational and mechanical complexity, and environmental adaptability. These differences reflect the inherent trade-offs between hardware design and algorithmic sophistication. Approaches with higher mechanical complexity, such as mechanical positioning systems, can achieve precise alignment while reducing computational requirements. In contrast, methods with simpler mechanical setups, including machine vision and sensor fusion, rely more heavily on advanced algorithms to maintain precision, especially under limited payload constraints or dynamic conditions. A similar relationship between mechanical complexity and control sophistication will also be observed in the context of battery swapping systems discussed in later sections.

4. Charging Mechanisms and Technologies

This section will provide an overview of basic charging technologies, which represent a key aspect of hardware in the case of the docking stations. Two basic categories will be covered, Physical Contact Charging and Wireless Power Transfer.

4.1. Physical Contact Charging

Direct contact charging involves establishing physical contact between the conductive elements on the UAV and conductive elements on the docking station. Depending on the design, the conductive elements can be standardized or unconventional connector designs or, on the other hand, conductive plates located on both the UAV and the docking station. Since the current transmitted to the UAV is of a DC nature, it is important to keep the polarity of the connection through additional electrical components. By making the conductive plates in a “chess-pad” configuration with alternating positive and negative conductive squares, landing accuracy issues can be mitigated to always ensure charging regardless of the exact position on the platform.

The combination of the “chess-pad” symmetrical configuration and polarity modulator circuitry allows the UAV to land and charge without issue regardless of variations in the landing procedure [14]. In this approach, the landing pad consists of square copper plates, as shown in Figure 6 that alternate between positive and negative polarity, while the onboard circuitry features six integrated bridge-rectifier diodes and a current indicator to ensure correct polarity and confirm successful docking. This design not only mitigates the need for precise landing accuracy but also supports different power sources, such as solar panels. Nevertheless, one of the major concerns with such a system lies in the possibility of arcing if the connection is not properly established, which can lead to the breakdown of contact materials and pose a potential fire hazard. Overall, this system has been reported to reach high efficiency, nearly equivalent to directly charging the UAV batteries.

4.2. Wireless Power Transfer

Wireless Power Transfer (WPT) systems represent the transfer of electrical energy using electromagnetic induction to counter the disadvantages of direct contact charging. These systems offer improved reliability in terms of automated charging by reducing the need to precise positioning of the UAV onto the charging pads. Since the power transmission is completely wireless, there is no damage being done to the system that would require maintenance over time, such as the contact pads in case of direct contact charging systems, making these systems more reliable and robust in this regard. Also, by eliminating the contact plates, there is no arcing and sparking which in certain environments could lead to fire hazards or even explosions. Wireless power transfer systems are generally categorized by transmission distance into near-field and far-field, and by method into radiative electromagnetic (EM) and non-EM (acoustic or optical). Figure 7 shows a schematic representation of the far-field WPT system.

Inductive charging represents the transfer of energy using the magnetic resonance between two coils positioned one above the other. This type of technology is often used for low-power charging systems such as smartphones and smartwatches where you can place the device on a charging dock, and the energy is transferred. The challenges of using such a system include misalignment tolerance [62,82], interference with other electromagnetic fields, and the need for lightweight components to avoid reducing payload. Landing accuracy directly impacts the coupling factor and transfer efficiency because the coils required for energy transfer need to be positioned optimally for maximum energy transfer, with each deviation from the optimal position reducing efficiency. Advanced mutual-inductance dynamic-predicted control schemes have been proposed to maintain constant-current charging in flight, improving robustness under varying UAV payloads and magnetic conditions [79]. Overall, a system can achieve DC-to-DC efficiency greater than 86% for constant-current constant-voltage (CC-CV) charging up to 500 W. Recent developments, such as the patented three-phase WPT system for UAVs [77], further improve inductive charging performance.

Capacitive charging is a wireless power transfer technology that uses electrical fields to transfer electricity. The main benefit is the ability to integrate capacitive coupling plates into surfaces on the landing platforms and UAV landing gears for high coupling capacitance and minimal aerodynamic influence. This in turn increases the efficiency and capacity of the whole system.

Laser power transfer uses laser beams of specific wavelengths and frequencies to power photovoltaic (PV) cells mounted on the UAV. This technology is primarily used in space and military applications because it can deliver large amounts of energy to the receiver using narrow beams and therefore offers potential for unlimited operation. Disadvantages of this technology include objects blocking the beam, mobility of the ground system, and performance in long-range flights. Additionally, laser beams are hazardous to human health and living environments [61]. Overall, the efficiency of the system is low because PV cells have low conversion efficiency (around 20%).

Near-field systems, such as inductive and capacitive charging, provide high power density and efficiency due to the close coupling between the transmitter and receiver. They are also safer and more environmentally friendly, as the electromagnetic fields remain confined to a small area with minimal radiation leakage. In contrast, far-field systems, including microwave and laser-based transmission, allow power delivery over long distances, but with lower power density and efficiency due to beam divergence and atmospheric losses. These systems raise safety and environmental concerns, as high-intensity radiation can pose a risk to humans and wildlife.

5. Battery Swapping Approaches

Extending the autonomy of multirotor UAVs is mainly a challenge of providing power to the systems so that they can operate for longer periods of time. Through the technologies presented so far, whether it is direct contact or wireless charging, this is achieved by directly charging the battery. This operation, of course, takes significant amounts of time because the batteries must be charged with lower currents to prevent overheating, which actually means that the multirotor battery takes longer to charge than the multirotor can be airborne.

5.1. Swapping Mechanisms on Fixed Platforms

To overcome the downside of the time needed to directly charge a battery, it is more efficient to mechanically swap the battery on the multirotor. By unplug the empty battery and plug in a charged battery, the effective time needed to become operational after landing amounts to the time needed to mechanically swap the batteries. As previously noted, the development of reliable landing and docking mechanisms for UAVs, whether fixed or mobile platforms, is far more complex than for ground robots. This stems from the dynamics of the multirotor UAVs and the fact that they operate in 3D space, while ground robots, in simple terms, can be treated as moving in 2D space. As a result, tasks such as positioning and battery replacement are much easier to solve. One example of such a system is the battery replacement platform for an agricultural robot, presented in [83]. Although challenging, research into multirotor UAV docking and automated battery replacement has been ongoing for some time. Early efforts have focused on low-cost, fixed battery replacement systems, such as the one presented in [75].

Research has been intensified in recent years in both the software and hardware aspects. In [63], the design and construction of a low-cost service station for quadrotors is presented, which uses position control and a simple mechanical mechanism to replace a discharged battery with a charged one automatically. Battery swapping mechanisms are, in essence, robotic devices [69,70] whose main task is to plug/unplug the battery from the UAV and then perform the corresponding action at the docking slot where the discharged battery needs to be charged, and the charged battery needs to be delivered to the UAV. Symmetrical design of the robotic mechanism and docking interface is often employed to ensure reliable alignment and minimize the risk of misconnection. The design of the robotic mechanism can be different. Most often, there are Cartesian configurations with translational degrees of freedom, as shown in Figure 8. But there are also versions where these mechanisms are presented with rotational degrees, such as the automated battery management system presented in [71].

In [64], the authors present a fully autonomous system for replacing batteries in small multirotor UAVs, consisting of a robotic ground station with a quick-charge power unit supported by solar panels. Experiments demonstrated that the swapping process can be completed in under three minutes. Additionally, alternative concepts such as the “Inverted Docking Station” solution [31] are being explored, where the UAV docks on a platform located above it, which is schematically shown in Figure 9.

Recent patent filings also show the trend of developments in UAV docking stations for battery swapping systems. Patent [49] demonstrated a system where the UAV can land in the docking bay of the platform and have the battery on it replaced with a fully charged one. Subsequently, the empty battery is placed on the rotating track to move out of the way. Patent [80] demonstrates a fully enclosed and automatic battery charging and swapping system. The cabin provides a way to close to protect the drone from elements and is equipped with landing and takeoff mechanisms, a battery swap and charging mechanism, and a lifting mechanism. Patent [78] presents another fully enclosed battery swapping system where the battery is changed using a manipulator. This enables precise positioning of the battery regardless of the UAV’s position.

5.2. Swapping Mechanisms Using Mobile Platforms

The combination of multirotors and UGV robotic platforms is emerging as one of the feasible solutions for increasing system autonomy. A docking platform mounted on the UGV robot allows the UAV to perform a precise landing and quick battery replacement, eliminating the need for a return to a fixed station or manual operator intervention. In addition to research papers, there are already commercial products that offer this approach. An example is the Autonomous Drone Ground Handling Station for Precision Landing [76] by Embention, which offers an integrated system for precision landing and automatic battery replacement on off-road vehicles. Such approaches extend the operational reach of UAVs, particularly in applications where long-term or repetitive activity is required, such as infrastructure inspections or surveillance.

Research has gone a step further, towards very complex concepts that include the docking of UAVs and changing batteries in flight. Reference [3] presents an experimental system in which a “flying battery”, representing a small quadcopter, docks on the platform of the main UAV, makes electrical contact, and transfers energy, which significantly increases the flight time. Furthermore, in [72], a mechanism was developed that enables extremely fast replacement of batteries during flight. In [8], a solution has been presented that tries to mimic the behavior of bee colonies, where the UAV automatically returns to the ground docking station when the onboard power level drops below a predetermined threshold. Then, numerous concepts such as the Aerostat-Based system for replacing batteries in the air have been proposed [11]. These approaches demonstrate how the boundaries of autonomy can be significantly pushed, but at the cost of high technical complexity and the need for sophisticated algorithms for guidance, positioning, and safe energy transfer.

When comparing different battery swapping approaches, a key trade-off lies between mechanical complexity, software sophistication, and payload constraints. Fixed, and partially UGV, and USV platforms face fewer payload constraints, allowing for more complex mechanisms that achieve higher alignment accuracy, but at the cost of increased maintenance. Conversely, systems with tighter payload constraints, such as aerial platforms, rely more on advanced perception and control algorithms to compensate for simpler mechanical designs. In general, higher mechanical complexity reduces the software requirements for precise positioning, and vice versa. The optimal design therefore depends on balancing mechanical robustness, control accuracy, and maintenance effort to ensure reliable and scalable UAV autonomy.

6. Discussion and Future Directions

Throughout this paper, a detailed overview of technological advancements in UAV docking systems has been presented. The current direction of the technology is to enhance UAV autonomy and increase flight endurance, thereby extending overall operational duration. Automated docking and charging reduces downtime by eliminating the need for the aircraft to return to an operator for manual battery replacement. Recent advances in mobile docking platforms enhance UAV autonomy and operational efficiency by reducing the distance and time required for the UAV to reach the docking station, as the platform can be positioned much closer to the work area.

When this mobility is combined with contact or wireless charging, a fully automated system can be achieved, with some companies already holding patents for UAV charging solutions, patent [81]. Recent advancements in autonomous UAV charging based on hybrid wireless power transfer concepts have been reported in [65], demonstrating improved energy efficiency and operational range. Since the operator no longer needs to manually change or recharge batteries, long-duration and extended-range operations can be conducted with proper planning, making this especially advantageous for missions involving repetitive tasks. Combining such systems with built-in solar charging, as shown in [50], potentially allows such systems to operate in remote areas for months without the need for any human intervention. The future of UAV docking systems lies in the integration of the reviewed technologies, representing a significant step toward minimizing operator intervention and deployment time while ensuring complete flight autonomy. It is of course important to maintain remote monitoring and control capabilities via, for example, mobile networks, while docking stations use renewable energy sources to power the station itself and the UAV.

Current struggles with advances in the field stem from the challenge of integrating vision systems, docking and charging mechanisms and other sensors into a cohesive autonomous system. Maintaining symmetric design in docking platforms and mechanical interfaces is important, as it ensures stable alignment of multirotor UAVs, and reduces the risk of positioning errors during autonomous operations. Each of these components is essential for achieving the required level of autonomy. In this paper alone, 42.9% of the 85 references mention utilizing machine vision systems to navigate the landing procedures. While the technology is very flexible in the way it can be adapted to the specific scenario it is used in, it also requires heavy computing power to manage all the calculations. This in turn adds complexity to the vehicles as well as cost. With the current state of technology, systems should be integrated to simplify the load on each individual system and on the overall computing power needed to perform precise calculations Implementing many different types of features into a single system comes with a trade-off between precision, reliability, cost and complexity. Sensor fusion approaches have been considered in some papers, around 20.2% of the 85 references utilize this method and it shows that there are certain advances in this field, but maybe not in the capacity that is needed Some solutions are very reliable and precise in their operation, but due to their complexity the cost is too high for consumer markets, and the intricacy of the system is not approachable by many. This is where a standardized system needs to be implemented to regulate and provide industry standard for levels of docking systems in accordance with charging protocols and safety regulations.

In order for such systems to be expanded and widely used, it is important to make them easily scalable. The system must be simple to implement and operationally use, making it accessible to a larger number of users. Simplified use and lower mechanical complexity significantly reduce the possibility of failures. These aspects are addressed given that the development of a heterogeneous robotic system is underway within the framework of the national research project “Advancing Autonomy: A Concept for a Multipurpose Ground-Aerial Robotic System—cAMGARS”, which focuses on the development of an integrated UGV-UAV robotic system for outdoor environments. In this concept, a tracked UGV serves as a mobile docking and charging platform equipped with a robotic battery exchange mechanism, while a multirotor UAV performs repetitive missions and autonomously lands on the UGV to replace depleted batteries with fully charged ones. The findings and insights from this review directly contribute to defining design parameters and scalability considerations for the cAMGARS system.

Since the systems potentially have a high degree of autonomy and can be controlled remotely, there is a risk of unauthorized takeover or cyberattacks. Such systems, due to their functional power and remote access, could be abused; therefore, it is necessary to implement highly secure protocols for communication and operation. Artificial intelligence (AI) is definitely a vision of the future, as in many areas in this field where new advances are always needed. With AI, deployment could be automated when certain criteria are met near a station so that the UAV can deploy and monitor the environment, as well as automatically process the data received. AI could also be used to coordinate multiple UAV systems in a coordinated swarm manner, which offers many advantages for certain types of operations. Furthermore, by integrating new energy sources that could be cheaper and easier to produce than conventional batteries, it would be possible to offset the carbon footprint of battery charging and production, as well as increase flight performance and system duration.

For the industry to adopt these systems on a large scale, many comparisons between autonomous and standard systems would need to be made. This would take into account all the costs associated with these systems, along with their remote control. Companies would also prefer systems that offer patented solutions, such as those presented in patents [49,66,73,74,78,80] for docking stations. This offers a more professional and concrete plan for the company to develop the system, and then for interested companies to invest and purchase. Regulatory legislation is also a difficult topic to discuss, as it varies from country to country. Creating a universal solution for a broad market would be too complex and would most likely not yet be harmonized across countries. Analysis of usage rates and legislation would play an important role in determining which countries would benefit from these systems.

7. Conclusions

Throughout this review paper, current advances in autonomous docking, charging, and battery-swapping systems for multirotor UAVs are summarized. Research shows that reliable operation depends on accurate positioning, robust landing, and efficient energy transfer. The design of docking systems, including symmetric layouts, plays a key role in ensuring UAV stability and reliable alignment. Fixed docking platforms are the most mature from a research and application perspective, while mobile solutions, which extend endurance, are gaining ground in academia and consequently in applications. Vision-based systems have taken a majority role for precise navigation and landing on platforms, with passive landing mechanisms also taking the lead in their respective category. This shows that simple navigation and positioning mechanisms are preferred over more complex mechanical systems. In the charging field, various technologies are presented, including physical contact charging methods and WPT techniques. Among the hardware solutions, battery swapping concepts are particularly emphasized, emerging as an attractive alternative for missions where charging downtime is unacceptable, as this autonomy leads to flight time extensions and significant reduction in human intervention requirements. The main contribution of this review lies in its systematic, hardware-oriented classification of autonomous multirotor UAV docking systems. By focusing on the structural, mechanical, electrical, and control design aspects of the reviewed solutions, the paper provides a coherent overview of current technologies and their integration principles. Through this systematic categorization, the review highlights the merits and demerits of existing systems in terms of accuracy, complexity, robustness, cost, and overall implementation feasibility.

With new advances in processing of sensor data, sensor fusion systems perform sub-centimeter level landing accuracy through advanced control algorithms. The overall focus is on implementing simpler navigation systems to provide a cheap and readily available way to land UAVs on docking stations. This in turn requires a reliable and standardized system for docking various UAVs, along with the required safety systems for long-term use. AI is starting to play a big role by helping refine the current solutions to new heights and to improve the systems that are already in place today.