Nomenclature

ABC

Artificial Bee Colony

AC

air cooled

AAM

advanced air mobiltiy

AI

artificial intelligence

APF

artificial potential field

BaFA

back-and-forth algorithm

BEC

battery eliminator circuit

BiCC

bipartite cooperative coevolution

BMS

battery management systems

CFD

computational fluid dynamics

DMFC

direct methanol fuel cell

DPA

dijkstra’s path algorithm

DC

direct current

EEA ^*

energy efficient A^*

EMS

energy management system

ECMS

equivalent consumption minimisation strategy

EEMS

external energy maximisation strategy

ESC

electronic speed control

E ${M^2}$ CPP

energy-aware multi-uav multi-area coverage path planning

EOL

end-of-life

EAOA

energy-aware obstacle avoidance

EaPP

energy-aware path planning

FNNSCP

A fuzzy neural network adaptive sequential convex optimisation

FL

fuzzy logic

GPS

Global Position System

GA

genetic algorithm

GNSS

Global Navigation Satellite Systems

HMPC

hierarchical model predictive control

LC

liquid cooled

MPC

model predictive control

ML

machine laearning

mCPP

multi-UAV coverage path planning

MO-GA

multi objective-genetic algorithm

MOT

maximum operating temperature

NFZs

non-flying zones

OGSE

online generalised shape expansion

OSD

on screen display

I-PSO

improved particle swarm optimisation

I

current

PID

proportional integral derivative

PPS

parallel partitioning along a side

P

power

PRM

probabilistic roadmap

PEMFC

proton-exchange membrane fuel cell

RL

reinforcement learning

RRT

rapidly exploring random tree

RUL

remaining useful life

SOC

state-of-charge

SOFC

solid oxide fuel cell

SOH

state-of-health

SARSA

state–action–reward–state–action

TIG

thermionic generator

TEG

thermoelectric generator

UAV

unmanned arial vehicle

UAM

urban air mobility

VTOL

verticle take-off landing

V

voltage

Other nomenclature and symbols are defined as they appear.

2.0 Current review studies

Since this paper advocates the combined consideration of hydrogen-based propulsion and path planning to increase the flight time and payload capacity of UAVs, review papers on these topics are mentioned in this section.

2.2 On the hydrogen-powered propulsion

Table2 compares the review papers on energy management and fuel cell technology of hydrogen-powered unmanned aerial vehicles. This comparison is based on the topics; the path planning, energy management, UAV system components and hybrid typologies, hydrogen storage technologies and conceptual design for hybrid propulsion. None of these review studies has addressed the topic of energy-aware path planning in detail, except for the suggestions for future works in Ref. [Reference Zhang, Qiu, Chen, Li, Liu, Liu, Zhang and Hwa62]. As mentioned in the Ref. [Reference Zhang, Qiu, Chen, Li, Liu, Liu, Zhang and Hwa62], energy management and path planning should be investigated together to improve the limited flight time of unmanned aerial vehicles.

Table 2. Comparison of existing review studies on hydrogen-powered UAVs

^a Remark for the column of components the listed review studies incorporate/compare: 1 - path planning, 2 - energy management, 3 - UAV system component and hybrid typologies, 4 - hydrogen storage technologies, 5 - conceptual design for hybrid propulsion.

Figure 4. UAV application areas and some examples of hydrogen-powered UAVs. a: delivery, b: spy, c: agriculture, d: emergency, e: inspection and surveying, f: UAM, g: agriculture, deployment of biocontrols, h: inspection. (adapted from Refs [Reference Boukoberine, Zhou and Benbouzid32, Reference Zhang, Qiu, Chen, Li, Liu, Liu, Zhang and Hwa62, Reference Xu, Huangfu, Ma, Xie, Song, Zhao, Yang, Wang and Xu68] (pictures for Refs [Reference Boukoberine, Zhou and Benbouzid32, Reference Boukoberine, Zhou and Benbouzid32, 73–77]).

3.0 Applications for hydrogen-based uavs

Small UAVs are currently a commercialised product used in many military and civilian areas. As UAVs’ payload capacity and flight durability increase, their usage will also increase simultaneously [Reference Mariscal, Depcik, Chao, Wu and Li15, Reference Huang, Li, Ma, Huang, Zheng and Song19, Reference Marqués, Montero, Sánchez-Diaz and Quintanilla70]. To improve these UAV characteristics, fuel cell hybridisation or only fuel cell-powered propulsion has increased the use of unmanned aerial vehicles in civil and military applications. In addition to flight durability, the level of autonomy of UAVs is another characteristic that directly affects the intensity of their use [Reference Mahmoud Zadeh, Powers and Bairam Zadeh71]. With advances in autonomy, we may run into the use of small UAVs in many new areas.

The applications of small unmanned aerial vehicles can be broadly classified as inspection and monitoring, delivery, agriculture, wireless coverage, military and entertainment. Figure4 summarises the leading commercial applications of fuel cell-powered unmanned aerial vehicles with their mission. In (Reference Oladosu, Pasupuleti, Kiong, Koh and Yusaf72) Refs [Reference Çnar, Kandemir and Donateo69] comprehensive reviews of the literature on fuel cell-powered UAVs are provided.

With the developments in energy storage and transform areas in recent years, hydrogen-based UAVs have now become commercial. The leading ones are shown in Table 3. They participate mostly in delivery, education, surveillance, inspection, agriculture and monitoring. Hydrogen-based propulsion is not limited to small UAVs but is now a potential solution to fossil fuels for aircraft of all sizes. Hydrogen-based propulsion systems are being designed for passenger transport, especially in urban air mobility. For example, the VTOL aircraft called AMSL Vertiia made its first flight in 2024 and has a range of up to 1,000 km. Additionally, the tilt-wing type hydrogen-powered eVTOL vehicle that Joby developed for UAM has completed successfully flight tests.

Table 3. Examples of fuel cell-powered commercialised UAVs

^a L: liquid, C: compressed;

^b Take-off mass;

^c Estimated flight time

Table 4 lists leading fuel cell producers for small unmanned aerial vehicles. This table only lists companies that produce fuel cells for small UAVs. Among these, intelligent energy offers air-cooled fuel cells in three different capacities named IR-SOAR 800, IE-SOAR 1.2 and IE-SOAR 2.4. These fuel cell module masses (excluding tank) are 1,450, 2,700, and 4,800 grams, respectively. Also, they operate between −5 and 40 ${{\rm{\;}}^ \circ }$ C degrees and can robustly provide energy up to 3,000 m altitude. Using the IE-SOAR 800, a maximum flight time of 15.6 hours was achieved in an unmanned aerial vehicle with a total take-off mass of 7.4 kg. The operating voltage range of the IE-SOAR 800 is 24–48 V, while the other two fuel cells are 50–70 V. Another leading fuel cell manufacturer is Horizon Fuel Cell Technologies, and Its subsidiary for small UAVs is H3 dynamics, which manufactures Aerocell 300–4000 air-cooled PEM fuel cells in various power capacities. The masses of these fuel cells range from 0.72 kg to 6 kg. Their specific power and power density of them are 550 W/kg and 230 W/L, respectively. Combining these fuel cells with suitable batteries can create hybrid power systems. For example, Doosan is a company that offers such a hybrid system. Doosan offers powerpack products called DP30M2S, DM30M2S and DM 15, which include the air-cooled fuel cell and the battery. Their masses are 6.9 kg, 11.1 kg and 8 kg, respectively, and the first two of them are designed for rotary-wing UAVs and the third for fixed-wing UAVs. The rated powers of these power packs are 2.7 kW, 2.7 kW and 1.25 kW, respectively. Also, Doosen is studying liquid-cooled fuel cell systems. Finally, Honeywell has two water-cooled fuel cell packs with masses of 1.8 kg and 4 kg and continuous output powers of 600 W and 1,200 W.

Table 4. List of leading manufacturers of fuel cell-based power banks for UAVs

4.0 Hybrid topologies and components of small uavs

4.1 Hybrid topologies of fuel cell powered UAVs

The UAVs generally use three different types of fuel cells: PEMFC, SOFC and DMFC. Especially in small-type unmanned aerial vehicles, PEMFCs are mostly preferred due to their low operation temperature and relatively high energy density. The restrictions to implementing PEMFCs alone in a propulsion system are mentioned in the introduction section of this study. These restrictions can be addressed by creating hybrid power systems, where batteries serve as the secondary energy source and fuel cells serve as the main energy source. Although the literature contains examples of hybrid systems that combine fuel cells with solar cells and supercapacitors, fuel cells are typically used in conjunction with batteries. Solar cells are only suitable for unmanned aerial vehicles with large wing areas. The configurations of hybrid system topologies with multiple energy components in this way are summarised in detail in Refs [Reference Tao, Zhou, Zicun and Zhang44, Reference Wang, Zhao, Li, Wang, Huang, You and Becker65, Reference Çnar, Kandemir and Donateo69]. This study only considers hybrid power systems that include batteries and fuel cells. These hybrid system topologies are given in Fig.5. Also, Table 5 evaluates fuel cell-driven power systems and prominent literature examples (only experimental ones).

Figure 5. Fuel cell powered propulsion system topologies (a and c: passive control b, d, and e: active control).

Table 5. Comparison of fuel cell-based hybrid topologies and application examples

(a and c: passive control b, d, and e: active control).

Figure 6. Categorisation of electronic components for fixed and rotary wing UAVs.

In Fig.5, the topologies a and c drive directly the motor without any control. Whereas, in topologies b, d and e, active energy management is applied by controlling the energy components with the help of DC/DC converters. Although not included in Fig.5, there are hybrid system studies consisting of SOFC and TIG/TEG in the literature [Reference Ren94]. However, these studies are only simulation studies. This hybrid system may be handled in future studies.

4.2 Components

Based on their wing configuration, UAVs are divided into three primary types: fixed-wing, rotary-wing and VTOL systems. Despite differences in aerodynamic design and propulsion mechanisms, all three types contain fundamental electronic components essential for navigation, communication, energy management, and control.

The avionics of UAVs can be broadly classified into several subsystems, including flight control systems, computing, communication, sensory, power, navigation and positioning [Reference Osmani and Schulz110]. Figure6 categorises these components, while Fig.7 illustrates a typical configuration for a fixed-wing UAV with a single motor. Fixed-wing UAVs often utilise a single electric motor for propulsion, but multi-motor configurations are also employed to enhance redundancy, energy efficiency and manoeuvrability. In contrast, VTOL UAVs rely on multiple motors for hover and cruise, necessitating more complex motor control electronics and energy distribution systems.

Figure 7. Electronic equipment of a small fixed-wing UAV with a single motor.

Recent developments in embedded systems facilitate real-time energy-aware path planning, especially in hybrid propulsion systems that combine diverse energy sources like fuel cells, lithium-ion batteries, solar cells and supercapacitors [Reference Hashim111]. These systems depend on high-performance flight controllers, such as the Holybro Pixhawk 6X and Cube, which combine STM32H7 microcontrollers with fast processors able to calculate power consumption and the best way to distribute energy in real-time. With DC/DC converters, power sharing between energy sources in hybrid propulsion systems can be achieved (manufacturers including Vicor, Current Logic, Texas Instruments and Linear Technologies) [Reference Martinez-Heredia, Colodro, Mora-Jiménez, Remujo, Soriano and Esteban112]. Also, the combined operation of EMS and BMS, capable of continuously measuring voltage, current and temperature enables high energy efficiency.

Real-time energy-aware path planning relies on advanced sensors that measure flight dynamics characteristics, environmental conditions and power characteristics on the relevant platform to determine trajectories while aiming to minimise energy consumption. High-accuracy inertial measurement units (IMUs) such as VectorNav VN-200 combined with RTK-GNSS modules such as Septentrio’s AsteRx-m3 and Ublox ZED-F9P provide centimeter-level positioning accuracy that allows for precise flight path corrections depending on atmospheric and terrain conditions. Moreover, advanced wind speed and air pressure sensors can detect environmental conditions that can impact excessive power consumption. Optical and LiDAR-based systems enable terrain-adaptive path planning, reducing power consumption by dynamically adjusting altitude and route selection. To process these complex real-time energy-aware path planning calculations, UAVs integrate AI-driven edge computing platforms like the NVIDIA Jetson Xavier NX and Qualcomm RB5, which execute reinforcement learning algorithms to predict the most energy-efficient flight routes. This is critical for UAVs with hybrid propulsion systems, whereby AI-aided decision-making helps dynamically allocate power between primary and auxiliary energy sources, optimising endurance while maintaining mission performance.

5.0 Energy-aware path planning for uavs

The purpose of this section is twofold, i.e. while discussing on energy-aware path planning approaches, it includes insights on the relationship of such techniques with UAV configuration (e.g. fixed-wing, VTOL, etc.).

The path planning term does not only comprise finding the shortest and optimal path but also safe flight by considering collision-free and energy awareness. The key terms motion planning, trajectory planning, and navigation are also directly related to path planning [Reference Aggarwal and Kumar51]. All of these terms are a measure of autonomy for UAVs and should be considered together in fully autonomous systems [Reference De Ruiter and Owlia113]. The Refs [Reference ul Husnain, Mokhtar, Mohamed Shah, Dahari and Iwahashi56, Reference Sepulveda and Smith114] divide the autonomy of UAVs into four levels: remotely controlled systems, automated systems, autonomous non-learning systems and autonomous learning systems. An important factor in achieving completely autonomous UAV flights is path planning algorithms. They are generally classified as traditional, mathematical models, machine-learning-based, meta-heuristics and hybrid algorithms. Figure8 illustrates a general classification of the UAV path planning algorithms.

Figure 8. Classification of the path planning algorithms for UAVs (adapted from Refs [Reference Aggarwal and Kumar51, Reference ul Husnain, Mokhtar, Mohamed Shah, Dahari and Iwahashi56]).

Table 6. Comparison of energy-aware path planning studies for small UAV

^a Exp.: Experimental, Sim. : Simulation, HILS: Hardware-in-the-loop simulation

^b Remark for the column of components the listed review studies incorporate/compare: 1 - Collision avoidance, 2 - multi-UAV, 3 - Comparison of studies, 4 - multi-physics, and 5 - real-time applicable and reliability.

Figure 9. Structure of energy-aware path planning of a small UAV with a hybrid propulsion system.

Figure8 classifies established path planning approaches from the current literature, each of the approaches having its unique characteristics. The listed approaches are compared and evaluated in Refs [Reference Mannan, Obaidat, Mahmood, Ahmad and Ahmad54, Reference Ait Saadi, Soukane, Meraihi, Benmessaoud Gabis, Mirjalili and Ramdane-Cherif57, Reference Yahia and Mohammed58, Reference Luo, Tian and Wang61]. Traditional path planning algorithms, such as A* and Dijkstra, are easy to implement and provide optimal solution in only static environments with simple obstacles yet suffering from high computational costs and poor performance when employed for dynamic and complex environments [Reference Ait Saadi, Soukane, Meraihi, Benmessaoud Gabis, Mirjalili and Ramdane-Cherif57]. Meta-heuristic algorithms generally exhibit easy implementation and scalability for complex environments but lack guaranteed optimality assurance and require parameter tuning [Reference Hooshyar and Huang115]. Machine learning-based methods provide optimal and reasonable results for dynamic environments and dynamic changes, albeit ML methods require extensive training, normally high-computational burden, and use of appropriate datasets. The more (analytical) mathematical approaches, like mixed-integer linear programming (MILP) and MPC, exhibit effective control and robustness but can be difficult to implement and normally require a model that can be complex (to encapsulate the required system information) and struggle with real-time execution [Reference Luo, Tian and Wang61]. While meta-heuristic and ML-based approaches are more effective in dynamic, complex environments and when considering uncertainties, traditional and mathematical methods are best suited for static environments. In addition, the selection of suitable algorithms depends on several factors, including the UAV’s wing configuration, mission requirements, propulsion type, adaptability to dynamic conditions, computational requirements, real-time applicability. Hybrid methods, such as the combination of MPC with RL, can balance real-time adaptability with optimal energy efficiency [Reference Choi, Jimenez and Mavris116–Reference Shiri, Park and Bennis118].

The main topics in path planning of UAVs are as follows: optimum path length, optimality, completeness, cost-efficiency, time efficiency, energy efficiency, robustness and collision avoidance. Since the path planning of a UAV depends on its mission, one or more of these main topics may be considered together in a certain mission. Given that UAVs’ flight duration and range are restricted by the energy components’ limited capacity, energy-aware path planning is a crucial research topic. Consequently, the energy-aware path planning topic has attracted great attention in the literature. Many studies deal with this topic in battery-based small UAVs.

The energy-aware path planning studies of small UAVs are compared in Table 6. Studies of energy-aware path planning for small UAVs are usually simulation-based, as shown in Table 6. To our knowledge, no experimental energy-aware path planning study has been conducted for small UAVs powered by hydrogen, despite recent years seeing the development of experimental energy-aware path planning for battery-only small UAVs. Also, energy management and energy-aware path planning should be considered together for safe and efficient flight in hybrid propulsion-powered UAVs. Figure9 illustrates a coordinated control technique that considered together the energy-aware path planning and energy management for a hybrid propulsion-powered UAV. The connection between energy management and energy-aware path planning is the power consumption of the UAV. In simulation-based studies, the power consumption of UAVs can be determined by creating a multi-physics model (model-based method). Or, an approximate power consumption can be determined by considering the UAV’s flight speed. However, actual flight tests (data-driven approach) can be used to estimate UAV power consumption. After determining the UAV power consumption, an energy-aware path planning algorithm can be implemented considering energy management. Here, the energy-aware algorithm can minimise the power consumption of the UAV and also take into account the health status parameters of its propulsion system components. In other words, energy-aware path planning covers all issues related to the power and health status of the propulsion system.

Since fixed-wing, VTOL, and rotary-wing UAVs operate under distinct flight conditions, their energy-aware path-planning algorithms have different characteristics in terms of power consumption. Fixed-wing UAVs, benefiting from higher lift forces than rotary-wing UAVs due to their fixed wings, have greater flight endurance but require a steady altitude and speed, necessitating smooth trajectories without sudden manoeuver changes [Reference Ait Saadi, Soukane, Meraihi, Benmessaoud Gabis, Mirjalili and Ramdane-Cherif57, Reference Liao, Zhu, Wu, Shaikh, Sharaf and Mostafa143]. In contrast, rotary- wing UAVs are characterised by hovering and high manoeuverability, leading to greater energy demands, while tilt-wing and tilt-rotor VTOL UAVs incorporate complex transitioning mechanisms between vertical and horizontal flight [Reference Keller, Thakur, Likhachev, Gallier and Kumar144]. Given these variations in UAV types, energy-aware path planning should account for UAV-specific power consumption behaviour influenced by factors such as flight speed, altitude, manoeuverability, acceleration and trajectory requirement. For a fixed-wing UAV, path-planning algorithms should focus on optimising altitude, utilising gliding for energy efficiency, leveraging wind assistance, minimising sharp turns and optimising waypoints [Reference Jafari, Saeedi and Pishro-Nik145]. In rotary-wing UAVs, reducing hover time, mitigating wind resistance and optimising manoeuvering are essential, whereas in VTOL UAVs, efficient mode transitioning, minimising vertical flight phases and maximising cruise efficiency should be targetted. Since fuel cells have a lower power density, they may limit the UAV’s manoeuvering capabilities and require smoother trajectories. This highlights the necessity of a joint analysis of path planning and energy management for hydrogen-powered UAVs.

6.0 Challenges and recommendations for future work

The most popular research topics in the literature on unmanned aerial vehicle path planning include complex environments, multiple UAV communication, real-time path planning and hybrid path planning algorithms. The main challenge in real-time path planning is the computational burden of the algorithms. As hardware technology develops, the number of real-time path-planning applications in UAVs will increase. In general, Veroni diagram (VD), rapidly exploring random tree (RRT), Q learning algorithm and particle swarm optimisation (PSO) algorithms are used in real-time path planning because they provide low computational burden [Reference Vashisth, Batth and Ward146]. Since many missions, such as surveillance and inspection, require multiple UAVs, the path planning topic for multiple UAVs is important [Reference Shanmugavel, Tsourdos, White and Żbikowski147]. The multiple UAVs can be homogeneous or heterogeneous. Path-planning based on time or energy minimisation and communication of UAVs are among the trending research topics in multiple UAV path planning. Also, the complex environment is one of the difficulties in the path planning of UAVs [Reference Zhao148]. Although some simulation studies use fixed obstacles and 2D environments to test path-planning algorithms, the research challenges are on dynamic obstacles and 4D (time, x, y, z) environments. Hybrid path-planning algorithms have been developed in recent years to improve the computational burden, robustness, convergence and completion properties of path algorithms [Reference Zhao, Huo, Li, Yu and Qi149]. Especially, new hybrid algorithms have been developed using meta-heuristics algorithms and machine learning algorithms [Reference Yu and Luo150, Reference Tutsoy, Asadi, Ahmadi, Nabavi-Chashmi and Iqbal151].

The trend research topics and future work recommendations briefly summarised above also are valid for the topic of energy-aware path-planning of UAVs, as they take into account the general path planning of UAVs. In addition, the specific future work recommendations for energy-aware path planning of UAVs are given below:

• • Considering multi-physics: In simulation-based energy-aware path planning studies of UAVs, the power consumption of the UAV is carried out either by the speed of the vehicle or by simple equations. To the best of the authors’ knowledge, only few studies considered energy-aware path-planning have considered the multi-physics of the UAV by modelling the dynamics of the battery, ESC, propeller and rigid body of the UAV in the literature [Reference Michel, Wei, Kong and Lin122, Reference Michel, Kong and Lin152, Reference Michel, Wei, Kong, Sinha and Lin153]. In particular, considering the multi-physics in energy-aware path planning studies of fuel cell-powered UAVs are even more important (due to the nonlinear structure of fuel cells), and it can be said that new studies are needed in the literature on this topic.

• • Considering energy management: Although various studies consider the energy management of hydrogen-propelled UAVs, to the authors’ best knowledge, only four studies in the current literature address energy management and path planning together [Reference Tian, Zhang and Yang48, Reference Yao, Wang, Zhou, Li, Liu and Li49, Reference Tian, Liu, Zhang, Shao and Ge154, Reference Tian, Liu, Zhang, Shao and Ge155]. In this review study the path planning of UAVs with hybrid propulsion systems supports the view of considering these simultaneously with energy management to facilitate more efficient and sustainable operations. Furthermore, the path planning of UAVs with hybrid propulsion systems has not received much attention in the literature. Therefore, energy management and energy-aware path planning of UAVs with hybrid propulsion systems should be considered together to guarantee safer and more efficient flights.

• • Considering multi-cost function: The path planning studies mostly consider one objective. These objectives may be energy efficiency, time, shortest path length, etc. For example, energy-aware path planning studies are given in Table 6. Similarly, some studies consider the minimisation of time and path length. Depending on the mission of the UAV, it may be necessary to develop path-planning algorithms that take several of these into account. Therefore, it can be said that according to the mission of the UAV, path-planning algorithms that take into account multiple objectives should be developed.

• • Considering altitude’s effect on the performance of the fuel cells: The changes in air temperature and pressure with altitude change directly affect the performance of the fuel cells [Reference Gong, Xing, Liang and Tu41, Reference Pratt, Brouwer and Samuelsen156]. Therefore, the altitude effect should be included in the energy-aware path planning of fuel-cell-powered UAVs. For this reason, the fuel cell performance due to altitude change should be examined, and an energy-aware path planning algorithm should be developed accordingly.

• • Considering the aging of energy components: Although many studies discuss the estimation, modelling and monitoring of parameters such as RUL and SOH of batteries or fuel cells in UAVs, very few studies consider these health indicators together with path planning [Reference Schacht-Rodrguez, Ponsart, Garca-Beltrán, Astorga-Zaragoza, Theilliol and Zhang157–Reference Rodriguez159]. Especially in UAM vehicles that perform delivery and transportation, continuous monitoring of SOH and RUL state parameters is important for a safe and efficient flight [Reference Mitici, Hennink, Pavel and Dong160]. In this context, SOH and RUL parameters of propulsion system energy components (battery or fuel cell) should be included in energy-aware path-planning studies of UAVs.

• • Structural complexity of hybrid systems: Although low hydrogen has a volumetric energy density, onboard hydrogen storage has challenges with aerodynamic efficiency and take-off mass [Reference Massaro, Pramotton, Marocco, Monteverde and Santarelli161]. Aerodynamic efficiency relates to the lift-drag ratio. The on-board hydrogen storage effect has negative effects on the aerodynamic efficiency of UAVs because it decreases the lift-drag ratio. Therefore, aerodynamically efficient designs of rotary-wing hydrogen-powered UAVs contain a research gap. Additionally, considering the time-varying wind and speed, aerodynamically efficient path-planning algorithms need to be developed [Reference Dobrokhodov, Jones, Walton and Kaminer162]. In fixed-wing UAVs, the hydrogen tank is placed inside the fuselage and is therefore less affected by aerodynamics than in rotary-wing UAVs. However, the air intake design for fuel cell ventilation must be taken into account [Reference Özbek, Yalin, Ekici and Karakoc108]. As a result, there is a need for designs that address aerodynamic efficiency and path planning together in both UAV types [Reference Apeland163].