Development of a Firefighting Drone for Constructing Fire-breaks to Suppress Nascent Low-Intensity Fires

Abstract

Korean forests are highly vulnerable to forest fires, which can severely damage property and human life. This necessitates the establishment of a rapid response system and the construction of firebreaks to prevent the spread of fires and protect key facilities. The existing firebreak construction methods can be classified into prevention- and response-stage methods. In the prevention stage, the progression and spread of fire are delayed, while in the response stage, primitive manual methods involving tools such as hooks are used, in addition to aerial deployment of water and fire retardants through helicopters. Herein, we propose the use of “fire-extinguishing drones” for firebreak construction during the initial, low-intensity stage of a fire before the deployment of firefighting personnel. We implement a continuous fire-extinguishing module capable of carrying six fire-extinguishing balls to verify its deployment accuracy and stable hovering capabilities. Through the operation of multiple drones using a ground control system and real-time kinematics to precisely generate designated automatic flight paths, we conducted experiments to assess the feasibility of firebreak construction by using fire-extinguishing drones to prevent the spread of wildfires. A firebreak construction field test was conducted to evaluate the accuracy of continuous fire extinguisher deployment, hovering performance during deployment, accuracy of the RTK-designated paths, and GCS performance. The proposed system achieved 100% performance on all indicators, except the accuracy of the RTK-designated paths.

1. Introduction

1.1. Background and Objectives of the Study

This study focuses on the prevention of grass fires (wildfires) and the response to such fires by utilizing drones. In South Korea, 62.6% of the total land is covered with forests, and coniferous forests, which are composed primarily of pine trees, account for 36.9%. Therefore, the region is highly susceptible to wildfires [1].

Over the past decade, on average, 481 wildfires have occurred in the country, resulting in the loss of 1087 hectares (per year) of forested land [2]. Grass fires, caused primarily by negligent activities, have been identified as the main cause. In the last five years alone, 6538 grass fires occurred, resulting in 234 casualties, and 95% of these were attributed to negligence [3].

Particularly alarming is the fact that 27% of the grass fires originated from activities such as the burning of trash or agricultural residues that spread into mountainous areas [2]. The impact was more severe when these fires occurred near residential areas and mountainous border zones. This study aims to investigate the use of drones for deploying swift responses and effective preventive measures in the early stages of grass fire outbreaks to ultimately minimize the resulting damage.

Because it is practically impossible to completely prevent wildfires, a prompt and systematic response system is required to limit the damage caused by them. Early suppression of wildfires and the establishment of firebreaks are crucial response strategies for preventing the spread of wildfires; however, the lack of personnel and equipment hinders the construction of firebreaks for wildfire containment, particularly in large mountainous areas, which are prone to wildfires. Therefore, the use of drones in the initial stages of firebreak construction can help overcome these limitations and compensate for the lack of firefighting personnel [4]. In this study, we aim to study the effectiveness and impacts of using drones in the initial stages of firebreak construction.

1.2. Study Methodology and Procedures

We present a prototype of a firefighting drone that can be used to construct firebreaks to prevent the initial spread of grass fires. This prototype is developed and validated experimentally. As the first step, the existing literature and case studies related to field fire suppression techniques and firebreak construction are reviewed. Continuous fire-extinguishing ball deployment, as well as the relevant case studies and experiments, are investigated. In the second step, focus group interviews (FGI) are conducted with three firefighting experts and three drone specialists to evaluate the feasibility of utilizing drones to continuously deploy fire-extinguishing balls for firebreak construction. In these interviews, the design of the proposed continuous fire-extinguishing ball-deployment module is evaluated. In the third step, a prototype based on the evaluated design is fabricated. In the fourth step, experiments are conducted to assess the accuracy of the fire-extinguishing ball-deployment module and automated flight of a ground control system (GCS)-guided drone equipped with the aforementioned module to predetermined target points. Multiple drone operations are executed, and real-time kinematic (RTK) technology is utilized to field test firebreak construction and evaluate the accuracy of the proposed fire-extinguishing ball-deployment module. These tests validate the feasibility of firebreak construction for containing field fires. The research procedures and methods are summarized in

Table 1. Study flow.

Introduction
1. The Literature Review and Case Studies	Review of the literature and case studies related to the use of fire-extinguisher drones and wildfire firebreak construction.
2. FGI Expert Survey	- Phase 1: interviews with experts to ascertain the functionality and potential of using fire-extinguisher drones for firebreak construction. - Phase 2: evaluation and review of the design of fire-extinguisher drones.
3. Equipment Production	Production of a consecutive fire extinguisher deployment module by considering expert opinions.
4. Experimental Validation	- Experimental tests to determine the accuracy of fire extinguisher deployment and autonomous drone flight to target locations. - Verification of the accuracy of firebreak construction and effectiveness of fire suppression using drones.
Conclusion	Achievements, limitations, and future research directions.

.

2. Theoretical Background

2.1. Issues in Wildfire Management and Firebreak Construction

Wildfire management comprises two important stages: prevention and response (firebreak construction). In the prevention stage, the risk of wildfire occurrence can be reduced, and effective responses to initial outbreaks can be provided by establishing fire prevention lines (i.e., firebreaks) [5]. These lines act as obstacles along the spread path of a wildfire, thereby delaying its propagation and facilitating efficient firefighting operations [6].

In the response stage, firebreaks are constructed by deploying personnel and helicopters. According to Article 2 of “Integrated Regulations for Wildfire Management”, firebreak construction with manual labor involves establishing firebreaks by utilizing natural terrain features (such as rivers and rocks) or employing methods such as clearing undergrowth, removing leaf litter, and digging trenches to create fire containment lines [7].

However, firefighting and prevention efforts that rely on human labor have physical limitations and raise safety concerns [4]. Although the safety of firefighting personnel has been evaluated as the most crucial element of the firefighting response, its importance has been ranked the lowest [8]. The use of helicopters for constructing firebreaks is highly effective for safeguarding residential areas and critical infrastructure [9]; however, aerial firefighting using helicopters is perceived as ineffective during nighttime, and helicopter pilots have reported experiencing hazardous situations during aerial firefighting operations [10].

The authors of this study aim to establish firebreaks for suppressing low-intensity wildfires caused by surface fires and spreading ground fires, rather than high-intensity fires. Wildfires tend to spread primarily from grass fires, and it is crucial to establish effective firebreaks to prevent and mitigate large-scale wildfires. Grass fires act as a major catalyst for high-intensity wildfires, and without proper control, wildfires are highly likely to spread expansively. Therefore, setting up appropriate firebreaks in advance is essential for effectively controlling the spread of high-intensity wildfires and preventing their occurrence.

However, firefighting activities in areas affected by wildfires present two main challenges. First, the burning of farmland and crop residues by farmers can escalate into wildfires or large-scale fires [11]. Despite the awareness that the burning of fields is a major cause of wildfires, farmers continue this practice [12]. Second, the areas in which fires occur are typically narrow and situated between rice paddies, which constrains the entry of fire trucks. An evaluation of the road conditions within rural communities with a focus on residents revealed a need for “increasing the road width to facilitate the passage of emergency vehicles such as firetrucks and ambulances” [13]. This signifies the necessity of increasing road width, especially in areas that are prone to wildfires and located between rice paddies, to facilitate easier access for emergency vehicles, including firetrucks and ambulances.

2.2. Case Studies on Fire-Extinguishing Drones

Drones significantly alleviate the aforementioned temporal and spatial constraints, which increases their utility and necessity. Both domestically and internationally, drone usage is being researched extensively in various fields, including exploration and disaster and fire monitoring. However, there is a relative lack of direct research specifically related to firefighting and fire response [14]. This highlights the need for research on the use of drones for fire suppression and response.

To increase the efficiency of wildfire response, Korea Forest Service has recently partially researched developed drones equipped with conventional, cylindrical, single-shot fire extinguishers as prototypes [15]. These drones are designed to respond swiftly to wildfires, suppress the spread of fires, and facilitate effective firefighting; moreover, they are effective even in areas that are inaccessible by ground vehicles and areas where roads are not readily available. The stability and effectiveness of these fire-extinguishing drone systems have been validated by conducting practical experiments [15].

Table 2. Summary of case studies related to fire-extinguisher drone research.

Study	Image	Fire Extinguisher Deployment Method/Fire Extinguisher Payload/Status of Experimental Validation
Abdel Ilas N. Alshbatat (2018) [16]		- To address the difficulties and risks associated with firefighting in high-rise buildings or mountainous terrains, a safe fire-extinguishing system consisting of six components (drone, robot, firefighting equipment, fire-extinguishing ball, collision-avoidance system, and camera) was proposed. - The system was demonstrated to be suitable for initial fire suppression, and the experimental results proved its ability to operate safely at heights where fire-extinguishing balls can be reloaded while maintaining drone stability.
Burchan Aydin et al. (2019) [17]		- The potential utility of fire-extinguishing balls in the proposed firefighting system, which complements the traditional firefighting methods using drones and remote-sensing technology, was examined. - The experimental results demonstrated that small-sized fire-extinguishing balls may not be effective against building fires but they can be effective for suppressing grass fires.
Rupali Patil et al. (2020) [18]		- A holder was designed to store fire-extinguishing balls for firefighting purposes. - The holder consisted of a railing system attached to a base plate composed of an acrylic sheet. - Using Arduino and servo motors, the release and blocking of the balls were controlled, while the rails guided the balls, and the barriers released and stopped the balls.
Nastaran Reza Nazar Zadeh et al. (2021) [19]		- A system that uses unmanned aerial vehicles (UAVs) for firefighting was proposed to facilitate quick, efficient access to the top floors of buildings and to improve firefighters’ visual access to affected areas. In this system, fire-extinguishing balls were used to prevent the spread of fire. - This system was based on a newly designed firefighting UAV equipped with a mechanism to shoot and drop fire-extinguishing balls. The UAV was controlled using a 2.4 GHz radio frequency signal and video monitoring controller to operationalize the shooting and dropping mechanism. - This firefighting UAV can be utilized effectively in the early stages of a fire, and its ability to enhance the safety and efficiency of firefighters was demonstrated.

summarizes the research conducted in [16], wherein the potential effectiveness of fire-extinguishing drones in outdoor locations was verified. In [17], experiments were conducted using fire-extinguishing balls to prove their efficacy in initial firefighting and reloading operations. These two studies validated the efficiency of using fire extinguishers in the early stages of wildfire suppression. In [18], a holder was designed for drones to store fire extinguishers using a railing-based system and releasing fire-extinguishing balls. This railing system can store and release multiple fire-suppressant balls in a controlled manner. In [19], a design that allows for the use of both launch and drop methods to deploy fire-suppressant balls was proposed.

These studies proved that fire-extinguishing drones are effective for rapidly responding to wildfires in mountainous areas and validated the efficiency of various fire extinguisher deployment methods. Proactive wildfire prevention and response can possibly be realized through continuous, controlled deployment of fire-suppressant balls. While research on single-shot fire-extinguisher deployment drones is underway overseas, studies on the reliability of the drones designed for continuous fire-extinguisher deployment are lacking. In this study, we implement six fire-extinguisher canisters and conduct practical research on firebreak construction through continuous fire extinguisher deployment to prevent the spread of fires.

2.3. Investigation of Fire-Extinguishing Ball

Fire-extinguishing balls are made of Styrofoam and can be transported easily using drones. Currently, Elide and AFO are the two main commercially available brands of fire-extinguishing balls. Elide’s product has a diameter of 15.5 cm, weight of 1.4 ± 0.1 kg, and firefighting coverage area of 2.5 m³. Meanwhile, AFO’s product has a diameter of 15.2 cm, weight of 1.3 ± 0.2 kg, and firefighting coverage area of 3 m³. Both brands claim that their products are effective against various types of fires, including general fires (Class A), fuel and gas fires (Class B), and electrical fires (Class C) [20,21]. However, specific information on their effectiveness in extinguishing wildfires is not provided. Based on its weight and coverage area, AFO’s product can be considered more effective for drone-assisted wildfire suppression. According to Aydin et al. [17], this product is particularly effective against grass fires or in areas with simple vegetation.

2.4. Enhancement of Fire Response Messures Using Cluster Flight System

Setting an appropriate flight path is crucial for establishing firebreaks in areas affected by wildfires. In this process, it is essential to equip fire-extinguishing drones with RTK technology, a real-time positioning method, to realize accurate location measurement and improve the performance of fire extinguisher deployment. Fire-extinguishing drones are used to deploy multiple fire extinguishers for establishing firebreaks and preventing the spread of wildfires. RTK-guided drones can deploy fire extinguishers accurately, which increases the efficiency and precision of firebreak construction. RTK technology is an extension of the Global Navigation Satellite System (GNSS) (a system that receives signals from satellites to determine the position of the receiver). The GNSS determines receiver positions by receiving signals from satellites but various factors such as atmospheric errors and errors in satellite elevation angles reduce its accuracy.

The precision of RTK technology is higher than that of basic GNSS positioning. The GNSS measures a position by receiving satellite signals, and its accuracy is in the order of meters because of atmospheric errors and errors in satellite elevation angles. Meanwhile, RTK technology calculates and corrects these errors by using real-time GNSS data received from a fixed reference point (base station), and its position measurement accuracy is in the order of centimeters for a moving receiver. Therefore, RTK technology is suitable for applications requiring enhanced accuracy, such as aerial photography [22].

Furthermore, it is challenging to deploy prevention and response measures effectively by operating a single drone over a large area covered with wildfires. Therefore, a cluster-flight system with high autonomy and cooperative capabilities is needed. An outdoor group-flight system based on RTK-GPS, which provides precise location information at the centimeter level, can fulfill these requirements [23]. The RTK-GPS system enhances the flight and operational performances of drone clusters, which demonstrates its potential for use in various UAV applications. Therefore, the RTK-GPS-guided cluster-flight system is expected to play a crucial role in the advancement of spatial information and drone-based operations, which can greatly enhance the deployment of effective response measures in situations such as wildfires.

3. Prototype Development of Fire-Extinguishing Drone

3.1. Demand Survey through Focus Group Interviews

Based on the item composition framework derived from the preliminary research process, FGIs were conducted with firefighting experts and drone specialists. In total, three firefighters, each with more than 20 years of firefighting experience, and three drone experts were selected, and the interviews were conducted on 12 September and 23 October 2021, respectively (

Table 3. Focus group interview (FGI) participants.

Professional Firefighter			Drone Engineer
Lee	68	Fire Chief	Nam	54	Drone Manufacturing Engineer
Bak	53	Wildfire Safety Expert	Bak	49	Drone Pilot and Mapping Expert
Song	65	Fire Chief	Lee	46	Drone Communication Expert

). Various opinions pertaining to the considerations and functionalities relevant to the design of the “Fire-Extinguishing Drone Prototype” were collected. A prototype drone was fabricated based on the inputs obtained from these interviews, which are summarized in

Table 4. FGI results.

	Key Issues from Expert Interviews
Issues Pertaining to Fire-Extinguishing Balls	1. Fire-extinguishing balls are ineffective for directly suppressing and preventing wildfires because they can collide with terrain features, such as rocks and trees, which can cause them to deviate from their intended drop locations; 2. The currently available fire-extinguishing balls are suitable for small-scale fires or scenarios with low-intensity fires; 3. It is necessary to increase the fire-suppression range of the existing fire-extinguishing balls and reduce their weight; 4. To prevent rolling, fire-extinguishing balls with alternative forms (e.g., hemispherical, cuboid) should be developed.
Challenges in Constructing Firebreaks	1. A large fleet of at least several dozen drones is required to construct firebreaks; 2. The use of drones for constructing firebreaks can be beneficial for protecting military facilities, rural villages near fire-prone areas, and cultural heritage sites; 3. From a safety perspective, using drones to construct firebreaks is more effective than using them directly for firefighting; 4. Firebreaks can be constructed easily on flat or low terrains but it is challenging to construct them in mountainous terrains.
Drone Functionality Improvements	1. Reconnaissance drones are needed to gather information about fires; 2. Overcoming obstacles related to inter-drone communication at various distances is important; 3. Adequate flight capabilities are required to reach wildfire-affected areas; 4. Autonomous flight capabilities through a GCS are needed; 5. Achieving high GPS positioning accuracy by using real-time kinematics (RTK) is crucial; 6. There is a risk of drone loss or crash owing to updrafts during drone flights over mountainous fire zones; 7. The use of coordinates to set drop points for fire-extinguishing balls and the use of downward-facing cameras are necessary; 8. Accurate information about fire locations and precise targeting of drop points are crucial; 9. Drone specifications must be determined logically by considering the weight and number of fire-extinguishing projectiles.

.

3.2. Verification and Discussion of Practical Scenarios through FGIs

The following diagram (Figure 1) illustrates the approach and problem-solving process for the issues outlined in the FGIs. The drone that carries fire-extinguishing balls is primarily meant for constructing firebreaks to suppress grass and marsh fires rather than fighting forest fires. This drone operates effectively on gently sloping fields. The use of vehicles to launch drones enhances the accessibility of drones to field fires and secures their flight capabilities ④.

Conceptually, the initial reconnaissance drone ⑤ collects information ⑥ about the location and progression of a grass or marsh fire. In the information-collection phase, the drone receives flight paths and coordinates for navigation ⑦. To prevent the rolling of a fire-extinguishing ball upon contact with the ground, the ball should split into two hemispheres upon contact with the ground ⑧. Based on the information obtained from the reconnaissance drone, the length and thickness of the fire line constructed through clustered drone flights are determined to ensure its effectiveness ⑨.

3.3. Development of Prototype Fire-Extinguishing Drone Based on FGIs

To prevent the spread of field fires, it is necessary to establish accurate flight paths and construct firebreaks by deploying multiple fire-extinguishing balls. In this study, a prototype of the proposed fire-extinguishing drone is fabricated based on the opinions obtained from the FGIs, the contents of which are summarized in Table 4. As depicted in Figure 2, the proposed fire-extinguishing module is composed of a Y-shaped frame that accommodates six fire-extinguishing canisters, that is a total payload of 7.8 kg (one canister weighs 1.3 kg). The Y-shaped module is suitable for the continuous deployment of fire-extinguishing balls, and it allows for precise targeting because each canister is released individually from the central lower part of the drone. Moreover, the module facilitates the use of downward-facing cameras. Furthermore, to maintain drone balance during canister deployment, hovering technology is used. The developed drone has enhanced capabilities that help it traverse mountainous terrains. To realize these capabilities, GCS and RTK technology are used to achieve automatic flight along precise routes and cluster flights. The drone frame is composed of lightweight carbon onyx material to reduce its weight. Additionally, the drone is equipped with multi-spectrum cameras and thermal cameras for surveying disaster- and fire-affected areas. The GCS device, which comprises a mini-PC and a monitor, improves the accuracy of drone by using Mission Planner 1.3.8 software for automated flight-path generation (Figure 3).

4. Field Experiment of Fire-Extinguishing Drone

4.1. Experiment to Test the Accuracy of Fire Extinguisher Deployment (First Experiment)

4.1.1. Accuracy Performance Test and Method of Fire Extinguisher Deployment at Different Altitudes

The accuracy of fire extinguisher deployment was tested experimentally in an open field at 666-2 Choji-dong, Danwon-gu, Ansan-si, Gyeonggi-do, Republic of Korea. On the day of the experiment, 10 December 2021, the weather conditions were cloudy with a wind speed of 4 m/s, which was deemed suitable for the field experiment. In Experiment 1, the accuracy of the fire-extinguisher deployment was tested at different altitudes, and a wooden box measuring 1.44 m² (=1.2 m × 1.2 m) was mounted on the drone for deploying the fire-extinguishing balls. At each altitude (10 m, 20 m, 30 m, 40 m, and 50 m), two fire-extinguishing balls were loaded into the aforementioned wooden box (1.44 m²).

4.1.2. Experimental Objectives

The objectives of the experiment were to ensure the stability of drone flight during the deployment of fire-extinguishing balls, verify the functionality of the continuous-deployment fire-extinguisher module, and conduct accuracy experiments to determine the optimal altitude for deploying the fire-extinguishing balls.

4.1.3. Experimental Plan

The fire-extinguishing balls were aimed at the target within the fire-suppression radius, which was calculated based on the drone’s flight altitude. A score of 10 points was awarded if the fire-extinguishing balls landed inside the wooden box, 8 points if they landed within the target area of 12.96 m² (=3.6 m × 3.6 m), and 6 points if they landed outside the target area. The average score was used to evaluate accuracy, with 10 points indicating high accuracy, 9 points indicating moderate accuracy, and scores below 9 points indicating low accuracy.

4.1.4. Experimental Results

The experimental accuracy results are depicted in Figure 4 and summarized in

Table 5. Accuracy of drop points based on altitude.

Performance Objectives	Performance Indicators	Quantity	Target Value		Average Score	Grade
			Inside Box (10)	Outside Box (10)
Accuracy by Drop Altitude	High altitude (10 m) drop (1 m/s or less)	2EA	20	-	10	High
	High altitude (20 m) drop (1 m/s or less)	2EA	10	8	9	Medium
	High altitude (30 m) drop (1 m/s or less)	2EA	8	6	7	Low
	High altitude (40 m) drop (1 m/s or less)	2EA	8	8	8	Low
	High altitude (50 m) drop (1 m/s or less)	2EA	8	8	8	Low
	10–50 m	10EA	6EA	4EA	8.4

.

The stability evaluation results are presented in

Table 6. Flight Stability and Deployment Accuracy Results.

	Performance Indicators	Target Value	Validation Method	Results
1	Drop accuracy	Altitude-specific drop tests	On-site evaluation	80%
2	Drone balance during drop	Fixed hovering	On-site evaluation	100%
3	Operation of drop module	Six fire-extinguisher modules	Functioning	100%
4	LTE video transmission	Real-time transmission	On-site evaluation	100%

.

Drone stability was secured by attaching the fire-extinguishing module and maintaining the designated flight altitude, which allowed for the successful completion of continuous fire-extinguishing ball-deployment missions. According to the experimental results, which are depicted in Figure 4 and summarized in Table 5, the deployment accuracy was highest at the altitude of 10 m and decreased with increasing altitude. However, according to Table 6, flight stability was maintained regardless of the deployment accuracy. During the experiments, features such as fixed hovering, module operation, and LTE video transmission functioned perfectly.

4.2. Firebreak Construction and Fire Suppression Experiment (Second Experiment)

4.2.1. Experimental Objective

Normal operation of the fire-extinguishing drone was verified by navigating it along the designated path to the location of the fire outbreak and inputting the points for deploying the fire-extinguishing balls. The effectiveness of the first firebreak construction (Drone 1) and second firebreak construction (Drone 2) at the designated coordinates using the fire-extinguishing drone was tested. Subsequently, a fire was simulated to assess the effectiveness of the constructed firebreaks.

4.2.2. Experimental Plan

As illustrated in Figure 5, the flight path to the fire area, deployment path of the fire-extinguishing drone, and landing site were planned. The experimental scenario for verifying firebreak construction is described in

Table 7. Scenario of demonstration experiment.

	Scenario of Demonstration Experiment
1	Fire incident near Ansan-Wastadium Development Area;
2	Transmission of fire location and confirmation of drone flight path;
3	Although Ansan-Wastadium and the fire location are located along a straight line, the presence of the Central Danwon District Office in between necessitated a route change;
4	Confirmation of the risk of fire spreading to nearby residential areas, which necessitated the construction of firebreak lines;
5	Inputting drop locations of fire-extinguishing balls using GCS equipment;
6	Sequential deployment of two fire-extinguishing drones (drop-type) stationed at Ansan-Wastadium;
7	Arrival at the fire scene approximately 2 min later at a distance of 500 m;
8	Accurate route navigation and assessment of on-site conditions using the RTK equipment and camera mounted on the drone;
9	Upon arrival at the scene, Drone 1 drops six fire-extinguishing balls to establish the first firebreak line;
10	Upon arrival at the scene, Drone 2 drops six fire-extinguishing balls to establish the second firebreak line;
11	Fire trucks and firefighters present at the scene directly extinguish the ignition point.

.

4.2.3. Performance Metrics Setup

The performance metrics were set, as summarized in

Table 8. Performance metrics.

	Performance Metric	Performance Result and Target	Basis for Setting Targets	Measurement Formula
1	Flight accuracy	Accuracy in reaching the mission location	Flight accuracy with RTK-GPS	Error range within approximately 10 m
2	Firebreak construction	Construction of first and second firebreaks	Six fire-extinguishing balls per drone	3 m × 6EA (1.3 kg) ≒ 18 m firebreak construction
3	Approach to fire suppression	Fire suppression through the first firebreak	Verification of remaining fire

. To ensure safety during the experiment, we aimed to utilize the drone for deploying the fire-extinguishing balls within a high-accuracy range of 10 m to establish the first and second firebreaks. Ignition was carried out along the projected fire path, and when the fire approached the firebreak, the deployed fire-extinguishing balls were detonated to suppress the fire.

4.2.4. Experimental Method

In the firebreak construction experiment conducted to prevent the spread of fire, two drones equipped with fire-extinguishing balls were used. As illustrated in Figure 5, the GCS was used to generate the automatic flight paths of multiple drones and verify mission routes. At the outset, the first drone was operated along a 500-m orbit to establish the first firebreak by deploying six fire-extinguishing balls at 1.5-m intervals. Then, after a 3-min interval, the second drone was operated to establish the second firebreak by deploying six fire-extinguishing balls at 1.5-m intervals in the designated area. The construction processes of the first and second firebreaks are illustrated in Figure 6.

4.2.5. Experimental Results

As summarized in

Table 9. Performance results based on performance metrics.

Performance Indicators	Target	Validation Method	Results
Drop accuracy	Deployment of 12 fire-extinguishing balls along the defense lines (Drone 1: 6 balls/Drone 2: 6 balls)	On-site evaluation	80%
Drone balance during deployment	Stable hovering	On-site evaluation	100%
Functionality of deployment module	Activation of six fire-extinguishing balls (1st line of defense)	Operational	100%
LTE video transmission	Real-time streaming	On-site evaluation	100%
RTKs	Accuracy of path	On-site evaluation	90%
GCS	Ground control system	On-site evaluation	100%

, an analysis of the experimental results confirmed that the first and second drones successfully deployed six fire-extinguishing balls consecutively. The GCS operated normally, and the hovering function maintained the drones’ balance during the deployment of the fire-extinguishing balls. The accuracy of the RTK-generated paths was 90%, and the error was less than 10 cm. Therefore, the RTK-guided drones outperformed the GPS-guided drones; however, there was some variability depending on the weather conditions and solar status.

During the fire experiment, the six fire-extinguishing balls deployed in the first firebreak activated successfully, resulting in 100% fire suppression. However, a few issues related to firebreak continuity were observed. As illustrated in Figure 7 and Figure 8, the spherical fire-extinguishing balls tended to roll upon deployment, which changed their initial positions. To address this problem, design improvements or system modifications to these fire-extinguishing balls are necessary. For instance, designing the balls such that they split into hemispheres upon impact with the ground could reduce the aforementioned positional shifts and potentially enhance the effectiveness of fire suppression.

5. Conclusions and Limitations

Wildfires occur frequently every year, and they cause severe loss of life and property. Therefore, firebreak construction in the field is necessary to prevent the spread of wildfires and protect residential and other critical facilities. Large wildfires often stem from activities such as field burning or crop residue burning. These ignition sources tend to spread rapidly, thus emphasizing the importance of early identification of fires and deployment of measures to prevent their escalation into large-scale wildfires. Proactive responses to ignition sources such as field burning contribute not only to wildfire prevention but also to the safeguarding of residential areas and critical infrastructure, thereby enhancing resource efficiency in firefighting efforts. However, the existing methods rely on manual labor using primitive tools such as hooks, which limits their efficiency and rapid-response capabilities.

In this paper, we proposed a fire-extinguishing drone to prevent or delay the spread of wildfires in the initial outbreak stages before the deployment of ground firefighting teams. We implemented a fire-extinguishing drone module capable of deploying six fire-extinguishing balls consecutively. Additionally, we demonstrated the effectiveness of controlling multiple drones using a GCS and the accuracy of the precise designated paths generated using RTK technology for wildfire containment.

We conducted several experiments to evaluate the deployment accuracy of fire-extinguishing balls. According to the results, the accuracy of deployment was 100% within a 10-m target range, and the deployment accuracy was 80% for heights of 20 meters or less. A firebreak construction field test was conducted to evaluate the accuracy of continuous fire extinguisher deployment, hovering performance during deployment, accuracy of the RTK-designated paths, and GCS performance. The proposed system achieved 100% performance on all indicators, except for the accuracy of the RTK-designated paths.

Continuous deployment of fire-extinguishing balls and firebreak construction proved to be effective for securing critical facilities and protecting designated areas. It facilitated early firebreak construction, thereby improving community safety, wildfire prevention, and operational stability during firefighting efforts. We anticipate that by using multiple units of the fire-extinguishing drones developed herein in tandem, the impact of fires can be minimized by mounting a rapid response, including measures to prevent the spread of wildfires and protect residential and other critical facilities.

However, a few issues emerged, such as the tendency of the spherical fire-extinguishing balls to roll upon deployment, which changed their initial positions. To address this problem, the design of these fire-extinguishing balls must be improved. Moreover, we could consider using methods such as splitting the spherical fire-extinguishing balls into hemispheres upon impact with the ground to reduce the aforementioned positional shifts and increase the effectiveness of fire suppression.