A Long-Range and Low-Cost Emergency Radio Beacon for Small Drones

Abstract

The increasing use of unmanned aerial vehicles (UAVs) in the commercial and recreational sectors has led to a heightened demand for effective recovery solutions after a crash, particularly for lightweight drones. This paper presents the development of a long-range and low-cost emergency radio beacon designed specifically for small UAVs. Unlike traditional emergency locator transmitters (ELTs), our proposed beacon addresses the unique needs of UAVs by reducing size, weight, and cost, while maximizing range and power efficiency. The device utilizes a global system for mobile (GSM)-based communication module to transmit location data via short message service (SMS), eliminating the need for specialized receivers and expanding the operational range even in obstacle-rich environments. Additionally, a built-in global navigation satellite system (GNSS) receiver provides precise coordinates, activated only upon impact detection through an accelerometer, thereby saving power during normal operations. Experimental tests confirm the extended range, high precision, and compatibility of the prototype with common mobile networks. Cost-effective and easy to use, this beacon improves UAV recovery efforts by providing reliable localization data to users in real time, thus safeguarding the UAV investment.

1. Introduction

In recent years, the commercial and civilian use of unmanned aerial vehicles (UAVs) has increased noticeably. Nowadays, the introduction of drones into national airspace systems has brought operational challenges, including safe and secure integration. Despite these challenges, the drone sector holds enormous promise. In particular, in the United States, the Federal Aviation Administration forecasts that the commercial drone sector would include approximately 955,000 small drones in 2027 and the recreational small drone fleet UAV will probably reach 1.82 million units by 2027 [1].

Commercial drones have an average unit price assumed to be between USD 2500 and USD 25,000. Recreational drones have an average unit price of USD 750. In both cases, users make a not insignificant investment when acquiring them. The fact that drones are continuously exposed to events during flight that might cause a crash raises concerns among users, since they can lose their UAVs and their investments. Some UAVs, depending on the regulations of the country in which they are going to fly and the category in which they are classified, must carry a lost command and control link recovery system, or a flight termination or safe landing system. However, there are many UAVs of other categories that do not have to carry such systems, for example, those for recreational use. Then, an UAV accident can occur due to external causes (such as a collision with a bird) or other causes such as mechanical failures or human errors. In some cases, depending on the terrain on which the drone has fallen, locating and recovering it might be rather difficult. There are so many UAV applications nowadays that it is worth equipping them with an emergency radio beacon. This is the case for UAVs used, for example, in wildlife control [2,3,4], agriculture [5,6,7], disaster response [8,9,10], etc.

This paper addresses the need to develop radio beacons adequate for locating small UAVs in distress after a crash accident. Traditional emergency locator transmitters (ELTs) used to locate manned aircrafts in an emergency are not suitable for finding small UAVs. They are not only too big and heavy (a common optimized ELT measures approximately 17 cm × 8 cm × 8 cm, and weighs 0.9 kg) but also interact with the international satellite system for search and rescue Cospas-Sarsat System [11,12,13], which utilizes expensive technology and human infrastructures that should be focused on saving people’s lives. The other types of emergency beacons that interact with the Cospas-Sarsat System, which are also not suitable for small drones, are the emergency position-indicating radio beacons (EPIRBs), designed for maritime navigation, and the personal locator beacons (PLBs), beacons for personal use that are usually activated manually (alpinists, hikers, etc.). In addition, the prices for all these distress radio beacons vary from USD 400 (the PLB) up to USD 10,000 (the ELT and EPIRB).

As a result, there is a demand for creating radio localization systems specifically designed for small UAVs, allowing them to be installed on these lightweight unmanned aircraft. This would facilitate their location on the ground after a collision, allowing the recovery of both the airframe and the payload. It is evident that factors such as the weight, size, and flight range of UAVs will dictate the requirements for the weight, dimensions, range, and battery life of the radio beacons they can carry. Furthermore, the proposed solution should aim to be as cost-effective as possible.

Although localization has been a significant research focus in recent decades, and there is an increasing demand for positioning UAVs, the limited localization systems available in the literature primarily target navigation for continuous control of UAVs. This presents a distinct challenge where precision and real-time performance are essential, resulting in systems that are often complex, expensive, and energy-intensive, and bulky and heavy. Moreover, most existing systems do not account for plane crash scenarios, meaning that they lack the ability to detect an accident and activate the entire system only at that moment, which could significantly reduce power consumption. In our previous work [14], we proposed a radio beacon for a mini UAV that met all requirements in size, weight, cost, precision, and user-friendliness. In addition, it had enough range and endurance. That design incorporated a low-cost microelectromechanical accelerometer to detect collisions and a global positioning system (GPS) module to obtain the precise location of the UAV. The main difference from existing localization systems for small UAVs, which were based on receivers that detected signal strength or signal direction, is that our design provided the user with the exact location of the crashed aircraft, considerably easing the recovery of the plane. Nevertheless, the range of that radio beacon was limited by the radio frequency (RF) module used to transmit the localization. The manufacturer of this module provided an outdoor RF line of sight (LOS) range of up to 40 km with a 2.0 dBi dipole antenna. We conducted tests in a quasi-direct path in the quasi-free space of more than 2.5 km between the radio beacon and the receiver. The experimental tests confirmed the range of 2.5 km. However, experiments showed that it was very difficult to achieve a 40 km LOS between the radio beacon and the receiver due to topographic characteristics and vegetation of the terrain. Furthermore, later experiments showed that the practical range decreased dramatically in the presence of obstacles (less than 10 km).

This work proposes a radio beacon based on our previously reported one [14] but with the primordial objective of providing a really long practical range—several tens of kilometers at least—regardless of obstacles and vegetation of the terrain around the UAV and the user, and all this to be verified experimentally. In addition, the other characteristics of the radio beacon (user-friendliness, small size and weight, long endurance, and reduced costs) have been addressed again to gain improvement with respect to the older version.

The technical and innovative contributions of this paper are as follows:

(1)

Proposal and application of another wireless communication strategy in the radio beacon that allows reaching really long distances (beyond 50 km).

(2)

Study and application of technical measures for the reduction of power consumption in the radio beacon.

(3)

Development of improvements in the code for data processing.

(4)

Integration of modules and components in a compact box to achieve a prototype close to a real final product.

(5)

Validation of our prototype radio beacon in different contexts.

The rest of this paper is as follows. Section 2 presents an update of the literature review we made in [14] to compare our work with somewhat related ones. Section 3 presents the proposed emergency radio beacon based on our previous work and describes the current improvements incorporated, along with details of its implementation. Section 4 shows the experimental tests performed to validate the prototype. Section 5 presents the results and discusses them. Finally, Section 6 draws the conclusion of this work.

2. Related Work

In [14], we reviewed the literature on localization systems and emergency radio beacons for UAVs, concluding that no specific emergency radio beacons for UAVs had been developed. The reviewed studies addressed other types of beacons, such as EPIRBs for vessels or portable GSM systems to locate individuals, but these solutions were complex, expensive, or impractical for UAVs. The review also considered UAV localization systems that used GPS and cameras for indoor navigation, along with vision and inertial sensor-based localization methods, but these did not meet the goal of locating a UAV in the event of a crash. Other studies proposed localization infrastructures requiring ground-based transmitting stations, but these were costly and lacked details on size, weight, or power consumption, limiting their practicality for UAVs in emergency environments.

For the current work, we updated that review. The result is that we found hardly any specific emergency radio beacons for UAVs. We found personal locator beacons (PLBs) and EPIRBs but they are not practical for small UAVs. In the following, we present the references found that are more related to our work.

In [15], it is proposed to use Bluetooth low-energy (BLE) wearables (such as smart bands, smart watches, and earbuds) as personal beacons in emergencies, even if the user is unconscious. They propose using BLE detection systems on ground and aerial robots to locate victims in complex disaster areas. However, the range they achieve is only 15 m.

The authors in [16] propose using iBeacon technology to determine the position of an entity in inner space, using the principle of trilateration, but this is not the problem we are focused on.

Study [17] proposes a LoRa technology-based vehicular communication system to improve emergency response times for road accidents by automatically sending crash information to emergency services. Using broadcast beacons, it transmits key data, such as location and accident details, over multi-hop routes to cover large areas with minimal latency. Simulations show that higher node density improves network coverage and packet delivery, though with a slight increase in latency.

The work [18] presents the design and implementation of an emergency message beacon system (EMBS) using the automatic position reporting system (APRS) protocol. The device benefits from a widely distributed radio network where modern wireless network infrastructure is unavailable (open seas or in the middle of a jungle). The system transmits an immediate beacon containing the current location when the unit crashes and automatically sends interval beacons to report location, system status, and an instant message. The unit works with either an existing or an impromptu APRS ad hoc network. This system would not only be used to alert search and rescue services, which is not our objective, but it also has a range of only 2.1 km.

In [19], a PLB is presented. Using the power of the Internet of Things, this distress beacon allows the wearer of the device to transmit a distress signal to the authorized Android device in less than two minutes. The signal consists of his location and photos of the conditions of his surroundings. Although the size, weight, autonomy, and range of this device are unknown and we cannot know whether it meets the requirements for a small UAV, the disadvantage compared with ours is that it needs to have access to a Wi-Fi network. So, coverage problems can easily arise.

In [20], a prototype proof-of-concept is presented for an emergency radio beacon focused on cyclists. The system is intended to automatically detect an accident and send an alert to relatives, friends, or emergency services, including the user’s location, in order to assist the person as soon as possible. Microelectromechanical (MEMS) inertial sensors are distributed between the bicycle and the helmet. The device attached to the bicycle also includes a GNSS receiver and a LoRa transceiver to send geolocation data over long distances until they reach a gateway; then, it is forwarded to the cloud to be processed and finally sent as an alert via SMS. This emergency radio beacon has some parts in common with ours (accelerometer to detect impact, GPS) but its maximum range is 7.6 km, much lower than ours. Furthermore, it has a packet loss of 18%. Its size, weight, and autonomy are unknown.

The proposal in [21] is a quadcopter that acts as an emergency beacon for large aircraft. Like our previous beacon [14], it has an accelerometer to detect impacts and a GPS transmitter to send the location. This UAV would be on board the aircraft. When the aircraft crashed or an emergency landing occurred, the quadcopter equipped with the GPS would depart from a certain location on the aircraft and transmit the location data to the GPS receiver. However, if an ELT placed on the tail of an aircraft is susceptible to fires or severe shocks due to aircraft accidents, this UAV is even more susceptible.

As we already reported in [14], we also extended the literature review to localization systems for UAVs and we found more references, but the existing designs are not well suited to be onboard this class of UAV, nor are they intended to report the localization of the collision scenario. For example, [22] presents a solution to the problem of UAV navigation, based on maintaining a constant radio signal between the UAV and the control system. Radio communication is maintained by building a mesh network based on LoRa data transmission technology modules throughout the entire robot path. The navigation system is a mesh network based on the radio beacon. But, it has nothing to do with emergency radio beacons. Thus, we have ruled out this type of work.

3. Proposed Emergency Radio Beacon

Figure 1 shows the emergency radio beacon system scheme. There are important changes that improve the performance in [14], as will be explained. The scheme in Figure 1 is straightforward and efficient. The radio beacon includes an accelerometer, a microcontroller, and a module with a wireless transmitter and a GNSS receiver. GNSS (Global Navigation Satellite System) is a satellite-based technology that provides geo-spatial positioning data to users with compatible receivers. It allows for determining location, navigation, and timing information anywhere on Earth, provided there is an unobstructed view of the sky. It includes multiple satellite constellations operated by different countries: GPS (USA), GLONASS (Russia), Galileo (European Union), BeiDou (China), NavIC (India, regional system with limited global capabilities), and QZSS (Japan, regional system complementary to global GNSS). A GNSS receiver can use signals from multiple constellations to improve accuracy, reliability, and availability, especially in areas with poor satellite visibility (urban canyons or mountainous areas). Thus, a GNSS receiver provides access to all global navigation systems, meaning it can be used around the world and in more remote areas than a GPS receiver. The accelerometer detects significant impacts with obstacles such as trees, birds, other UAVs, or the ground, enabling substantial power savings by keeping the system in sleep mode (microcontroller) and the GNSS and transmitter off until an impact occurs. The GNSS receiver identifies the beacon’s location when the UAV is down, and the wireless transmitter communicates this position to the receiver. The ground station operator only requires a mobile phone.

This scheme offers several benefits. First, the radio beacon not only transmits a distress signal on a designated radio frequency but also directly supplies the coordinates generated by the GNSS receiver. Another benefit is that the radio beacon can remain in sleep mode during regular flight operations, greatly saving power (as GNSS receivers in standard mode can consume a significant amount of energy). Furthermore, the radio beacon is an external component with its own independent battery, and its GNSS receiver operates separately from the UAV’s navigation GNSS.

Figure 2 shows the flow diagram of the main program of the microcontroller. Although it appears similar to [14], now we have also made improvements to reduce power consumption, as will be explained later. The operational logic of the proposed radio beacon is as follows:

• The system is initially powered on when the UAV is prepared for take-off, either manually by the ground pilot or automatically by autopilot software.
• After powering on, the supply control circuit we have incorporated is deactivated and the wireless and GNSS module stays switched off. In addition, the microcontroller enters sleep mode.
• In the event of a crash or abrupt landing, the accelerometer detects the impact. If it senses an acceleration of more than five Gs, it sends an interrupt signal to wake up the microcontroller, which then verifies if this acceleration indicates a genuine impact. If confirmed, the radio beacon activates to transmit its location; otherwise, the microcontroller returns to sleep mode to save power.
• Once activated, the microcontroller starts the supply control circuit, and this connects the battery to the wireless and GNSS module. This gathers information (e.g., location and time) and then forwards these data to the wireless transmitter.
• The wireless transmitter relays the information to the ground station or receiver. It will periodically resend this information at preset intervals with updated GNSS data, continuing until the user locates the UAV and manually turns off the beacon or until its battery is depleted.

In this work, as in [14], we have also assumed a level of 5 Gs for the accelerometer to detect a strong impact and generate an interruption [14] and a 5 min interval as adequate to transmit the location to the receiver (the GNSS could take 2 or 3 min to obtain a fix and, in addition, the transmitter saves power because it does not use a continuous transmission mode).

The component used to detect impacts is the low-power three-axis MEMS accelerometer ADXL345. It features four sensitivity ranges from +/− 2 G to +/− 16 G. It consumes approximately 175 μA in measurement mode. Its size is 25 mm × 19 mm × 3.14 mm. It weighs 1.27 g. The accelerometer is the component responsible for detecting the impact of the UAV by measuring a sudden change in its acceleration. In [14], it was established that, in a mini UAV with the radio beacon inserted at the center of its structure, an acceleration on any axes greater than 5 Gs during at least 10 ms shows a strong impact. Furthermore, we have now selected a rechargeable Li-po battery of 11.1 V and 2000 mAh, size of 75 mm × 35 mm × 25 mm, and weight of 140 g.

Nevertheless, in this work we have proposed and implemented several innovative improvements to our previous radio beacon to extend its performance, as described in the following subsections.

3.1. Wireless Transmitter

The first main difference between the current proposed radio beacon and that of [14] is the transmitter. In the old version, it was a low-cost low-power RF module (Xbee PRO 868) that transmitted the localization with an outdoor RF line-of-sight range up to 40 km with a 2.0 dBi dipole antenna. A range of 2.5 km between the radio beacon and the receiver was experimentally confirmed. Nevertheless, in later studies and experiments, it was noticed that it was very difficult to achieve a 40 km LOS between the radio beacon and the receiver due to topographic characteristics and vegetation of the terrain. In addition, the practical range decreased dramatically in the presence of obstacles (less than 10 km).

In the current version, we have opted for a transmitter that sends the location through a short message service (SMS) text message. SMS is a telecommunications protocol that allows the exchange of short text messages between mobile devices, typically over a cellular network. Each SMS message is limited to 160 7-bit characters or 70 2-byte characters depending on the character encoding used. This change has a number of advantages.

• It increases ease of use, as most of the population today has a smartphone with GPS. The smartphone acts as the receiver of the message, and by simply clicking on the coordinates link, it displays a map with the position of the beacon and the real-time position of the mobile phone.
• It considerably reduces the price compared with the previous version.
• It uses a wide-coverage network that is already available.

3.1.1. Mobile Network

For the radio beacon to be able to send a text message with the coordinates, it must have access to a mobile telephone network. There are different types of mobile networks (generations 2G, GRPS, 3G, 4G, 5G, etc.), so it is necessary to choose the most suitable option for our application. The main difference between them is the increased speed of data exchange in the most recent generations. The speed issue on this device is not very important, as the amount of data transmitted is very small for any of the generations of mobile networks. What an emergency beacon really requires is the widest possible coverage. After studying the coverage map of mobile networks in Spain, we have decided to use the 2G network (GSM network). In Spain, POS terminals or data phones, for example, which are used to pay by card in most shops, use 2G technology. There are other countries, such as the United Kingdom, where 2G is still widely used for older phones devices and for Internet of Things (IoT), such as smart meters, eCall systems, and vehicle tracking systems, to avoid the high cost of patent licenses for new technologies. In this work, it was decided to use it for the following two reasons: (1) it is one of the networks with the most coverage at the moment, and (2) GSM communication modules are much cheaper than the rest (the processors required to use the GSM system are much simpler since the information is processed more slowly and in smaller quantities). For example, in the analysis of possible modules for this work, it was found that, while a 2G module such as the SIM800L costs less than EUR 4, a 4G module such as the SIM7600G can cost EUR 45, that is, ten times more.

Furthermore, it is easy to update the beacon to use another mobile network, since most SIM (subscriber identity module) communication modules use the same AT command language (Hayes Commands) to configure themselves.

3.1.2. Combined Communications and GNSS Module

In [14], an RF module is used as the transmitter and another breakout is used for the GPS module. In the proposed radio beacon, we use a single board (SIM868 development board) that combines the processing GSM and GNSS in a chip of 17.6 mm × 15.7 mm × 2.3 mm and 1.5 g.

3.2. Microcontroller

In [14], a self-contained board based on a high performance 32-bit microcontroller (Atmel SAM3X8E ARM Cortex-M3 CPU) was used. Its operating voltage was 3.3 V and its size was 101.52 mm × 53.3 mm × 10 mm (length × width × height), with 36 g as weight. Now, we used a much smaller board, based on a microcontroller Atmega328P, with 3.3 V and 8 MHz. We demonstrated that this board has enough power for our application, while it has a size as small as 33.6 mm × 18.5 mm × 10 mm and a weight of only 4.6 g.

3.3. Ground Station

The second main difference between the current proposed radio beacon and that of [14] is at the ground station. In the old version, the receiver was a second RF module with an Xbee USB adapter that received the location data from the radio beacon and sent them to a laptop. When entering the latitude and longitude coordinates into a map application, the user could find the location of the downed UAV. In the current version of the radio beacon, a SMS text message arrives on the user’s mobile phone. This text message is a link that the user must click to open the free Google Maps application (version 24.49.06.703287473) and show the position of the crashed UAV using the geographic coordinates.

3.4. Other Implementations Details

Two improvements have been introduced in the proposed radio beacon to reduce energy consumption and give the system greater autonomy. One has been completed through the incorporation of new hardware and the other through software optimization.

3.4.1. Hardware Improvement

In [14], after powering on the radio beacon, both the GPS and microcontroller entered sleep mode until the UAV crashed. Now, the microcontroller enters sleep mode but the communication and GNSS module stays off because we have added a control circuit for the connection of the SIM868 module to the battery. Only if a shock has been detected, this circuit allows the SIM868 module to connect to the battery and, therefore, receive the power necessary to operate. Disconnecting this module from the power supply saves much more energy than leaving it in sleep mode.

The control circuit is made of two PNP transistors, three resistors, and an electrolytic capacitor. A prototype of it is shown in Figure 3. This circuit is used to connect and disconnect the power supply on the positive side of the load (communication board with SIM868 module), maintaining low power consumption in both the on and off state.

3.4.2. Software Improvement

Much of this saving can be achieved through programming. The code has now been written to optimize consumption, eliminating unnecessary processes. In this regard, the following actions have been taken in the final code:

• Serial communication functions via USB have been eliminated.
• Functions have been used to “sleep” the system processor when it is not needed at its maximum capacity. In addition, some functions have also been turned off, such as the ADC module (analog digital converter module) or the BOD module (brown out detection module).

4. Experimental Tests

We developed a prototype for the emergency radio beacon for small UAVs to test its functionality. Figure 4a illustrates this prototype. To facilitate outdoor testing, a box was designed to contain the radio beacon. It is shown in Figure 4b. This box was printed using fused deposition modeling with a filament 3D printer and has holes for the USB connection and the microcontroller power supply. In addition, it includes two holes for the GSM and GNSS antenna cables. A Velcro strap is used to open and close it during testing. The dimensions of this test case are 94 mm × 34 mm × 34 mm. A compact system has been achieved that can be incorporated into small UAVs, both inside and outside. Although it does not yet have certain characteristics of a final product, such as shock resistance, impact absorption, water tightness, and buoyancy, it can be considered a representation of one.

The operation of the radio beacon was tested through various experimental setups. Initially, each module was tested individually, followed by testing the entire system.

In the first step, we checked the operation of the SIM868 module. When asked for position data, the module returns a data frame like the following: “+CGNSINF: 1,1,20210516210418.000,37.308973,−5.887029,61.248,0.00,108.8,1,,1.3,1.6,1.0,,13,6,,,43,,”. This frame has the following structure: +CGNSINF: GPS status, positioning status, date and time UTC (coordinated universal time), latitude, longitude, altitude above sea level, speed, heading, positioning mode, reserved 1, HDOP, PDOP, VDOP, reserved 2, visible GNSS satellites, used GNSS satellites, used GLONASS satellites, reserved 3. The operation of text message sending was also tested. Previously, an activated micro SIM card must be inserted into the dedicated slot on the communications board.

In the developed code, due to the way the SIM868 module returns information, this information is extracted and separated before sending it by text message. In the case of our prototype, only the latitude and longitude will be sent. The aim is to facilitate the work of the operator in charge of finding the beacon by directly sending a link with the coordinates to the Google Maps site. To achieve this, it will be sent in the following way: http://maps.google.com/maps?q=loc:latitude,longitude (accessed on 5 March 2024). The first check is with the radio beacon while at the Higher Technical School of Engineering of Seville (Spain). In this experiment, Google Maps confirmed that these coordinates correspond with the Higher Technical School of Engineering of Seville. To check the location precision, we measured the position with the GNSS receiver in different locations, in both urban and rural environments, obtaining some pairs of coordinates and we obtained our “real” locations pointed out in Google Maps. Then, we calculated the difference in meters between GNSS locations and locations according to the map application. In general, we can say that the results obtained confirm that, in the areas where the considered UAV is going to fly (rural areas, outdoors, far from any great obstacles), the accuracy of the GNSS receiver given by its manufacturer is valid and it is very good.

Second, we checked the response of the accelerometer to taps, double taps, vibrations, and strong shocks to ensure that it was reading the data correctly, including when it was motionless. This helped us define a series of events to discriminate possible false detections of crash events. After that, we programmed it to send an interruption to the microcontroller when it detected an acceleration greater than 5 G during at least 10 ms, which is the threshold established in [14] to detect great impacts. We also considered the readings of the accelerations as discriminatory against hard take-offs and landings. We then performed impact tests in the laboratory and read the data when it was dropped directly on a table. In addition, we moved it quickly into the air and crashed into a rigid object and deformable object. In any case, we verified that the interruption was activated with sufficient impact and that the lectures were greater than 5Gs during 10 ms on any of the three axes x, y, z. We also checked other levels of acceleration and crash duration to detect strong impacts to try and ensure the correct operation of the radio beacon. Furthermore, as already stated in [14], we supported the decision on trigger level and minimum crash time using telemetry obtained from real flights of two different UAVs.

The power control circuit was also verified. Finally, the operation of the entire system was verified. Tests were carried out in an urban area, more precisely, in the city of Seville (Spain). First, the radio beacon and the receiving phone were set only a few meters from each other, in an indoor area, and we checked that the whole system worked correctly. Once the radio beacon was activated, after hitting the accelerometer in a manual way, the link to the location was sent by SMS to the receiving phone, showing its proper functioning. We also carried out experiments in an outdoor urban area with many buildings and trees and checked more than a 10 km range. After the first tests and verification of the correct functioning of the system was carried out, a final version of the program to be loaded into the finished beacon was edited. The intention was to eliminate all unused functions, with the aim of lightening the program, thus allowing the processor to have greater fluidity and less load, which also extends its useful life. This process involved eliminating all debugging options.

Table 1. Comparison of program storage and dynamic memory usage.

Program	Program Storage [Used Bytes/Total Bytes] (Percentage)	Dynamic Memory [Used Bytes/Total Bytes] (Percentage)
Original version	13,528/32,256 (41%)	1362/2048 (66%)
Light version	11,716/32,256 (36%)	805/2048 (39%)

shows a comparison between program storage and dynamic memory usage. With the simplified version of the code, there was a reduction in dynamic memory of 27% and storage space of 5%. When a high percentage of dynamic memory is used, operational problems begin to occur, since the free space is used by local variables.

Other outdoor tests were carried out in the province of Seville. A range of more than 50 km was verified, even in bad weather conditions. We confirmed the excellent coverage of the 2G network of the company Movistar, which was the telephone operator used for the radio beacon. In fact, the 2G network has coverage in practically the entire Spanish territory.

The radio beacon was mounted on a multirotor UAV, as can be seen in Figure 5. An impact experiment was carried out in what could be a flight zone far from populated areas. The multirotor was dropped from a height of approximately one meter. The casing was also hit several times, simulating the crash of the UAV. In all cases, an SMS was received on the receiving phone with a link to the location of the radio beacon, showing its proper functioning. A time of thirteen seconds was measured from the impact until the SMS was received. Figure 6 shows the result of double clicking on the link contained in the received SMS. The Google Maps site opens and the location of the radio beacon, and therefore the UAV, is displayed with high precision.

Although, basically for economic reasons, we were only able to perform limited tests, we can say that our design meets the primordial objective of providing a really long practical range—several tens of kilometers at least—regardless of obstacles and vegetation of the terrain around the UAV and the user. In addition, the other characteristics of the radio beacon (user-friendliness, small size and weight, long endurance, and reduced costs) were addressed again to gain improvement with respect to the older version regarding most requirements (maximum weight of 0.3 kg, minimum endurance of 2 h, maximum volume of the radio beacon of 6 cm × 6 cm × 12 cm, and minimum range of 10 km).

5. Results and Discussion

In this section, all the results obtained in the development of the emergency beacon prototype for small UAVs developed in this work will be compiled. The results will be analyzed, which will allow us to make a comparison with other existing emergency radio beacons and drone recovery technologies, along with works related in some way to ours. In addition, based on these results, improvements and modifications to the design will be proposed.

Regarding dimensions, our previous radio beacon [14] had a size of 101.5 mm × 53.3 mm × 50 mm, while the beacon proposed here has a slightly smaller size, 94 mm × 34 mm × 34 mm. The weight is similar in both cases, around two hundred grams. These sizes and weights can be easily accommodated by most small UAVs, and, of course, by larger ones. It is estimated that any UAV with a maximum take-off weight (MTOW) greater than one-and-a-half kilograms could carry this radio beacon without problems.

The consumption of the emergency radio beacon was measured in two situations. With the first version of the program, it was verified that the required current is 55 mA when the radio beacon is at rest (before an impact) and 90 mA when it has been activated (i.e., a power of 660 mW in the first case and 1080 mW in the second). With the light version of the code, it was verified that the required current is 42 mA when the radio beacon is at rest (before an impact) and 90 mA when it has been activated (i.e., a power of 546 mW in the first case and 1080 mW in the second). It was verified that the light version of the code saves on consumption.

Once consumption has been determined, the autonomy of the system can be estimated. Our radio beacon, which is powered by its own battery, has an endurance of approximately 24 h of transmission, whereas the system in [14] could have 15 h of transmission. Both times are enough to find the UAV. In sleep mode, both can operate for several days.

Regarding the cost of the prototype, the total price of the components used in this circuit amounts to about sixty euros (EUR 60). If we used a generic UNO microcontroller board with the Arduino design, the price of the beacon would be less than EUR 40. Thus, for less than EUR 40 in materials, a functional and automatic radio beacon for UAVs has been developed. This cost is minimal compared with the price of one of the UAVs on which it could be used, which may be carrying devices such as a light detection and ranging (LIDAR) system or video cameras, with costs to the order of thousands of euros. This beacon is much cheaper than the UAV location methods on the market and reduces the price of the first version (USD 193). Furthermore, it is not necessary to operate a specific receiver module, as in [14] (the receiver cost USD 69), because the location can be received directly on the mobile of the user (saving another USD 69).

Finally, the first objective pursued, that of significantly increasing the range of the radio beacon, has also been achieved. For our previous radio beacon [14], it was very difficult to achieve a 40 km LOS between the radio beacon and the receiver due to topographic characteristics and vegetation of the terrain. Furthermore, the practical range diminished dramatically in the presence of obstacles (less than 10 km). Now, we are using the GSM network, one of the networks with the most coverage at the moment. Distance is no longer a problem for our current radio beacon.

There is another aspect worth mentioning. It is recommended that the duration of execution of the processes in an emergency situation be as short as possible. With a shorter duration, the location of the crashed UAV can be reached earlier. This could prevent the spread of a possible fire in a natural environment, and even the theft of the aircraft. The time elapsed has been measured from the time the impact takes place to the reception of the text message on the receiving mobile phone. However, it must be taken into account that this time will depend largely on several factors as follows:

• The ambient temperature, since the GNSS antenna works worse at low temperatures.
• The location and orography of the terrain, which affects GSM coverage and satellite visibility.
• The time of day: the number of visible satellites can vary from time to time.
• The processing of the mobile phone that receives the message.

The average time in different tests has been less than 20 s, exceeding it only on rare occasions when it did not connect with a sufficient number of satellites because the GNSS antenna was inside a building. In addition, the hot positioning time (when the antenna has been used moments before) has always been less than a second. It is worth highlighting the good results of the SIM868 GNSS module, considerably superior to the SIM808 module, even managing to establish communication inside some buildings.

Furthermore, a new version of an emergency beacon for small unmanned aircraft has been developed with the following characteristics:

• Automatic detection of accidents (with accelerometer).
• Fast and precise positioning (use of GNSS).
• Wide coverage (GSM).
• Sending position by text message (SMS).
• Low manufacturing and use costs. (<EUR 40, SMS).
• Reduced energy consumption (various measures to save energy).
• Low weight (around 100 g).
• Compact size (94 × 34 × 34 mm).

The results obtained make this beacon a very attractive and useful device, valid for a large number of small and medium UAVs. The proposed emergency radio beacon significantly improves upon the existing systems identified in the literature, addressing key limitations in terms of range, practicality, and suitability for small UAVs. Unlike traditional emergency locator transmitters (ELTs), emergency position-indicating radio beacons (EPIRBs), or personal locator beacons (PLBs) that are unsuitable due to their size, weight, or reliance on costly satellite infrastructures [11,12,13], our design is specifically optimized for also lightweight UAVs, offering a compact, low-cost, and power-efficient solution. We have demonstrated that our current beacon is better than our previous one [14], especially in terms of range. Existing localization systems, such as those utilizing Bluetooth low-energy (BLE) [15] or iBeacon technology [16], achieve limited ranges (e.g., 15 m for BLE). While these solutions are adequate for specific applications, they are impractical for UAV recovery in outdoor or large-area scenarios. Similarly, systems using LoRa or automatic position reporting systems (APRSs) [17,18,20] exhibit longer ranges but remain constrained by significant limitations. For example, the LoRa-based beacon for cyclists [20] achieves only 7.6 km with notable packet loss (18%), whereas APRS-based systems [18] report ranges as low as 2.1 km, insufficient for practical UAV recovery.

Moreover, solutions relying on ground-based infrastructures [17,22] or Wi-Fi connectivity [19] face challenges related to cost, availability, or coverage in obstacle-rich or remote environments. These issues highlight the need for a more robust and autonomous approach, which our beacon addresses by leveraging GSM-based communication. This enables a practical range exceeding 50 km, verified experimentally, and ensures compatibility with existing mobile network infrastructures even in challenging terrains.

In addition, unlike the quadcopter-based emergency beacon proposed in [21], our design does not depend on an additional UAV for localization, avoiding increased susceptibility to damage during crashes. The integration of an accelerometer to activate the system only upon impact further enhances power efficiency, a feature not consistently addressed in existing designs.

Currently, there are also some commercial transponders that can be used to track mini UAVs, but we have not considered them because they are not focused on the problem we are facing (designing a simple user-friendly low cost emergency radio beacon), but they are designed to provide a response when they receive an RF interrogation or even broadcast information like position, speed, direction, height, and identification of the UAV. They are used in navigation and, because of their complexity and sophisticated functionality, they are typically high cost (several thousand dollars).

Overall, our beacon’s combination of extended range, compact size, lightweight design, reduced cost, and user-friendliness represents a substantial advancement over the systems reviewed. By addressing the unique requirements of small UAVs in crash scenarios, our solution bridges a critical gap in the field of UAV recovery technologies.

However, there are some issues that we intend to improve in the future. One is the choice of the microcontroller; if a smaller one was used, both the size and weight of the beacon could be significantly reduced. Another issue is that, in a few years, the network we use in this project, the 2G network, may be completely eliminated and replaced by some other network such as 4G or 5G, which right now do not have as much coverage in the territory. In that case, the SIM868 module could be changed for another module valid in the new network. This task would not be complicated, since this type of module uses the same language (Hayes code) for the configurations.

We would also like to include methods for the beacon to detect emergency situations, even before the possible impact occurs. For example, if an aircraft loses the command signal mid-flight and the return-to-home system does not exist or fails, the radio beacon could be activated by a signal from the flight controller in the event of loss of the control link.

Another option would be to check whether there is no movement after an impact, since, moments after an accident, the movement should be zero, except for precision errors. To achieve this, it would be very useful to use the ADXL345 accelerometer’s inactivity interruption.

Other parameters, in addition to the location coordinates, could also be sent, for example, the height, the time of data collection, or the precision of the coordinates.

We could also include the possibility of charging the beacon’s battery through the power supply system of the UAV on which it is mounted. In this way, the need to charge the beacon every certain number of flight hours would be avoided. To this end, a charging system must be designed that does not affect the operation of the beacon.

Regarding the exterior design, a casing could be designed with some of the following characteristics: light weight, resistant material, water tightness, buoyancy, fire resistance, and shock absorption, with planned detachment if necessary. In this way, it would be guaranteed that the beacon would continue to function in the vast majority of emergency situations.

6. Conclusions

In this work, a low-cost long-range emergency radio beacon for small UAVs has been developed that meets the essential requirements of precise localization and autonomy in accident conditions, proving to be an effective solution to improve the recovery of drones in disaster situations. Throughout this work, significant advances have been made compared with previous prototypes, highlighting the incorporation of a combined communications and GNSS module, an SMS-based communication strategy that eliminates the dependence on specialized receiver devices and improves the effective range of the system, even in environments with obstacles. Experimental tests demonstrate that the prototype meets the objectives of coverage and rapid response to impacts, and the portability and low energy consumption characteristics required for its application in small drones. With a manufacturing cost reduced to less than EUR 40 and an autonomy of up to 24 h in transmission mode, this device represents an economical and efficient solution compared with commercial alternatives, providing UAV operators with a reliable tool to minimize losses and protect their investment. This radio beacon is presented as a viable solution for a wide range of commercial and recreational UAVs.

Future improvements are planned in terms of size and weight reduction, adaptation to future telecommunication networks (4G/5G), and the design of a robust and lightweight housing that increases its ability to survive various types of accidents. In addition, the integration of advanced functionalities, such as emergency detection before impact and the inclusion of charging through the UAV’s power source, can further enhance the utility and flexibility of this beacon in real-world contexts.