- Email: [email protected]
Unmanned aerial vehicles (drones) to prevent drowning
Resuscitation 127 (2018) 63–67
Contents lists available at ScienceDirect
Resuscitation journal homepage: www.elsevier.com/locate/resuscitation
Clinical paper
Unmanned aerial vehicles (drones) to prevent drowning☆ a
a
b
c
Celia Seguin , Gilles Blaquière , Anderson Loundou , Pierre Michelet , Thibaut Markarian a b c
c,⁎
T
Emergency Medical Service – SAMU 40, Centre Hospitalier Layné, Mont de Marsan, France Health Assessment Research Unit, EA 3279, Department of Public Health, Aix-Marseille Université, Marseille, France Emergency Department, Hôpital de la Timone, UMR MD2 P2COE, Aix-Marseille Université, Marseille, France
A R T I C LE I N FO
A B S T R A C T
Keywords: Drone Drowning Unmanned aerial vehicles Rescue Lifeguards Simulation
Background: Drowning literature have highlighted the submersion time as the most powerful predictor in assessing the prognosis. Reducing the time taken to provide a flotation device and prevent submersion appears of paramount importance. Unmanned aerial vehicles (UAVs) can provide the location of the swimmer and a flotation device. Objective: The objective of this simulation study was to evaluate the efficiency of a UAV in providing a flotation device in different sea conditions, and to compare the times taken by rescue operations with and without a UAV (standard vs UAV intervention). Several comparisons were made using professional lifeguards acting as simulated victims. A specifically-shaped UAV was used to allow us to drop an inflatable life buoy into the water. Results: During the summer of 2017, 28 tests were performed. UAV use was associated with a reduction of time it took to provide a flotation device to the simulated victim compared with standard rescue operations (p < 0.001 for all measurements) and the time was reduced even further in moderate (81 ± 39 vs 179 ± 78 s; p < 0.001) and rough sea conditions (99 ± 34 vs 198 ± 130 s; p < 0.001). The times taken for UAV to locate the simulated victim, identify them and drop the life buoy were not altered by the weather conditions. Conclusion: UAV can deliver a flotation device to a swimmer safely and quickly. The addition of a UAV in rescue operations could improve the quality and speed of first aid while keeping lifeguards away from dangerous sea conditions.
Introduction Drowning is still a leading causes of unintentional death worldwide [1]. Unintentional drowning occurs in diverse locations (lakes, rivers and coastlines), different weather conditions and affects adults as well as children. A clear picture of drowning is therefore challenging to brush. Despite this heterogeneity, a recent meta-analysis on drowning outcome at the scene has emphasized the importance of rapid care factors as short submersion durations (< 5 min) and short Emergency Medical Service (EMS) response times (< 9 min) [2]. Consequently, submersion time seems to be the most powerful predictor in assessing the prognosis of drowning [2–6]. This highlights the importance of prevention and early reaction in the drowning process [7]. A unique and “universal Drowning Chain of Survival” has been recently proposed to guide the life-saving steps for lay and professionals rescuers [8]. To match these results and recommendations [2–6,9], most bodies of water open to the public and offering watersports in France are supervised by lifeguards. After the recognition of victim distress and call for help, the next ☆ ⁎
priority is to interrupt the drowning process by providing flotation to the victim [8]. Nevertheless, sea condition, currents and wave size could interfere with rescue operations especially on the sea coast. Providing a flotation device using a drone equipped with an inflatable life buoy could be a new way to reduce the time taken by lifeguards to get a flotation device to the swimmer, particularly in difficult conditions. Drones have previously been reported as having the potential to transport an Automated External Defibrillator (AED) in case of Out of Hospital Cardiac Arrest (OHCA) before the emergency medical services arrive [10]. Drones may also have the potential to reduce the time taken to provide cardio-pulmonary resuscitation in drowning related OHCA by their recognition and alert ability [11] though not by a direct intervention yet. The aim of this simulation study was to evaluate the efficiency of a drone in providing the location of a drowning victim and dropping an inflatable life buoy into the water. This test was performed and compared with standard rescue operations in different locations and under different sea conditions on the Atlantic coast of France.
A Spanish translated version of the abstract of this article appears as Appendix in the final online version at https://doi.org/10.1016/j.resuscitation.2018.04.005. Corresponding author at: Emergency Services, Hôpital de la Timone, UMR MD2 P2COE, Aix-Marseille Université, Marseille, France. E-mail address: [email protected] (T. Markarian).
https://doi.org/10.1016/j.resuscitation.2018.04.005 Received 14 January 2018; Received in revised form 19 March 2018; Accepted 9 April 2018 0300-9572/ © 2018 Elsevier B.V. All rights reserved.
Resuscitation 127 (2018) 63–67
C. Seguin et al.
Fig. 1. Unmanned Aerial Vehicle (UVA) patrol intervention. A: Unmanned Aerial Vehicle (UVA); B: Simulated victim; C: Drop life buoy; D: Inflatable life buoy
Methods
by one of the doctors taking part in the study. The doctor and a lifeguard recorded the time between the alarm being rung to the life buoy being received by the simulated victim. Following this, two strategies were compared and analyzed. All the lifeguards involved in this study were specifically trained police lifeguards and were not told that the study results would be analyzed or published to ensure rescue conditions would remain as close to standard rescue as possible. These lifeguards usually took part in beach patrol every year, and had done so for several years.
This is a prospective simulation study comparing two methods for locating a simulated drowning victim and providing them with a life buoy. This study was conducted on three beaches of the Atlantic coast of France in different sea conditions. All tests were performed during daylight over two months – July and August 2017. Conditions on the Atlantic coast of France
Standard rescue operation by lifeguards (SRO)
The Atlantic coast of France mainly consists of a 200 km beach stretching from the Gironde estuary to the Spanish border. Although the sea temperature is warm – between 18 and 24 °C in the summer period – the coast is subject to currents and large waves. Moreover, the effects of the current on this particular area of beach create dangerous pools of water called “baïnes”. Each summer, these conditions cause dozens to drown. All sea conditions were recorded during the tests. The calm, moderate and rough conditions have been defined according to the French legislation for beach security. These levels mainly depend on tide, swell size and wind and the local authority for beach security chose the level each morning in summer period. In order to mimic real conditions, no weather restrictions were set as the lifeguards considered that the beach was open to swimmers. Sea conditions were categorized as calm, moderate and rough.
Each lifeguarding team was composed of one lifeguard located at the bottom of the sand dune who looked for the victim using binoculars. We do not recorded the time for this first lifeguard to visually locate the simulated victim using binoculars. Two or three other lifeguards were on a watchtower near the beach. Once the victim was located and the alarm had been rung, the lifeguard nearest the shore went into the sea (using swim fins) with a life buoy while another used a jet ski on the beach to head towards the victim. The time taken from starting at the beach (both for the lifeguard and the jet-ski) to reaching the simulated victim was recorded using a waterproof GPS tracker. It appears important to report separately the results for the swimmer and the jet-ski since the availability of a jet-ski is not systematic for the French coast and to provide respective contribution of the swimmer, the UAV and the jet-ski.
Test scenario
UAV patrol (intervention)
For each test, a simulated victim (lifeguard) chose a position in the sea at different distances from the beach (between 100 and 200 m), within a predefined search area of 50000 m2. This area reflects the standard ‘swimming allowed’ area for a lifeguard team during the summer period. The distance from the beach was chosen as reflecting the most frequent location of drowning related to rip currents and baïnes in French Atlantic coast conditions. The other lifeguards were blinded to the location of the simulated victim and the alarm was rung
A lifeguard with special training for flying a UAV was positioned on the sand dune and launched the drone to a height of 50m. With live video streaming from the UAV to an iPad, the lifeguard was able to fly the UAV and watch its journey. Once the UAV reached the simulated victim, the live video (real-time 720p HD video, with a range of more than 2 km) meant that the lifeguard could assess the situation and 64
Resuscitation 127 (2018) 63–67
C. Seguin et al.
Discussion
decide whether or not to drop the life buoy (Fig. 1). Afterwards, we analyzed the times between the UAV arriving over the simulated victim and the life buoy being dropped into the water. The life buoy was a specifically dedicated self-inflated one (Zodiac™). Once the buoy was in water, a salt-starter allowed an immediate self-inflation. This method allowed a drop of life buoy very close to the victim. The drone used was a Helper® Drone (www.helper-drone.com), specifically shaped for bad weather conditions (wind resistant up to 80 km/h in gusts of wind, empty weight: 4.5 kg, maximum take-off weight: 6 kg, payload of 1.5 kg). In order to ensure the tests were safe, the drone was equipped with a ballistic parachute which ejected from its casing from a height of 3 m, limiting the risk of injury in the event of a crash. In addition, a Helper® Vision program was available – a geographic information system which lets you create no-fly zones and altitude restrictions. The drone was also programmed with a return-home program (using GPS), which could be used if the signal from the drone pilot was lost or if the battery was low. These flight tests were carried out within line of sight and were approved by the Direction Générale de l’Aviation Civile (the French national aviation authority).
Our study was the first to demonstrate that a UAV is able to provide a flotation device faster than standard rescue operations using lifeguards. In terms of unintentional drowning deaths worldwide, significant efforts have been made to improve prevention measures and first aid policies. Considering a “drowning timeline”, as recently proposed, the successive steps of the drowning chain of survival include a reaction phase closely associated with mitigation [7]. If the victim is not able to reach the shore by himself, the first challenge for a rescue team is to locate the victim in distress and to start following the drowning chain of survival. Conversely, the fact that the swimmer is not in distress appears to be important in preventing the unnecessary involvement of rescuers. The UAV streams a live video to a trained lifeguard, allowing them to assess the situation fully and respond in an appropriate way. The concept of using a UAV for this purpose has been analyzed by Claesson et al. who demonstrated that it enables us to locate submerged victims earlier than with standard procedure, which leads to earlier resuscitation for drowning victims [11]. In the same way, our results showed that using a UAV reduces the time taken to locate the simulated victim, and that these times did not vary depending on sea conditions (something that was not evaluated by the Claesson study performed in calm conditions only). Regardless of the conditions, the fact that the victims are most frequently unable to cry for help when struggling and about to drown supports the value of using video streaming [8]. Using a UAV could also help us identify a victim at high risk of drowning, demonstrating behaviors such as a near-vertical body position, ineffective downward arm movements, ineffective pedaling or kicking leg actions, and making little or no forward progress through the water [12]. Once it has been confirmed that the victim is drowning, the next step is to trigger the alert and initiate the rescue process including first aid and emergency medical services. During this time the victim should be removed from the water as soon as possible, although following the drowning chain of survival, a flotation device must be provided to prevent or stop submersion. A recent meta-analysis has demonstrated that the strongest factor associated with favorable outcome after drowning was submersion duration [2]. Indeed, reducing submersion durations to less than 6 min was associated with better outcome [2,3]. Nevertheless, when the victim is taken out of the water, it is crucial that lifeguards do not become victims themselves by engaging in dangerous behavior. In this way, it has been reported that in order to reduce the risk during this step of the chain of survival, a rescuer must provide a flotation device to assist the victim [8]. Our study confirms this risk since in two cases the rescue team was unable to use a jet ski due to the size of waves. Moreover, our results showed that it took longer for the lifeguard to reach the simulated victim in poorer sea conditions. These conditions were associated with longer times to keep the simulated victim out of the sea when calm and rough sea conditions were compared. The use of UAV could prevent these consequences without increasing the time taken to provide the victim with a life buoy, as demonstrated in our study. The potential value of using a UAV has recently been outlined in several studies. Claesson et al. have shown that a UAV could be used to locate submerged victims early on, and therefore could contribute to earlier onset of cardio-pulmonary resuscitation [11]. Our study proposes a further use for the UAV: not only is it able to improve the situation assessment by providing close range observation, but it can also become actively involved in the drowning chain of survival itself by providing the victim with a life buoy. Similarly, in another recent study Claesson et al. demonstrated that there may be a potential value in using a UAV to deliver an Automated External Defibrillator (AED) in Out of Hospital Cardiac Arrest (OHCA) and reducing time to defibrillation [10]. For the drowning chain of survival, this contribution appears of paramount importance in poor weather conditions where the UAV is able to reach the victim and drop the flotation device more
Statistics For descriptive statistics, IBM SPSS Statistics version 21 was used. Continuous variables are expressed as means ± SD. Comparisons of means values between the two groups were performed using Student’s ttest or the Mann-Whitney U test according to the data distribution (Kolmogorov-Smirnov test). A p-value of < 0.05 was considered to be significant.
Ethical approval Approval from an Ethics Committee was not required for this study, in accordance with current French law. The tests were included in standard training for lifeguard teams and the use of UAV modified the training process. This study and local rescue operations were supported by the Aquitaine region.
Results During the study period, 28 tests were carried out under different sea conditions: 8 in calm conditions, 13 in moderate conditions and 7 in rough conditions. The average tidal coefficient (the ratio of the mean tidal range, in a given place) was 65 ± 22. The average distance between the first lifeguard location (near the shore) and the simulated victim was 118 ± 63 m. The average time for the UAV to reach the simulated victim was 70 ± 38 s. The global comparison showed that the life buoy could be grabbed by the simulated victim faster with UAV than with SRO (87 ± 37 vs 160 ± 94 s, p < 0.001). The difference in results between UAV use and SRO was more significant in moderate and rough sea conditions (Fig. 2 and Table 1). When the Jet Ski was used, the average time taken to reach the simulated victim was 105 ± 25 in calm conditions, 178 ± 25 in moderate conditions and 142 ± 39 s in rough conditions (Table 1). In two tests, the rough conditions did not allow the Jet Ski to reach the simulated victim safely, due to the size of the waves (data not included). The average time taken to get the simulated victim out of the water (i.e. on the shore) once rescuers reached the victims was 134 ± 22 s. in calm conditions, 209 ± 65 s. in moderate conditions and 284 ± 97 s. in rough conditions (p < 0.05 between calm and rough). During the tests, the UAV was used three times for real rescue operations with success (data not shown for ethical reasons). 65
Resuscitation 127 (2018) 63–67
C. Seguin et al.
Fig. 2. Time to get buoy according to operation mode and sea conditions.
Study limitations
Table 1 Time to reach the simulated victim and get buoy according to operation mode and sea conditions.
UAV Time Life Buoy grab SRO Time Jet-Ski Time
Calm conditions N rescues = 8
Moderate conditions N rescues = 13
Rough conditions N rescues = 7
64 ± 25 10.5 ± 4 93 ± 47 * 105 ± 25 §
68 ± 41 13 ± 5 179 ± 78 178 ± 25
84 ± 34 15 ± 8 198 ± 130 * 142 ± 39 §
* §
In our study, we were able to report the delay between the UAV arrival over the simulated victim and the grab of the self-inflating life buoy. Conversely, we were not able to analyze the time taken to visually locate the simulated victim by the first lifeguard with binocular. The use of lifeguards as victims who are more likely to be able to grab buoy compared to a real casualty represent another limitation since most victims will unlikely be able to swim for long distance to reach a buoy. Nevertheless, the location of UAV just few meters over the victim controlled by a lifeguards and the self-inflation ability of the buoy should limit this distance and thus the inability. For the swimmer lifeguard and the Jet-Ski we have considered that once the swimmerlifeguard or the Jet-Ski were close to the victim, the grab of the buoy by the victim was immediate. The distance from the shore in our study could appears as quite far but these distances represents the picture of drowning in Atlantic French coast related to rip currents and baïnes. Indeed, the time taken for victims to be located and to assess the distress is associated with a location of victim beyond 80 m. Moreover the distances reported were between the lifeguard location (near the watchtower) on the beach and the simulated victim what mainly include sea distance but also some meters on the beach. Nevertheless, when drowning incidents occur within 50 m of shoreline, UAV and lifeguard response times could be similar.
UAV Time: delay from UAV launch to arrival over the victim Life buoy grab: delay from UAV arrival over the victim to life buoy grab by the simulated victim. SRO Time: delay for the swimmer lifeguard to reach the simulated victim. Jet-Ski Time: delay for the lifeguard on the Jet-Ski to reach the simulated victim. UAV: Unmanned aerial vehicle; SRO: standard rescue operations; *: p < 0.05; NS: p > 0.05. * p < 0.05 for UAV Time versus SRO Time. § p < 0.05 for UAV Time versus Jet-Ski Time.
rapidly and in safe conditions for the rescuers. The UAV evaluated in our study has been designed for these conditions and tested in windy and rainy conditions beforehand, in order to mimic real life. As previously mentioned, the development of UAV use for medical applications requires changes in legislation and securing the radio connection with the UAV itself in the event of a fault. In fact, a great deal of caution must be taken to allow flights over urban areas or, in our case, beach areas. Our study was the first in France to receive an official authorization from the Direction Générale de l’Aviation Civile (the French national aviation authority). This authorization was associated with the organization of an air corridor between the UAV take off location and the shore, and the restriction of UAVs only being used within the range of pilots’ line of sight (a radius of one kilometer). Moreover, the UAV tested was equipped with a ballistic parachute which could be rapidly ejected from its casing from a maximum height of 3 m from the ground.
Conclusion Our study is the first to demonstrate that a specifically-designed UAV can deliver a flotation device safely and rapidly. Taking account of the value in using a UAV for medical applications, the UAV is not only a witness and a link but could also participate in the drowning chain of survival. The addition of a UAV in rescue operations improves the quality and speed of first aid while keeping lifeguards away from dangerous sea conditions.
66
Resuscitation 127 (2018) 63–67
C. Seguin et al.
Conflicts of interest
[5] Ballesteros MA, Gutiérrez-Cuadra M, Muñoz P, Miñambres E. Prognostic factors and outcome after drowning in an adult population. Acta Anaesthesiol Scand 2009;53(August (7)):935–40. [6] Youn CS, Choi SP, Yim HW, Park KN. Out-of-hospital cardiac arrest due to drowning: an Utstein Style report of 10 years of experience from St. Mary’s Hospital. Resuscitation 2009;80(July (7)):778–83. [7] Szpilman D, Tipton M, Sempsrott J, Webber J, Bierens J, Dawes P, et al. Drowning timeline: a new systematic model of the drowning process. Am J Emerg Med 2016;34(November (11)):2224–6. [8] Szpilman D, Webber J, Quan L, Bierens J, Morizot-Leite L, Langendorfer SJ, et al. Creating a drowning chain of survival. Resuscitation 2014;85(September (9)):1149–52. [9] Claesson A, Svensson L, Silfverstolpe J, Herlitz J. Characteristics and outcome among patients suffering out-of-hospital cardiac arrest due to drowning. Resuscitation 2008;76(March (3)):381–7. [10] Claesson A, Fredman D, Svensson L, Ringh M, Hollenberg J, Nordberg P, et al. Unmanned aerial vehicles (drones) in out-of-hospital-cardiac-arrest. Scand J Trauma Resusc Emerg Med 2016;24(October (1)):124. [11] Claesson A, Svensson L, Nordberg P, Ringh M, Rosenqvist M, Djarv T, et al. Drones may be used to save lives in out of hospital cardiac arrest due to drowning. Resuscitation 2017;114(May):152–6. [12] Lanagan-Leitzel L. Identification of critical events by lifeguards, instructors and non-lifeguards. Int J Aquat Res Educ 2012;20:3–14.
The authors declare that they do not have any conflicts of interest. Sources of funding This work was supported by a grant by the Nouvelle Aquitaine region. References [1] Szpilman D, Bierens JJLM, Handley AJ, Orlowski JP. Drowning. N Engl J Med 2012;22:2102–10. 31,366. [2] Quan L, JJLM Bierens, Lis R, Rowhani-Rahbar A, Morley P, Perkins GD. Predicting outcome of drowning at the scene: a systematic review and meta-analyses. Resuscitation 2016;104(July):63–75. [3] Quan L, Mack CD, Schiff MA. Association of water temperature and submersion duration and drowning outcome. Resuscitation 2014;85(June (6)):790–4. [4] Suominen P, Baillie C, Korpela R, Rautanen S, Ranta S, Olkkola KT. Impact of age, submersion time and water temperature on outcome in near-drowning. Resuscitation 2002;52(March (3)):247–54.
67
Contents lists available at ScienceDirect
Resuscitation journal homepage: www.elsevier.com/locate/resuscitation
Clinical paper
Unmanned aerial vehicles (drones) to prevent drowning☆ a
a
b
c
Celia Seguin , Gilles Blaquière , Anderson Loundou , Pierre Michelet , Thibaut Markarian a b c
c,⁎
T
Emergency Medical Service – SAMU 40, Centre Hospitalier Layné, Mont de Marsan, France Health Assessment Research Unit, EA 3279, Department of Public Health, Aix-Marseille Université, Marseille, France Emergency Department, Hôpital de la Timone, UMR MD2 P2COE, Aix-Marseille Université, Marseille, France
A R T I C LE I N FO
A B S T R A C T
Keywords: Drone Drowning Unmanned aerial vehicles Rescue Lifeguards Simulation
Background: Drowning literature have highlighted the submersion time as the most powerful predictor in assessing the prognosis. Reducing the time taken to provide a flotation device and prevent submersion appears of paramount importance. Unmanned aerial vehicles (UAVs) can provide the location of the swimmer and a flotation device. Objective: The objective of this simulation study was to evaluate the efficiency of a UAV in providing a flotation device in different sea conditions, and to compare the times taken by rescue operations with and without a UAV (standard vs UAV intervention). Several comparisons were made using professional lifeguards acting as simulated victims. A specifically-shaped UAV was used to allow us to drop an inflatable life buoy into the water. Results: During the summer of 2017, 28 tests were performed. UAV use was associated with a reduction of time it took to provide a flotation device to the simulated victim compared with standard rescue operations (p < 0.001 for all measurements) and the time was reduced even further in moderate (81 ± 39 vs 179 ± 78 s; p < 0.001) and rough sea conditions (99 ± 34 vs 198 ± 130 s; p < 0.001). The times taken for UAV to locate the simulated victim, identify them and drop the life buoy were not altered by the weather conditions. Conclusion: UAV can deliver a flotation device to a swimmer safely and quickly. The addition of a UAV in rescue operations could improve the quality and speed of first aid while keeping lifeguards away from dangerous sea conditions.
Introduction Drowning is still a leading causes of unintentional death worldwide [1]. Unintentional drowning occurs in diverse locations (lakes, rivers and coastlines), different weather conditions and affects adults as well as children. A clear picture of drowning is therefore challenging to brush. Despite this heterogeneity, a recent meta-analysis on drowning outcome at the scene has emphasized the importance of rapid care factors as short submersion durations (< 5 min) and short Emergency Medical Service (EMS) response times (< 9 min) [2]. Consequently, submersion time seems to be the most powerful predictor in assessing the prognosis of drowning [2–6]. This highlights the importance of prevention and early reaction in the drowning process [7]. A unique and “universal Drowning Chain of Survival” has been recently proposed to guide the life-saving steps for lay and professionals rescuers [8]. To match these results and recommendations [2–6,9], most bodies of water open to the public and offering watersports in France are supervised by lifeguards. After the recognition of victim distress and call for help, the next ☆ ⁎
priority is to interrupt the drowning process by providing flotation to the victim [8]. Nevertheless, sea condition, currents and wave size could interfere with rescue operations especially on the sea coast. Providing a flotation device using a drone equipped with an inflatable life buoy could be a new way to reduce the time taken by lifeguards to get a flotation device to the swimmer, particularly in difficult conditions. Drones have previously been reported as having the potential to transport an Automated External Defibrillator (AED) in case of Out of Hospital Cardiac Arrest (OHCA) before the emergency medical services arrive [10]. Drones may also have the potential to reduce the time taken to provide cardio-pulmonary resuscitation in drowning related OHCA by their recognition and alert ability [11] though not by a direct intervention yet. The aim of this simulation study was to evaluate the efficiency of a drone in providing the location of a drowning victim and dropping an inflatable life buoy into the water. This test was performed and compared with standard rescue operations in different locations and under different sea conditions on the Atlantic coast of France.
A Spanish translated version of the abstract of this article appears as Appendix in the final online version at https://doi.org/10.1016/j.resuscitation.2018.04.005. Corresponding author at: Emergency Services, Hôpital de la Timone, UMR MD2 P2COE, Aix-Marseille Université, Marseille, France. E-mail address: [email protected] (T. Markarian).
https://doi.org/10.1016/j.resuscitation.2018.04.005 Received 14 January 2018; Received in revised form 19 March 2018; Accepted 9 April 2018 0300-9572/ © 2018 Elsevier B.V. All rights reserved.
Resuscitation 127 (2018) 63–67
C. Seguin et al.
Fig. 1. Unmanned Aerial Vehicle (UVA) patrol intervention. A: Unmanned Aerial Vehicle (UVA); B: Simulated victim; C: Drop life buoy; D: Inflatable life buoy
Methods
by one of the doctors taking part in the study. The doctor and a lifeguard recorded the time between the alarm being rung to the life buoy being received by the simulated victim. Following this, two strategies were compared and analyzed. All the lifeguards involved in this study were specifically trained police lifeguards and were not told that the study results would be analyzed or published to ensure rescue conditions would remain as close to standard rescue as possible. These lifeguards usually took part in beach patrol every year, and had done so for several years.
This is a prospective simulation study comparing two methods for locating a simulated drowning victim and providing them with a life buoy. This study was conducted on three beaches of the Atlantic coast of France in different sea conditions. All tests were performed during daylight over two months – July and August 2017. Conditions on the Atlantic coast of France
Standard rescue operation by lifeguards (SRO)
The Atlantic coast of France mainly consists of a 200 km beach stretching from the Gironde estuary to the Spanish border. Although the sea temperature is warm – between 18 and 24 °C in the summer period – the coast is subject to currents and large waves. Moreover, the effects of the current on this particular area of beach create dangerous pools of water called “baïnes”. Each summer, these conditions cause dozens to drown. All sea conditions were recorded during the tests. The calm, moderate and rough conditions have been defined according to the French legislation for beach security. These levels mainly depend on tide, swell size and wind and the local authority for beach security chose the level each morning in summer period. In order to mimic real conditions, no weather restrictions were set as the lifeguards considered that the beach was open to swimmers. Sea conditions were categorized as calm, moderate and rough.
Each lifeguarding team was composed of one lifeguard located at the bottom of the sand dune who looked for the victim using binoculars. We do not recorded the time for this first lifeguard to visually locate the simulated victim using binoculars. Two or three other lifeguards were on a watchtower near the beach. Once the victim was located and the alarm had been rung, the lifeguard nearest the shore went into the sea (using swim fins) with a life buoy while another used a jet ski on the beach to head towards the victim. The time taken from starting at the beach (both for the lifeguard and the jet-ski) to reaching the simulated victim was recorded using a waterproof GPS tracker. It appears important to report separately the results for the swimmer and the jet-ski since the availability of a jet-ski is not systematic for the French coast and to provide respective contribution of the swimmer, the UAV and the jet-ski.
Test scenario
UAV patrol (intervention)
For each test, a simulated victim (lifeguard) chose a position in the sea at different distances from the beach (between 100 and 200 m), within a predefined search area of 50000 m2. This area reflects the standard ‘swimming allowed’ area for a lifeguard team during the summer period. The distance from the beach was chosen as reflecting the most frequent location of drowning related to rip currents and baïnes in French Atlantic coast conditions. The other lifeguards were blinded to the location of the simulated victim and the alarm was rung
A lifeguard with special training for flying a UAV was positioned on the sand dune and launched the drone to a height of 50m. With live video streaming from the UAV to an iPad, the lifeguard was able to fly the UAV and watch its journey. Once the UAV reached the simulated victim, the live video (real-time 720p HD video, with a range of more than 2 km) meant that the lifeguard could assess the situation and 64
Resuscitation 127 (2018) 63–67
C. Seguin et al.
Discussion
decide whether or not to drop the life buoy (Fig. 1). Afterwards, we analyzed the times between the UAV arriving over the simulated victim and the life buoy being dropped into the water. The life buoy was a specifically dedicated self-inflated one (Zodiac™). Once the buoy was in water, a salt-starter allowed an immediate self-inflation. This method allowed a drop of life buoy very close to the victim. The drone used was a Helper® Drone (www.helper-drone.com), specifically shaped for bad weather conditions (wind resistant up to 80 km/h in gusts of wind, empty weight: 4.5 kg, maximum take-off weight: 6 kg, payload of 1.5 kg). In order to ensure the tests were safe, the drone was equipped with a ballistic parachute which ejected from its casing from a height of 3 m, limiting the risk of injury in the event of a crash. In addition, a Helper® Vision program was available – a geographic information system which lets you create no-fly zones and altitude restrictions. The drone was also programmed with a return-home program (using GPS), which could be used if the signal from the drone pilot was lost or if the battery was low. These flight tests were carried out within line of sight and were approved by the Direction Générale de l’Aviation Civile (the French national aviation authority).
Our study was the first to demonstrate that a UAV is able to provide a flotation device faster than standard rescue operations using lifeguards. In terms of unintentional drowning deaths worldwide, significant efforts have been made to improve prevention measures and first aid policies. Considering a “drowning timeline”, as recently proposed, the successive steps of the drowning chain of survival include a reaction phase closely associated with mitigation [7]. If the victim is not able to reach the shore by himself, the first challenge for a rescue team is to locate the victim in distress and to start following the drowning chain of survival. Conversely, the fact that the swimmer is not in distress appears to be important in preventing the unnecessary involvement of rescuers. The UAV streams a live video to a trained lifeguard, allowing them to assess the situation fully and respond in an appropriate way. The concept of using a UAV for this purpose has been analyzed by Claesson et al. who demonstrated that it enables us to locate submerged victims earlier than with standard procedure, which leads to earlier resuscitation for drowning victims [11]. In the same way, our results showed that using a UAV reduces the time taken to locate the simulated victim, and that these times did not vary depending on sea conditions (something that was not evaluated by the Claesson study performed in calm conditions only). Regardless of the conditions, the fact that the victims are most frequently unable to cry for help when struggling and about to drown supports the value of using video streaming [8]. Using a UAV could also help us identify a victim at high risk of drowning, demonstrating behaviors such as a near-vertical body position, ineffective downward arm movements, ineffective pedaling or kicking leg actions, and making little or no forward progress through the water [12]. Once it has been confirmed that the victim is drowning, the next step is to trigger the alert and initiate the rescue process including first aid and emergency medical services. During this time the victim should be removed from the water as soon as possible, although following the drowning chain of survival, a flotation device must be provided to prevent or stop submersion. A recent meta-analysis has demonstrated that the strongest factor associated with favorable outcome after drowning was submersion duration [2]. Indeed, reducing submersion durations to less than 6 min was associated with better outcome [2,3]. Nevertheless, when the victim is taken out of the water, it is crucial that lifeguards do not become victims themselves by engaging in dangerous behavior. In this way, it has been reported that in order to reduce the risk during this step of the chain of survival, a rescuer must provide a flotation device to assist the victim [8]. Our study confirms this risk since in two cases the rescue team was unable to use a jet ski due to the size of waves. Moreover, our results showed that it took longer for the lifeguard to reach the simulated victim in poorer sea conditions. These conditions were associated with longer times to keep the simulated victim out of the sea when calm and rough sea conditions were compared. The use of UAV could prevent these consequences without increasing the time taken to provide the victim with a life buoy, as demonstrated in our study. The potential value of using a UAV has recently been outlined in several studies. Claesson et al. have shown that a UAV could be used to locate submerged victims early on, and therefore could contribute to earlier onset of cardio-pulmonary resuscitation [11]. Our study proposes a further use for the UAV: not only is it able to improve the situation assessment by providing close range observation, but it can also become actively involved in the drowning chain of survival itself by providing the victim with a life buoy. Similarly, in another recent study Claesson et al. demonstrated that there may be a potential value in using a UAV to deliver an Automated External Defibrillator (AED) in Out of Hospital Cardiac Arrest (OHCA) and reducing time to defibrillation [10]. For the drowning chain of survival, this contribution appears of paramount importance in poor weather conditions where the UAV is able to reach the victim and drop the flotation device more
Statistics For descriptive statistics, IBM SPSS Statistics version 21 was used. Continuous variables are expressed as means ± SD. Comparisons of means values between the two groups were performed using Student’s ttest or the Mann-Whitney U test according to the data distribution (Kolmogorov-Smirnov test). A p-value of < 0.05 was considered to be significant.
Ethical approval Approval from an Ethics Committee was not required for this study, in accordance with current French law. The tests were included in standard training for lifeguard teams and the use of UAV modified the training process. This study and local rescue operations were supported by the Aquitaine region.
Results During the study period, 28 tests were carried out under different sea conditions: 8 in calm conditions, 13 in moderate conditions and 7 in rough conditions. The average tidal coefficient (the ratio of the mean tidal range, in a given place) was 65 ± 22. The average distance between the first lifeguard location (near the shore) and the simulated victim was 118 ± 63 m. The average time for the UAV to reach the simulated victim was 70 ± 38 s. The global comparison showed that the life buoy could be grabbed by the simulated victim faster with UAV than with SRO (87 ± 37 vs 160 ± 94 s, p < 0.001). The difference in results between UAV use and SRO was more significant in moderate and rough sea conditions (Fig. 2 and Table 1). When the Jet Ski was used, the average time taken to reach the simulated victim was 105 ± 25 in calm conditions, 178 ± 25 in moderate conditions and 142 ± 39 s in rough conditions (Table 1). In two tests, the rough conditions did not allow the Jet Ski to reach the simulated victim safely, due to the size of the waves (data not included). The average time taken to get the simulated victim out of the water (i.e. on the shore) once rescuers reached the victims was 134 ± 22 s. in calm conditions, 209 ± 65 s. in moderate conditions and 284 ± 97 s. in rough conditions (p < 0.05 between calm and rough). During the tests, the UAV was used three times for real rescue operations with success (data not shown for ethical reasons). 65
Resuscitation 127 (2018) 63–67
C. Seguin et al.
Fig. 2. Time to get buoy according to operation mode and sea conditions.
Study limitations
Table 1 Time to reach the simulated victim and get buoy according to operation mode and sea conditions.
UAV Time Life Buoy grab SRO Time Jet-Ski Time
Calm conditions N rescues = 8
Moderate conditions N rescues = 13
Rough conditions N rescues = 7
64 ± 25 10.5 ± 4 93 ± 47 * 105 ± 25 §
68 ± 41 13 ± 5 179 ± 78 178 ± 25
84 ± 34 15 ± 8 198 ± 130 * 142 ± 39 §
* §
In our study, we were able to report the delay between the UAV arrival over the simulated victim and the grab of the self-inflating life buoy. Conversely, we were not able to analyze the time taken to visually locate the simulated victim by the first lifeguard with binocular. The use of lifeguards as victims who are more likely to be able to grab buoy compared to a real casualty represent another limitation since most victims will unlikely be able to swim for long distance to reach a buoy. Nevertheless, the location of UAV just few meters over the victim controlled by a lifeguards and the self-inflation ability of the buoy should limit this distance and thus the inability. For the swimmer lifeguard and the Jet-Ski we have considered that once the swimmerlifeguard or the Jet-Ski were close to the victim, the grab of the buoy by the victim was immediate. The distance from the shore in our study could appears as quite far but these distances represents the picture of drowning in Atlantic French coast related to rip currents and baïnes. Indeed, the time taken for victims to be located and to assess the distress is associated with a location of victim beyond 80 m. Moreover the distances reported were between the lifeguard location (near the watchtower) on the beach and the simulated victim what mainly include sea distance but also some meters on the beach. Nevertheless, when drowning incidents occur within 50 m of shoreline, UAV and lifeguard response times could be similar.
UAV Time: delay from UAV launch to arrival over the victim Life buoy grab: delay from UAV arrival over the victim to life buoy grab by the simulated victim. SRO Time: delay for the swimmer lifeguard to reach the simulated victim. Jet-Ski Time: delay for the lifeguard on the Jet-Ski to reach the simulated victim. UAV: Unmanned aerial vehicle; SRO: standard rescue operations; *: p < 0.05; NS: p > 0.05. * p < 0.05 for UAV Time versus SRO Time. § p < 0.05 for UAV Time versus Jet-Ski Time.
rapidly and in safe conditions for the rescuers. The UAV evaluated in our study has been designed for these conditions and tested in windy and rainy conditions beforehand, in order to mimic real life. As previously mentioned, the development of UAV use for medical applications requires changes in legislation and securing the radio connection with the UAV itself in the event of a fault. In fact, a great deal of caution must be taken to allow flights over urban areas or, in our case, beach areas. Our study was the first in France to receive an official authorization from the Direction Générale de l’Aviation Civile (the French national aviation authority). This authorization was associated with the organization of an air corridor between the UAV take off location and the shore, and the restriction of UAVs only being used within the range of pilots’ line of sight (a radius of one kilometer). Moreover, the UAV tested was equipped with a ballistic parachute which could be rapidly ejected from its casing from a maximum height of 3 m from the ground.
Conclusion Our study is the first to demonstrate that a specifically-designed UAV can deliver a flotation device safely and rapidly. Taking account of the value in using a UAV for medical applications, the UAV is not only a witness and a link but could also participate in the drowning chain of survival. The addition of a UAV in rescue operations improves the quality and speed of first aid while keeping lifeguards away from dangerous sea conditions.
66
Resuscitation 127 (2018) 63–67
C. Seguin et al.
Conflicts of interest
[5] Ballesteros MA, Gutiérrez-Cuadra M, Muñoz P, Miñambres E. Prognostic factors and outcome after drowning in an adult population. Acta Anaesthesiol Scand 2009;53(August (7)):935–40. [6] Youn CS, Choi SP, Yim HW, Park KN. Out-of-hospital cardiac arrest due to drowning: an Utstein Style report of 10 years of experience from St. Mary’s Hospital. Resuscitation 2009;80(July (7)):778–83. [7] Szpilman D, Tipton M, Sempsrott J, Webber J, Bierens J, Dawes P, et al. Drowning timeline: a new systematic model of the drowning process. Am J Emerg Med 2016;34(November (11)):2224–6. [8] Szpilman D, Webber J, Quan L, Bierens J, Morizot-Leite L, Langendorfer SJ, et al. Creating a drowning chain of survival. Resuscitation 2014;85(September (9)):1149–52. [9] Claesson A, Svensson L, Silfverstolpe J, Herlitz J. Characteristics and outcome among patients suffering out-of-hospital cardiac arrest due to drowning. Resuscitation 2008;76(March (3)):381–7. [10] Claesson A, Fredman D, Svensson L, Ringh M, Hollenberg J, Nordberg P, et al. Unmanned aerial vehicles (drones) in out-of-hospital-cardiac-arrest. Scand J Trauma Resusc Emerg Med 2016;24(October (1)):124. [11] Claesson A, Svensson L, Nordberg P, Ringh M, Rosenqvist M, Djarv T, et al. Drones may be used to save lives in out of hospital cardiac arrest due to drowning. Resuscitation 2017;114(May):152–6. [12] Lanagan-Leitzel L. Identification of critical events by lifeguards, instructors and non-lifeguards. Int J Aquat Res Educ 2012;20:3–14.
The authors declare that they do not have any conflicts of interest. Sources of funding This work was supported by a grant by the Nouvelle Aquitaine region. References [1] Szpilman D, Bierens JJLM, Handley AJ, Orlowski JP. Drowning. N Engl J Med 2012;22:2102–10. 31,366. [2] Quan L, JJLM Bierens, Lis R, Rowhani-Rahbar A, Morley P, Perkins GD. Predicting outcome of drowning at the scene: a systematic review and meta-analyses. Resuscitation 2016;104(July):63–75. [3] Quan L, Mack CD, Schiff MA. Association of water temperature and submersion duration and drowning outcome. Resuscitation 2014;85(June (6)):790–4. [4] Suominen P, Baillie C, Korpela R, Rautanen S, Ranta S, Olkkola KT. Impact of age, submersion time and water temperature on outcome in near-drowning. Resuscitation 2002;52(March (3)):247–54.
67