If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
AED delivery at night – Can drones do the Job? A feasibility study of unmanned aerial systems to transport automated external defibrillators during night-time
Department of Anaesthesiology, Intensive Care, Emergency Medicine, Transfusion Medicine and Pain Therapy, EvKB, University Hospital of Bielefeld, Campus Bielefeld-Bethel, Bielefeld, Germany
Department of Anaesthesiology, Intensive Care, Emergency Medicine and Pain Therapy, University Hospital of Bielefeld, Campus Municipal Hospital, University of Bielefeld
Department of Anaesthesiology, Intensive Care, Emergency Medicine, Transfusion Medicine and Pain Therapy, EvKB, University Hospital of Bielefeld, Campus Bielefeld-Bethel, Bielefeld, Germany
Corresponding author at: Department of Anaesthesiology, Intensive Care, Emergency Medicine, Transfusion Medicine and Pain Therapy, EvKB, University Hospital of Bielefeld, Campus Bielefeld-Bethel, Burgsteig 13, 33617 Bielefeld, Germany.
Department of Anaesthesiology, Intensive Care, Emergency Medicine, Transfusion Medicine and Pain Therapy, EvKB, University Hospital of Bielefeld, Campus Bielefeld-Bethel, Bielefeld, Germany
In their recent guidelines the European Resuscitation Council have recommended the use of Unmanned Aerial systems (UAS) to overcome the notorious shortage of AED. Exploiting the full potential of airborne AED delivery would mandate 24 h UAS operability. However, current systems have not been evaluated for nighttime use. The primary goal of our study was to evaluate the feasibility of night-time AED delivery by UAS.
The secondary goal was to obtain and compare operational and safety data of night versus day missions.
Methods
We scheduled two (one day, one night) flights each to ten different locations to assess the feasibility of AED delivery by UAS during night-time. We also compared operational data (mission timings) and safety data (incidence of critical events) of night versus day missions.
Results
All missions were completed without safety incident. The flights were performed automatically without pilot interventions, apart from manually choosing the landing site and correcting the descent. Flight distances ranged from 910 m to 6.960 m, corresponding mission times from alert to AED release between 3:48 min and 11:20 min. Night missions (m:night = 7:26 ± 2:29 min) did not take longer than day missions (m:day = 7:59 ± 2:27 min). Despite slightly inferior visibility of the target site, night landings (land:night = 64 ± 15 sec) were on average marginally quicker than day landings (land:day = 69 ± 11sec).
Conclusions
Our results demonstrate the feasibility of UAS supported AED delivery during nighttime. Operational and safety data indicate no major differences between day- and night-time use. Future research should focus on integration of drone technology into the chain of survival.
Out-of-hospital cardiac arrest (OHCA) claims more than 60.000 lives per year in Germany and only less than 10% of the patients affected survive without permanent neurological deficit.
However, survival rates of 50% or higher could be achieved if public access defibrillation (PAD) and Basic Life Support (BLS) performed by first responders or bystanders were immediately available.
This eye watering discrepancy illustrates the compelling need for optimising care for patients suffering OHCA. Unlike in our neighboring countries, in Scandinavia and the Netherlands, AED are not comprehensively available in Germany, particularly not in rural regions. According to the German Cardiac Arrest Registry and confirmed by own unpublished data, only 2% of OHCA patients currently receive early bystander defibrillation. Most OHCA happen in residential areas,
where AED are hardly ever found. Instead, many AED are placed in public locations and are inaccessible out of hours, further weakening the chain of survival during nighttime.
Automated External Defibrillators Inaccessible to More Than Half of Nearby Cardiac Arrests in Public Locations During Evening, Nighttime, and Weekends.
Optimizing Outcomes After Out-of-Hospital Cardiac Arrest With Innovative Approaches to Public-Access Defibrillation: A Scientific Statement From the International Liaison Committee on Resuscitation.
Drones delivering automated external defibrillators—Integrating unmanned aerial systems into the chain of survival: A simulation study in rural Germany.
These studies have also confirmed that employing UAS could significantly shorten the time to first defibrillation, especially in sparsely populated rural regions. The cardiac arrest research group at Karolinska Institute, Stockholm have demonstrated that airborne AED delivery to the site of a cardiac arrest is possible, safe and potentially faster than the ambulance service.
These convincing results have prompted the European Resuscitation Council and the International Liaison Committee on Resuscitation to consider airborne AED delivery in their recent guidelines.
Optimizing Outcomes After Out-of-Hospital Cardiac Arrest With Innovative Approaches to Public-Access Defibrillation: A Scientific Statement From the International Liaison Committee on Resuscitation.
Integration of airborne AED delivery with smartphone dispatched community first responders was holistically evaluated in our previous project, the overarching MV|Life|Drone study,
Drones delivering automated external defibrillators—Integrating unmanned aerial systems into the chain of survival: A simulation study in rural Germany.
where we conducted 50 full scale simulations, commencing with a bystander finding a cardiac arrest victim and ending with the first shock delivered by a first responder. In line with previous research, MV|Life|Drone suggests using AED-drones in combination with smartphone activated first responders is feasible, safe and probably reduces the time to first defibrillation. Using the identical setup, we are now in the position to capitalise on these results and focus our current research on issues highlighted by MV|Life|Drone, without the need to repeat the full scale simulations. One of the key issues identified, was the lack of research into night operability of airborne AED networks.
Drones delivering automated external defibrillators—Integrating unmanned aerial systems into the chain of survival: A simulation study in rural Germany.
Drones delivering automated external defibrillators—Integrating unmanned aerial systems into the chain of survival: A simulation study in rural Germany.
we intended to narrow this gap with the perspective to improve 24 h availability of PAD.
The primary goal of our study was to evaluate the feasibility of night-time AED delivery by Unmanned Aerial Systems (UAS). The secondary goal was to obtain and compare operational and safety data of night versus day missions.
Methods
Study design
This is an open, controlled simulation study, carried out in the rural northeast of the German federal state of North-Rhine-Westphalia; 20 drone flights to 10 different sites within a 7 km radius around the UAS base were scheduled. The UAS base was located on a remote farm. The target sites were arbitrarily chosen locations, representing a mixture of private homes and everyday public sites, typical for a rural environment. Two flights to each location were conducted; one during night time and one during day time, with the day flights serving as control.
The need for ethical approval was waived by the Ethics Committee of the University of Münster, Münster, Germany (No 2022-015-f-S). In line with German legislation, we followed the exemption regulations for emergency service UAS operators for Beyond the Visual Line of Sight (BVLOS) operations.
This included preparing a detailed, formal concept of operations and a specific operational risk assessment for all flights. Before the study commenced, the district authorities, the regional Emergency Dispatch Centers as well as the operator of the regional air ambulance were informed and a mobile phone hotline established to ensure airspace deconfliction.
Materials
We used the same equipment as described in the MV|Life|Drone project
Drones delivering automated external defibrillators—Integrating unmanned aerial systems into the chain of survival: A simulation study in rural Germany.
: The UAS was a multipurpose, octocopter (Fig. 1; type “Ceptor Transport” manufactured by Globe UAV, Germany) with the corresponding Ground Control Station (GCS). The UAV can operate between temperatures of −20 and +50 °Celsius. A high-resolution infrared night vision system with focusable infrared laser beam (Globe UAV, Germany) was added. The drone was operating out of a heated, self-operating hangar (Fig. 1). The top speed was 80 km * h−1 and the maximum flight duration was 40 minutes. The AED was a Schiller FRED Easyport (0.86 kg; manufactured by Schiller, Obfelden, Switzerland), which was transported in a protective hardcase and deployed by a purpose-built carrier attached to the bottom of the aircraft. According to the manufacturer short-time exposure to temperatures below 0 °C, as caused by a drone flight in cold weather conditions, should not affect the AED’s performance.
Fig. 1Drone in heated hangar, which provides all-weather protection. The hangar opens automatically on system activation.
Based on the geo-references, the GCS proposed direct flight corridors to the target sites. These corridors led over built up areas and no-fly-zones, which we had to avoid, as our missions were for research purposes, not allowing us to use the full spectrum of the emergency service exemptions. Therefore, the pilot, who was located at the base, adjusted the corridors manually directly before the launch of the drone (Video 1). All flights were performed automatically, BVLOS, controlled by the GCS and monitored by the pilot through live data transmission and live video stream by on-board camera. The pilot, had the possibility to intervene and take over flight control at any time, if required.
Simulations
The simulation started with a mock alert. The coordinates of the target location were then copied into the GCS, which calculated a direct flight corridor. The pilots, who all were certified drone operators in line with EU legislation, adapted the flight corridor manually as described above. On system activation the port opened, the drone took off, ascended to cruise height and flew automatically BVLOS within the flight corridor to the target location. The Concept of Operation entailed flying at standard cruise speeds of 40–50 km * h−1 during day and night operations. Under real life conditions the drone would fly at full speed of 80 km * h−1. Two observers, in direct contact with the pilot via mobile phone, were present at the landing sites. On arrival of the drone at the target coordinates, the pilot chose the landing site, and the drone landed automatically with manual corrections of the descent. The simulation ended with the release of the AED from the carrier after the rotors had come to a halt.
Data acquisition and processing
The primary outcome was measured by comparing the number of successful landings at the target site against the number of aborted missions (daytime versus night-time). The secondary outcomes were assessed by comparing the mission times (as per Table 1.), the incidence of untoward incidents during flight (near misses, emergency landings, collisions), approach of target location (visibility at landing site), and landing at scene (safe touch down – clear of any objects/ person). Data were recorded by the GCS as well as by the observers who tracked all steps from alert to release of the AED at scene. Video transmissions of the drone cameras were recorded to compare the visibility of the landing zones, which were graded by the drone pilots as excellent, good, satisfactory or poor. The grading was confirmed by crosschecking the video recordings by a second drone pilot.
Table 1UAS mission timings.
t0
Alert, pilot receives target coordinates, start of flight preparations
t1
UAS activation, port opens
t2
Take-off drone
t3
Drone has reached cruise height of 60 meters above ground
t4
Drone has reached target coordinates, start landing procedure
Landing on scene is the most hazardous phase of the mission and is critically depending on sufficient visibility of the landing site. Poor visibility at this stage is the most likely cause of delay or mission abort. All other flight phases were conducted automatically and were as such not daylight depending.
For data interpretation we applied descriptive statistics and confidence intervals.
Results
During March and April 2022 twenty missions to ten different locations were completed as planned. Comprehensive data on these flights are depicted in Table 2. The flights were performed automatically and no pilot interventions were necessary, apart from choosing the landing site after arrival at the target coordinates and manually correcting the automatic descent. The distance covered by the UAV ranged from 910 to 6.960 m with corresponding mission times (t0-t6) ranging from Tm = 3:48 to 11:20 min. The average mission times were m:day = 7:59 ± 2:27 min and m:night = 7:26 ± 2:29 min. The average time from alert until the aircraft reached cruising height (t0–t3) was (s) = 1:40 ± 0:10 sec. The average cruise speed of the aircraft was 45 ± 6 km * h−1 Wind speeds ranged from 10 to 35 km * h−1 and temperatures varied from 9 to 16° Celsius. The visibility at the landing sites were rated excellent during daylight and good during night operations.
Table 2Comprehensive flight data.
Destination
1 (day)
1 (night)
2 (day)
2 (night)
3 (day)
3 (night)
4 (day)
4 (night)
5 (day)
5 (night)
Alarm time (t0, CEST)
07:05 pm
10:50 pm
07:39 pm
00:05 am
06:33 pm
00:36 am
05:13 pm
00:58 am
04:07 pm
09:20 pm
Port opening (t1, min:sec)
00:42
00:39
00:29
00:25
00:32
00:26
00:26
00:25
00:22
00:17
Take-off (t2, min:sec)
01:19
01:20
01:13
01:10
01:09
01:06
01:09
01:11
01:10
01:04
Cruise height reached (t3, min:sec)
01:50
01:51
01:44
01:40
01:38
01:37
01:38
01:40
01:42
01:35
Start landing (t4, min:sec)
09:30
09:44
07:55
08:36
08:10
06:15
08:55
08:10
06:40
05:18
Touch down (t5, min:sec)
10:40
10:46
08:51
09:28
09:17
07:10
09:51
08:44
07:56
06:40
Drop off AED (t6, min:sec)
10:42
10:48
08:52
09:29
09:18
07:11
09:52
08:46
07:58
06:42
Wind speed (km * h−1)
11–26
10–20
11–26
10–20
11–26
10–20
10–17
11–26
11–26
15–30
Wind direction
NW
NNW
NW
NNW
NW
NNW
NW
NNW
NW
E
Flight direction
E
E
NE
NE
NNE
NNE
E
E
NW
E
Precipitation
No
Rain
No
No
No
No
No
No
No
No
Landing site
Private home
Private home
Farm
Farm
Field
Field
Building site
Building site
Forest trail
Forest trail
Visibility
excellent
satisfactory
excellent
good
excellent
good
excellent
good
excellent
good
Flight distance (km)
6.25
5.95
5.42
5.34
4,02
3.64
4.72
4.87
2.92
2.98
Average cruise speed (km * h−1)
48.9
45.3
52.6
46.2
36.9
47.1
38.9
45.0
35.3
48.1
Direct aerial distance (km)
5.43
5.43
4.61
4.61
3.63
3.63
3.56
3.56
2.87
2.87
Destination
6 (day)
6 (night)
7 (day)
7 (night)
8 (day)
8 (night)
9 (day)
9 (night)
10 (day)
10 (night)
Alarm time (t0, CEST)
04:35 pm
09:33 pm
02:11 pm
09:50 pm
02:36 pm
10:07 pm
02:57 pm
10:20 pm
05:55 am
10:54 pm
Port opening (t1, min:sec)
00:18
00:15
00:12
00:10
00:16
00:09
00:19
00:11
00:48
00:27
Take-off (t2, min:sec)
01:04
01:02
00:54
01:00
01:07
00:59
01:39
01:01
01:27
01:13
Cruise height reached (t3, min:sec)
01:34
01:33
01:25
01:31
01:39
01:29
02:08
01:33
01:56
01:33
Start landing (t4, min:sec)
05:10
04:22
03:04
02:49
04:30
04:09
04:00
04:18
10:09
09:45
Touch down (t5, min:sec)
06:39
05:30
04:03
03:46
05:38
05:30
05:22
05:24
11:18
11:05
Drop off AED (t6, min:sec)
06:40
05:31
04:04
03:48
05:39
05:32
05:23
05:26
11:20
11:06
Wind speed (km * h−1)
11–26
15–35
5–15
10–25
5–15
10–25
11–26
10–20
11–27
10–20
Wind direction
NW
E
NW
E
NW
E
NW
E
NW
E
Flight direction
NNE
NNE
S
S
SSW
SSW
S
S
E
E
Precepitation
No
No
No
No
No
No
No
No
No
No
Landing site
Factory
Factory
Football field
Football field
Car park
Car park
Chapel
Chapel
Lake side
Lake side
Visibility
excellent
good
excellent
good
excellent
good
good
good
good
good
Flight distance (km)
2.98
2.08
1.08
0.91
2.33
2.33
1.5
1.43
6.96
6,65
Average cruise speed (km * h−1)
39,50
44,31
39,27
42,00
49,05
52,43
48,21
31,2
50,82
49,67
Direct aerial distance (km)
1.76
1.76
0.9
0.9
2.33
2.33
1.5
1.5
6,21
6.21
Table 2 Operational data of 20 drone flights from the UAS base to 10 different destinations (one day flight and one night flight).
On one night mission however, heavy rain caused infrared reflections on approach leading to a just satisfactory view of the site; the landing could be completed safely nevertheless. The average day landing time was land:day 69 ± 11sec, whereas the average night landing time was land:night 64 ± 15 sec. The 95% confidence interval for the day/night difference in landing time was CI = −3.5–14.5 sec.
A video recording of a night landing sequence is supplied in the additional material (Video 2).
With regard to system safety no emergency interventions of the UAS pilots were necessary and no near misses, emergency landings or collisions occurred. The on-site observers deemed all landings safe and did not report any safety concerns.
Additional observations:
•
During landing, the night vision system gave an excellent high-resolution view of objects within the cone of infrared light. Objects outside the cone however, remained difficult to visualize (Video 2).
•
The visibility of the drone is, due to its bright position lighting, much better at night than during day time.
•
The bright position lighting of the drone might help bystanders and first responders to locate the aircraft after landing on scene.
•
During preparatory night flights in early March 2022, the drone got repeatedly into heavy rain and snowfall with temperatures falling below 0 °C. Interestingly this did not affect the performance of the system and a safe return to base was possible in all cases.
Drones delivering automated external defibrillators—Integrating unmanned aerial systems into the chain of survival: A simulation study in rural Germany.
has identified several fields for further research into airborne AED delivery. The most important field was night operation capability, as efficient utilisation of airborne AED delivery mandates 24 h operability of UAS networks. Our current results confirm that AED delivery at night can be accomplished by UAS and that associated safety aspects are manageable. All day and night missions were completed without any safety incident. The flights were carried out in automatic flight mode and pilot interventions were only required for choosing the landing site and correcting the descent of the drone. Overall, night missions did not take longer than day missions.
Earlier research has not addressed night flight and night landing capabilities of AED-UAS
; therefore, our study represents a first step towards closing this knowledge gap and opening the door to 24 h operability. Our findings are based on only twenty flights conducted under mild weather conditions. Nevertheless, the results are encouraging and justify further reaching evaluation of 24 h operability in comprehensive field trials. Daytime flights in heavy weather had been conducted successfully with the same UAS type during the MV|Life|Drone study.
Drones delivering automated external defibrillators—Integrating unmanned aerial systems into the chain of survival: A simulation study in rural Germany.
For this trial, we excluded heavy weather conditions as a potential confounder. During preparatory night flights however, we had encountered heavy rain and snowfall as well as temperatures below 0 °C. Notably, these adverse weather conditions did not affect the performance of the UAS. Obviously, there are variations in performance between different UAS types and we assume that the heavy octocopter we have employed is probably better suited for flying under poor weather conditions than lighter models.
Collision avoidance is of paramount importance in BVLOS flight operations. Effective airspace deconfliction in uncontrolled airspace (below 300 m above ground level) could be achieved with transponder systems, making UAS detectable and trackable for other aircraft and UAV pilots. Unfortunately, transponder systems are not legally required yet, despite appropriate systems being available and in use.
Airspace deconfliction at night appears to be easier to achieve than during day time; the uncontrolled airspace is used foremost during the day (by recreational aviators) and not at nighttime. Also, at night the UAS position lighting is well visible over several miles, further reducing the collision risk.
Landing at the target site is regarded the most critical phase of the operation. In previous studies the drones landed mostly directly at the scene to deliver the AED,
Drones delivering automated external defibrillators—Integrating unmanned aerial systems into the chain of survival: A simulation study in rural Germany.
without any major incident being reported. Landing BVLOS at night under low light conditions presents a challenge, which in civilian practice has not been addressed before. Our infrared-laser night vision system ensured an at least satisfactory overview of all target sites, allowing for a safe approach and landing on scene. Despite slightly inferior visibility of the target sites at night, the night landings were marginally quicker. The 95% confidence interval for difference in landing time (CI = −3.5 to 14.5 sec) was clinically insignificant in the context of the overall flight duration.
The working group at Karolinska Institute have recently adopted an alternative way of landing the defibrillator: the AED was lowered by winch, with the drone hovering over the landing site without making ground contact.
Drones delivering automated external defibrillators—Integrating unmanned aerial systems into the chain of survival: A simulation study in rural Germany.
This has not been confirmed under low light conditions yet. However, our on-site observations suggest, that the landing spots are sufficiently illuminated by the position lights of the UAV, hence we do not feel that AED retrieval is less safe or effective at night.
The directions for future research and development on airborne AED delivery were comprehensively depicted in recent open access publications.
Drones delivering automated external defibrillators—Integrating unmanned aerial systems into the chain of survival: A simulation study in rural Germany.
All authors highlighted the need for integrating this novel technology into existing emergency medical systems and combining it with first responder activation. Regarding night operability, future research and development should address automated routing, airspace deconfliction, night vision systems, landing methods of the AED and night functionality of different UAS models in heavy weather conditions. We have not experienced any safety issues in our trial; however, firm conclusion may be drawn from larger implementation trials. In the light of the existing evidence and the rapidly evolving technology, the next step would be moving to translational research projects to develop and explore integrated models of PAD. Such trials are complex, but possible under German legislation. They would require a trans-sectoral approach involving community responder schemes, AED registries, Emergency Medical Services, cardiac arrest registries, district authorities, civil aviation authorities and, last not least, capable UAS technology partners.
Limitations
The study was conceived as a feasibility study and conducted in a safe and planned setting with predefined and surveyed landing sites. The flight operations were performed in flat and sparsely populated areas, hence less challenging than in mountainous regions or densely populated cities.
We have observed differences in flight distances between the UAS base and the corresponding target locations. These were caused by the manual corrections of the flight routes, which the pilots did not consistently perform in the same manner. Under real life conditions flight regulatory exemptions would have applied, allowing them to use the shortest route, without major detours. Automated routing would make the need for manual correction redundant.
The concept of operations entailed flying at regular cruise speeds of 40–50 km * h−1. This was the first time that night flight under BVLOS condition was evaluated in Germany. We therefore chose not to test the system to its limits. Under real life conditions the maximum system speed of 80 km * h−1 would be used.
Specifications between different UAS types differ significantly; accordingly, our results only refer to the UAS investigated.
Given the limited numbers of flights we have conducted and the controlled environment we were operating in, our results must be regarded preliminary and need to be confirmed in further field trials.
Conclusion
Our results demonstrate the feasibility of UAS supported AED delivery during nighttime. Operational and safety data indicate no major differences between day- and night-time operations. We conclude that airborne AED delivery represents a promising approach towards 24/7 PAD cover, especially in rural regions. Automated routing, flying at maximum speed and using the full spectrum of flight regulatory exemptions, would help to decrease the current response time of airborne AED delivery. Further translational research should focus on the integration of drone technology into the chain of survival.
Funding
The study was funded by EvKB University Hospital, Bielefeld, Germany and supported by a 15.000 € research grant from the Laerdal Foundation, Stavanger, Norway. The study sponsor had no involvement in the study design, in the collection, analysis and interpretation of data nor the writing of the manuscript and the decision to submit for publication.
Disclosures
All authors declare that they have no Conflict of Interest regarding this publication.
Appendix A. Supplementary material
The following are the Supplementary material to this article:
Automated External Defibrillators Inaccessible to More Than Half of Nearby Cardiac Arrests in Public Locations During Evening, Nighttime, and Weekends.
Optimizing Outcomes After Out-of-Hospital Cardiac Arrest With Innovative Approaches to Public-Access Defibrillation: A Scientific Statement From the International Liaison Committee on Resuscitation.
Drones delivering automated external defibrillators—Integrating unmanned aerial systems into the chain of survival: A simulation study in rural Germany.
00047-3&id=gr1.jpg){kind=link}