Robohub.org
 

Tackling the European refugee crisis with solar-powered UAVs: A fully autonomous 26 hour search-and-rescue flight

by
24 August 2016



share this:
T=0h. Autonomous (hand-) launch.

T=0h. Autonomous (hand-) launch.

One year after having demonstrated the 81-hour continuous solar-powered flight that is still the current world record in flight endurance for all aircrafts < 50kg total mass, the AtlantikSolar UAV has completed its next milestone by demonstrating the first-ever fully autonomous (from launch to landing) solar-powered perpetual flight with significant payload (Color + Thermal Camera) in a 26-hour Search-and-Rescue (SaR) mission.

While the 81-hour and 28-hour endurance record flights were important milestones that demonstrated the perpetual endurance capability of AtlantikSolar, they required manual pilot control for launch and landing and did not carry any aerial imaging payload. However, the missions that our team really cares about – Search-and-Rescue missions relying on long-endurance aerial sensing to support authorities e.g. in the European refugee crisis currently unfolding over the Mediterranean Sea – require both of these elements, i.e. ease-of-use through full launch-to-land aircraft autonomy as well as significant payloads to help the rescue teams with the detection of victims on land and sea.

The 26-hour solar-powered Search-and-Rescue flight performed by AtlantikSolar AS-3 from July 19th – 20th 2016 demonstrated exactly that, and is, therefore the first-ever flight worldwide to combine:

  • Perpetual flight: 26-hours of solar-powered day/night- and thus energetically-perpetual flight.
  • Full aircraft autonomy: No pilot stick moved within 26-hours of flight.
  • SaR payload: The aircraft carried a 10 Watt 300g payload (1 Color camera, 1 Thermal Camera, 1 ODROID onboard computer with WLAN) and performed victim detection from the air during day and night.
  • Environment-aware: The aircraft performed automatic thermal updraft tracking for increased energetic efficiency and to speed up the battery recharge.

Flight summary

The 26-hour flight started on July 19th at 18:02 local time in Hinwil, CH with full batteries. The launch was performed fully autonomously: After all system checks were complete, the aircraft was tossed in the air via a hand-launch and then continued automatically and without pilot interaction towards its first loitering waypoint.

T=2h. Circling above while live-streaming camera data.

T=2h. Circling above while live-streaming camera data.

The respective flight control technology is based on an ETHZ/3DR Pixhawk autopilot with a custom flight controller designed at the Autonomous Systems Lab. Equipped with a color camera, thermal camera, an onboard computer and wireless LAN, the aircraft began live-streaming the respective images. The full payload was operated until around 21 o’clock, after which the color camera was switched off because of the insufficient lighting conditions. The batteries were at 95% state of charge at that time, having started their discharge shortly before 20 o’clock. Note that operating a fully-fledged SaR payload consuming 10W is a significant challenge for perpetual flight on a small-scale solar-powered UAV such as AtlantikSolar, and the overall power consumption of ~60W was thus closely tracked during the first hours of the evening flight.

 T=5h. Flying at night with position indicator LEDs on.

T=5h. Flying at night with position indicator LEDs on.

T=5h. The airplane is seen loitering in the top left, performing the SolAIR Search-and-Rescue mission.

T=5h. The airplane is seen loitering in the top left, performing the SolAIR Search-and-Rescue mission.

The actual testing of the Search-and-Rescue capabilities started just before 23:00. With the aircraft now flying in total darkness, the infrared camera served as the only source of aerial imaging information. As visible in the video, the payload system clearly manages to find the victim lying in low grass (and surrounding houses, tents, cars, the streets and especially the warm electric generator) at 23:05. The victim detection was performed manually by the ground station operator based on the live-streamed images this time. However, the next AtlantikSolar test flight will have automatic on-board victim detection (as already tested in the ICARUS search and rescue project) implemented. These demonstrated capabilities are integral elements for perpetual flight and aerial victim detection missions. We are convinced they can be of great help in solving pressing issues, such as the European refugee crisis of 2015 and 2016 above the Mediterranean Sea.

T=5h. Live-streamed thermal camera images with annotation. Note the victim that can easily be detected.

T=5h. Live-streamed thermal camera images with annotation. Note the victim that can easily be detected.

T=13h. Ground Station on the next morning. We’ve received first sun here.

T=13h. Ground Station on the next morning. We’ve received first sun here.

Despite heavy winds of up to 8m/s, the Search-and-Rescue support activities with the infrared camera payload were continued throughout the rest of the night. The sun hit at 06:20 local time on July 20th and the minimum battery state of charge was reached with 26% battery energy remaining at 08:04. Considering the additional payload power consumption (and mass) of 10W , this energetic margin – achieved on July 20th, and thus about one month after the solar solstice on June 21st – is a success and even exceeds the margin of 23% predicted by our simulations. To accelerate the battery recharging process, the aircraft also implemented autonomous thermal updraft tracking that was enabled during the late morning once thermal updrafts were encountered (see video). Multiple “free” altitude gains of >100m were achieved this way – again without any pilot interaction. The batteries were fully re-charged at 15:30.

T=26h: Final approach during the autonomous landing.

T=26h: Final approach during the autonomous landing.

As a final step, and after 26-hours of solar powered flight, the aircraft performed a fully autonomous landing at the Hinwil, CH airfield. Using its lightweight LIDAR (Light Detection And Ranging) sensor to measure its distance to the ground, AtlantikSolar – a hard to fly aircraft that can usually only be steered and landed by extremely experienced pilots – could safely perform the automatic landing. We consider the demonstrated full flight autonomy a vital step to allow search-and-rescue support teams, which usually do not possess extensive UAV flight training, to benefit from the significant advantages of solar-powered and in general high-performance UAVs.

T=26h. Landed autonomously.

T=26h. Landed autonomously.

AtlantikSolar AS-3 after its 26-hour fully autonomous solar-powered SaR mission.

AtlantikSolar AS-3 after its 26-hour fully autonomous solar-powered SaR mission.

The small operating post in Hinwil, CH.

The small operating post in Hinwil, CH.

Live-streamed RGB camera image. Our tents are clearly visible.

Live-streamed RGB camera image. Our tents are clearly visible.

Further information

Detailed design and technical information on the UAV platform can be found in “Oettershagen P, Melzer A, Mantel T, Rudin K, Lotz R, Siebenmann D, Leutenegger S, Alexis K, Siegwart R (2015), A Solar-Powered Hand-Launchable UAV for Low-Altitude Multi-Day Continuous Flight. In: IEEE International Conference on Robotics and Automation (ICRA)” [Download] .

An additional up-to-date publication, including the research for this flight, is planned.

Some more impressions from the flight were broadcasted live through our twitter account.

Acknowledgements

This research was funded through the research project SolAIR – Solar-powered Automated Aerial Imaging and Reconnaissance Using Infrared Cameras. SolAIR is a project funded under Armasuisse contract #043-12. The AtlantikSolar project is in addition funded with ETH Zurich’s internal resources, private supporters, and the European Union FP7 Search-And-Rescue research projects ICARUS and SHERPA .

Multiple project partners and collaborators have contributed towards making this important milestone possible, and we’d like to thank all of them for their various and ongoing support. Finally, we are grateful towards the Hinwil and Rafz model aeroplane clubs for providing the airfield!

Pilots: Rainer Lotz, Adrian Eggenberger. Development and Operations Team (Autonomous Systems Lab): Philipp Oettershagen, Rainer Lotz, Amir Melzer, Thomas Mantel, Bartosz Wawrzacz, Konrad Rudin, Thomas Stastny, Timo Hinzmann, Dieter Siebenmann, Dr. Gregory Hitz, Prof. Dr. Roland Siegwart.

The AtlantikSolar project was developed at the Autonomous Systems Lab at ETH Zurich


If you liked this article, you may want to read:

See all the latest robotics news on Robohub, or sign up for our weekly newsletter.



tags: , , , ,


Philipp Oettershagen is a research assistant and PhD student at ETH Zurich’s Autonomous Systems Lab (ASL).
Philipp Oettershagen is a research assistant and PhD student at ETH Zurich’s Autonomous Systems Lab (ASL).





Related posts :



Robot Talk Episode 101 – Christos Bergeles

In the latest episode of the Robot Talk podcast, Claire chatted to Christos Bergeles from King's College London about micro-surgical robots to deliver therapies deep inside the body.
06 December 2024, by

Robot Talk Episode 100 – Mini Rai

In the latest episode of the Robot Talk podcast, Claire chatted to Mini Rai from Orbit Rise about orbital and planetary robots.
29 November 2024, by

Robot Talk Episode 99 – Joe Wolfel

In the latest episode of the Robot Talk podcast, Claire chatted to Joe Wolfel from Terradepth about autonomous submersible robots for collecting ocean data.
22 November 2024, by

Robot Talk Episode 98 – Gabriella Pizzuto

In the latest episode of the Robot Talk podcast, Claire chatted to Gabriella Pizzuto from the University of Liverpool about intelligent robotic manipulators for laboratory automation.
15 November 2024, by

Online hands-on science communication training – sign up here!

Find out how to communicate about your work with experts from Robohub, AIhub, and IEEE Spectrum.
13 November 2024, by

Robot Talk Episode 97 – Pratap Tokekar

In the latest episode of the Robot Talk podcast, Claire chatted to Pratap Tokekar from the University of Maryland about how teams of robots with different capabilities can work together.
08 November 2024, by





Robohub is supported by:




Would you like to learn how to tell impactful stories about your robot or AI system?


scicomm
training the next generation of science communicators in robotics & AI


©2024 - Association for the Understanding of Artificial Intelligence


 












©2021 - ROBOTS Association