Vincent Brückner left the board. Thanks to Vincent for all his work done for SearchWing as board member! René did an audit of the financials and he is also elected as new cash auditor. The activities and the financials of SearchWing e.V. are described in the Tätigkeitsbericht 2023 in german.

Event on technologies for People on the Move

States around the world are using a hodgepodge of surveillance technologies to counter unwanted migration. However, these tools can also be used to support refugees and asylum seekers: Small and large associations and many people in solidarity operate aircraft and drones for sea rescue, use data from satellites, experiment with sensor technology, analyse publicly available data or follow digital traces to bring human rights violations to justice. Other projects are developing apps to pool various sea rescue data or protect people from pushbacks. These technologies not only benefit refugees on their way to Europe; in many cases, the projects also want to change the political framework with their technical work.

At the event, we will give a brief overview of the new technologies used by state surveillance agencies and then show examples of how these can also be utilised to benefit People on the Move and social movements.

Friday, 1 December, 7:30 pm
Aquarium, Skalitzer Straße 6, Berlin
U1/ U3/ U8 Kottbusser Tor

Organisers: SpaceEye, SearchWing, CILIP

The SearchWing summer camp 2022 took place from August 9th to 19th, 2022 on Helgoland. With the excellent support of the Alfred Wegener Institute (AWI), we were able to carry out flight tests at sea. We were a total of eleven participants from Augsburg, Berlin and Hamburg with six students from the Augsburg University of Applied Sciences from the faculties of mechanical engineering, computer science and electrical engineering.

The Alfred Wegener Institute maintains the Wilhelm Mielck House on Heligoland as accommodation for guest researchers and we stayed there during the summer camp. The house has a well-equipped self-catering kitchen and a beautiful living room.

The Alfred Wegener Institute not only has beautiful guest houses but also a perfectly equipped laboratory that is equipped for the analysis of biological samples. For us, with the connected wet area, it was an ideal place for working on our drones.

We were able to use two ships from AWI for the tests at sea. The Aade is operated by a professional sea crew who took us to the desired flight zones for testing. With the “diving boat” we were able to drive to the closer test areas when there was little sea.

The aim of the flight tests was to test our drones under realistic conditions at sea. These were take-offs from the ship and landings in sea water, but also testing the range under sea conditions.

Because of the corona restrictions, a maximum of three researchers could ride on the Aade. The drone is launched by hand and then flies from a predetermined flight plan.

After the drone has landed in the sea water, it is brought back on board with a boat hook.

We had permission to fly for three test areas at sea. One test area is designated by the test center for maritime technologies of the Fraunhofer Institute for Manufacturing Technology and Applied Materials Research IFAM and has already been used there for drone tests. Another test area was around the N4 buoy north of Helgoland and one at the south port. With the friendly support of VfL Fosite Helgoland, we were able to use the sports field of the municipality of Helgoland for take-offs and landings on land. The flight tests from the sports field are easier because you don’t have to take the boat out to sea first.

All flights were performed within the pilot’s line of sight as visual line-of-sight (VLOS) flights. You cannot take off anywhere on Helgoland without a permit, since the Helgoland airfield and the heliport for the sea rescue helicopters of the naval aviation squadron are on the island and almost the entire island and the sea are designated as nature reserves. We have therefore obtained approval from the Schleswig-Holstein aviation authority for the flight tests.

During our flight tests, we also examined the possible range of the drone, among other things. On August 15th we did a flight test with “Elly” at a flight altitude of 120m. The next picture shows the map view of the evaluation of the log data from the flight. We flew in a triangle north of Helgoland with an average speed of 68 km/h.

We observed the battery voltage during the flight. Our drone has a lithium polymer battery with a nominal voltage of 14.8 volts and a capacity of 10,000 mAh. When the battery is fully charged, it has a voltage of 16.8 volts. The battery is empty at a voltage of 13.2 volts.

On this flight we flew a distance of 91 km in one hour and twenty minutes. The battery voltage towards the end of the flight was 13.7 volts. At around 12:30 p.m. we went full throttle for a short time. You can see the battery voltage drop from 14.9 to 14.4 volts. At an average speed of 68 km/h, consumption is higher than at a lower speed. The maximum speed during this flight was 103 km/h.

Most tests were conducted in sunny weather with relatively little wind. Towards the end there were more waves and also more wind on two days. In total we performed 13 flights during the 9 days with 8 hours in the air and a total flight distance of 461 km. The two drones passed the tests well and were further tested in subsequent tests in Burriana. The SLA printed electronics boxes with the M8 connectors for the servos withstood the salt water. According to the manufacturer, we used waterproof servos from Traxxas – they show corrosion damage on the screws relatively quickly. On the Extron motors, we swapped the bearings for VA bearings. We no longer have premature fill-ups due to salt water.

We would like to thank the Augsburg University of Applied Sciences for the financial support of the students with the costs of the excursion, the Alfred Wegener Institute for the friendly reception, the IFAM for the friendly reports in connection with the approval, the municipality of Helgoland for help with the flight permits and the permission to use the sports field for the flight tests, the aviation authority Schleswig-Holstein for the patience with the approval, the lower nature conservation authority Pinneberg for the timely assessment of the impact of the flights on the flora and fauna on Heligoland despite the lack of staff, the airport, the SAR branch and HTM Helitravel for the daily individual releases and Verein Jordsand for advice on how to deal with the many birds on the island.

Friedrich Beckmann

Feel free to join by following the registration link.

Following is a copy about the contents of the meetup:

Following our 26th Conference THE KILL CLOUD (25-27 March) we are hosting a follow-up meetup in collaboration with SearchWing. SearchWing develops and builds unmanned aerial vehicles to find people in distress at sea.

About Searchwing

Everyday people flee war, persecution and poverty and are looking for a place of safety. On their journey they have to cross the sea such as the Mediterranean Sea or the Atlantic Ocean. While crossing the sea they are in great danger. NGOs like Sea-Watch e.V., Sea-Eye e.V or Open Arms do search and resuce at sea.

The SearchWing project develops unmanned aerial vehicles and systems to help these organizations improve the search and rescue efforts at sea. The automatic SearchWing UAV is equipped with two cameras and can cover an area of 100 square kilometers at an altitude of 550 meter. The UAS has already completed several trial runs in the Mediterranean Sea, and in December 2021 it was used during an actual mission near the Canary Islands in the Atlantic Ocean.

SearchWing helps all humans in distress regardless of origin, nationality, religious or political believes. They support the request for safe and legal pathways and support the request to establish a European search and rescue organisation.

Twitter @srchwng

Matthias Monroy

Knowledge worker, activist, editor of the German civil rights journal Bürgerrechte & Polizei/CILIP. First in touch with military and police repression in the early 90’s. Focussing on policing in the European Union, migration control, undercover cops, internet monitoring, surveillance and interception technologies, police gadgets, satellite intelligence, drones. Publishing in left-wing newspapers, online-media and netzpolitik.org.

During the meet-up

• We will start with a general introduction to the situation in the Mediterranean Sea and SearchWing’s activities
• An introduction about the use of drones by Frontex and EMSA in the context of border control by Matthias Monroy, giving an overview about the situation in the Mediterranean Sea.
• The second half of the presentation will give an overview about the SearchWing’s UAV/UAS and show how SearchWing builds a UAV with off the shelf components and free software.
• After a Q&A and discussion session participants have the chance to take a closer look at the UAVs and the software system such as the boat detection system.

The meetup is part of Disruption Network Lab’s all year round community programme

To collect as many images of boats on water as possible, a 2 km x 2 km area was flown over with different approach angles and heights. Such a flight pattern as shown in Figure 1 has a total flight distance of about 70 km. A real search pattern would have a distance of 2km between flight lines, since we cover a search range of about 2km with our cameras, i.e., we search a strip about two kilometers wide below the drone with the cameras. So a flight with a flying distance of 70km covers an area of 140 square kilometers in a proper search mission.

The target flight height of the drone is 550m above sea level. In 12 flights, different tests were flown at 200m, 400m, 550m and 700m at different times of the day and weather conditions. The amount of data collected comprises 110GB in about 23000 images. In the following picture 2 there is a inflatable boat next to the Sea-Watch 4…. (see original picture).

The images have a resolution of 3280 x 2464 pixels and are captured with a Raspberry Pi V2 camera. There are two cameras on the drone, each with a 30 degree downward sideways angle. This is the first time we have taken pictures from the target height of realistic boats under realistic conditions.

Figure 3 shows an enlarged section of figure 2. In this case, too, there are two boats on the picture detail. The evaluation of the images is supported by SearchWing image analysis software, which is adapted to the evaluation of images from the sea.

Figure 4 shows the view of the photo in the SearchWing image analysis software. The software is web-based and can be easily accessed in a browser. The images shown were taken in high winds and it can be seen that the software highlights many waves and white crest as possible boats. For each image, the drone’s position at the time of capture, the drone’s direction of flight, and the orientation of the image with regard to the compass direction are shown.

Figure 5 shows the image section without any markings of the automatic object recognition.

Figure 6 shows the image section with the markings of the automatic image evaluation. With the newly collected data, the machine image evaluation can be further improved.

As during the test mission with Resqship e.V., the take-off and recovery of the drone proved to be unproblematic. Problems with the transfer of the mission plan and the parameter download still need to be investigated. Feedback from the crew on Sea-Watch 4 has been positive and helpful. According to crew members, the drone can be quite helpful in searching for shipwrecked people compared to searching with binoculars. The waterproof housing of the battery is not yet satisfactory. We are very happy to have now, for the first time, real photos taken by our drone system on the Mediterranean Sea.

We would like to thank Sea-Watch e.V. for their support during the drone tests at sea.

The 4 planes are packed and are now getting shipped to the port where the ship is already waiting.

Our pilot Jonas will also set off next week and then accompany the mission on board. We hope to be able to support the crew on the rescue ship with our planes in the search for people in distress.

BR24 also reports on the use in this article: https: // www .br.de / news / bavaria / regensburger-seenotretter-set-drones-an, Rs9iXuD

We also gave a small lecture about our project and the current status, the lecture
is available online:

Hartmut Gieselmann from the c’t editorial team then interviewed us and that’s how it is
came to the publication on c’t:

https://www.heise.de/newsticker/meldung/36C3-Open-Source-Drohne-der-Hochschule-Augsburg-sucht-Fluechtlinge-im-Mittelmeer-4624125.html

The design is done with OnShape CAD software and is available online. I used the following CAM software:

• Deskproto 7.0 for the 3D CAM. This software reads the STL file and produces G-Code.
• Step-Four XpertMILL is the CNC control software

The fuselage is manufactured with the Step-Four XPERT 1000-S mill in the Modellbauwerkstatt of the faculty of architecture and civil engineering.

I cut a smaller EPP block with RG20 density with a styrocutter from a large block. The size is 155mm x 75mm x 900mm.

The milling is done in three steps. First the EPP block is flattened to a height of 70mm. This done via 2.5D milling, i.e. it is derived from a 2D sketch rectangle where the height is set in the CAM software. Then the 3D milling process starts and the inner side of the fuselage is milled. After that is finished, the block is turned around and the 3D outer shape is milled. Some more effort is needed to control this turn, because the CAD data expects a block width of 150mm but in fact the width was 155mm. So I manually moved the data in the CAM controller up by 5mm.

The fuselage milling process takes about 3 hours. I depends pretty much on the step width during 3D shape milling. Smaller stepwidth means more precise shape but if you move the drill only 0.2mm in each lane then it takes long. The flattening in the beginning was done with 80% of the drill diameter as stepwidth and only took about 5 minutes.

The EPP material is not very stiff compared to the EPO material from the mini talon. The milled fuselage part has a weigth of 30g while the mini talon fuselage part has a weight of 60g.

Petrosilius made an introduction to the challenge where he described the team setup and the problem setting. The SearchWing drone takes about one image per second and most of the images just contain water. The images are currently classified manually. It would be a great help if the image classification could at least be supported with computer vision. If a classification could be done during flight on the drone, then the team on the ship can react faster to a spotting.

More than 50 volunteers participated in the hackathon.

Everything happened in “The Drivery” which is located in the Ullsteinhaus in Berlin.

A place which fits perfect to hipsters of all kinds. The location includes kitchens, gym, theater and

screens for doing all kinds of important things. Food, drinks and the location were provided by Deep Berlin.

A get together barbecue took place on a roof terrace where the team members could get to know each other and discuss the plans.

Already the barbecue with papers on the table…

On Saturday we started with a breakfast at nine. The teams started hacking in a hackathon room where tables and screens were provided.

There was also a NVIDIA DGX GPU server available which was prepared with Docker Images where each team could run trainings. An NVIDIA support engineer was on site to help with the station. Also thanks for that support.

So you could hack behind flowers. There were seven teams planned but there was nobody on Team G and Team F was a one person team. The majority of participants was more interested in image processing and machine learning.

On Sunday morning the programming continued. The final presentations were scheduled for 15.00 at the big theater. This screen is actually a screen and not a beamer. All team members were on stage when the results were presented.

SearchWing Augsburg will establish a community for all the volunteers who want to continue to support us in the future with computer vision for civil search and rescue missions. We are currently collecting and summarizing the results from the hackathon. The plan for the community is to have

• A mailing list for the SearchWing Computer Vision group
• A github location to integrate the results
• A regular developer phone call and a physical meetup in Berlin

After preparing and running the hackathon Petrosilius is sleeping for the next four weeks…

I had a very warm welcome and during the discussions the usefulness of the drone was emphasised by RESQSHIP e.V.. So I hope that we can support the missions soon with a new release of our rescue drone.