Detailed testing, OGN compatibility
- Andreas
- Jun 15
- 3 min read
Updated: Jun 17
On Friday, May 23, 2025, our partner MSW Aviation carried out an extensive test flight with two “Votec” motorized aircraft, HB-YLA and HB-YMV. The pilots performed a series of test cases at high speed, which we take a closer look at below. To reduce the risk of an accident, one machine was equipped so that smoke was continuously emitted to make the machine more visible. The following screenshot is from the GoPro recording from one of the two aircraft involved in the test:

A Maxx II variometer from our partner XC Tracer GmbH was installed in each of the two machines, running a test implementation of the OCAP firmware. Once per second, the firmware broadcasts an ADS-L data packet with an OCAP-specific extension. It also continuously receives ADS-L data packets sent by nearby aircraft and executes the collision warning algorithm. Warnings are indicated on the display and acoustically and recorded in log files.
The following video shows the recording of one of the test cases:
The following tests have been performed during this flight:
Flying towards each other from behind (rear-end collision), 4 tests
Fly towards each other at right angles and cross, 2 tests
Looping and approach when coming down, 2 tests
Laterally offset flight, 1 test
Pursuit without approaching the front aircraft, 1 test
Arc with collision course, 4 tests
A list with the individual test cases as well as the corresponding flight paths and reference output from the algorithm are available in the source code repository: https://github.com/aschweiz/project-ocap/tree/main/testflights/20250523_birrfeld
By using ADS-L data packets, the entire test flight could be tracked in the Open Glider Network (OGN), which has been able to receive and evaluate ADS-L data packets for some time now. As the flying object type “paraglider” was sent out, both aircraft appear as (very fast) gliders (7B and 79) in the OGN display:


We have extracted 14 test cases from this test flight so that they can be re-played in the simulation environment. This allows verify and analyze the collision warning algorithm in detail. The following screenshot shows test case 13 in the simulation environment:

Another big advantage of the recorded test cases is that they can be used in the future to verify changes to the algorithm. If, for example, a change or improvement is added to the implementation, it’s possible to run all 14 test cases automatically and compare the output of the automated test run to a reference output to see if something has changed.
The following screen recording shows how the simulation environment automatically runs the 14 test cases from this test set. For every test case, the simulator starts, loads and prepares the test and then runs it. Note that the time runs much faster when running the automated tests so that the result is available as quickly as possible. (The text that is visible for a short time between test cases is the log output on the command line.)
The algorithm performed very well in this test. The next steps are to make a few last small improvements and finalize the documentation. We are also planning to carry out a larger test with ADS-L and OCAP in Switzerland, more on this blog soon!
コメント