NEOD2

Space objects detection from the ground is a critical aspect of space research, astronomy, satellite tracking. However, the presence of noise from human activities, atmospheric interference, sensor artefacts and light pollution makes this task challenging.

During the first NEOD project (2025), LIST has designed, trained and benchmarked AI models for the automatic detection of meteoroids in the Earth’s atmosphere using radio data, captured by BRAMS (Belgian RAdio Meteor Stations), a dedicated network operated by the Royal Belgian Institute for Space Aeronomy (BIRA-IASB). These data were mainly provided during “quiet” conditions, with the usual meteor background activity and not during active meteor showers. The computations were optimized and parallelized on MeluXina (the supercomputer provided by LuxProvide). 

As a collaboration between Royal Belgian Institute for Space Aeronomy and LIST, NEOD2 will contribute to two main objectives:

• On one hand, the AI models will be fine-tuned and applied to a large BRAMS data set in order to provide a database of meteor echo detections. The first objective will be to combine these data with existing Python packages to reconstruct a lot of meteoroid trajectory and speed in the Earth’s atmosphere and help the BRAMS network reach operationability.
• On the other hand, similar AI models will be tested and validated with different data sets corresponding to more complex BRAMS data obtained during active meteor showers. These data sets will come from the citizen science project, the Radio Meteor Zoo. The goal will eventually be to obtain automatic detections of complex meteor echoes with a measure of the confidence level. If the latter is high, detection will be considered final. For those with a lower confidence level, detection will be uploaded to the Radio Meteor Zoo 2.0, a future modified version of the citizen science project where users will only have to validate or not the detections provided by the AI models with low confidence. 

All computations will be optimized and parallelized on the MeluXina supercomputer.

The resulting pilot will be used as an enabling technology in the value chain for the Space Domain Awareness, with the potential to be licensed to interested public/private partners for fees and may be adapted for other space-related purposes.