The CERN is running a Kaggle competition, TrackML, that is a key factor for the success of the HL-LHC project. They already know that the current solution for r

The CERN is running a Kaggle competition, TrackML, that is a key factor for the success of the HL-LHC project. They already know that the current solution for reconstructing particle trajectories from raw sensor data, based on Kalman filtering, will suffer from combinatorial overcharge if the number of collision events goes up 10x. Even if they extrapolate computing power going up in the next years. Thus their approach of looking for a better track reconstruction solution, based on ML. This is not just a science popularization contest, it's vital for making the HL-LHC succeed. Without it, the accelerator, the new sensors and data infrastructure will collect the raw data, but the physicians won't be able to process it, so they won't get any new insights. https://www.kaggle.com/c/trackml-particle-identification https://www.kaggle.com/c/trackml-particle-identification