The objective is to use machine learning to identify magnetospheric boundaries (specifically planetary and interplanetary shocks crossings and magnetopause crossings) in spacecraft in situ data. The boundaries are identified by a discontinuity both in magnetic field and in the spectrum of high frequency waves. These two types of measurements are available on many planetary missions. Data from Earth's missions (Cluster, THEMIS) can be used for training.

As a first result, we strive for successful identifications of the boundaries with a good accuracy. The perfect result would be a classification of boundaries (shock, magnetopause, inbound, outbound) and the correct handling of multiple crossings.