Funded by Horizon Europe, the TRESSPASS project aimed to develop a risk-based border control model. It explored combining algorithmic video surveillance and processing of passenger data records (PNR) to detect "suspicious" travellers at air, land and maritime borders, to be submitted to enhanced security checks by border control officers. The following "suspicious" patterns were selected by the project to train the detection algorithm: "(a) mindless turn, (b) peculiar interest to the security units, and (c) avoiding the security units".