The study consists in observin a sample of paragliders released from a zipline on their own harnesses. Before being released, the pilots simulated catching a sail by synchronizing the brakes with flashing LEDS and were put in cognitive cognitive overload position by reciting a series of words. It is important to note that all of the pilots had undergone static handle access training before the test. Mat, the study’s director, points out that the delays would have been significantly higher in real situations. In many releases, we first notice a tilt of the body with a grip on the risers, even a good brake from the left hand. 80% of pilots turn their heads towards the handle and in 85% of cases, the hand arrives reflexively on the hip. In all cases, the grips are quite good. The natural movement of the throw is backwards keeping the close to the body, even if the parachute is block. In this case, the paraglider will change the position of the grip to use your biceps. The study shows that a slow motion in two stages throws it out of the reserve parachut. The inertia given to the pod by the internal movement may even lead to entanglement suspension lines of the parachute. In real conditions, the pilot suffered a serious shoulder trauma. 16% of the pilots had a belly parachute. In this configuration, if the access is to the handle and the movement is more fluid, this advantage can be counteracted by an an upward movement of the the emergency pocket thus reducing the efficiency of the extraction. Sometimes, if the strap connecting the handle to the pod is too long, extraction is more difficult and slower. 1. The study clearly shows then to fundamental under stress, in particular by movement of the arms close to the body. The ejector, dissociating the the triggering of the ejectio, makes it possible to position a handle close to the body and clearly visible. 2. The study can’t make a case or throwing in any particular direction or velocity, that depends much more on the configuration of the main canopy.At least the ejector will quickly move the pod away from the harness. this highlights difficulties in extracting the pod, these these difficulties being countered by vertical movements of the arm. It is clear that in real situation some pods could not have been extracted.Also, ejector has been calibrated to be effective at more than 4 G by a thrust of about 70 daN.