I guess that's important in light of the chart you included on page 6 of your report showing the accidents from 2002 to 2012. It actually shows, in absolute numbers, a decline, and as was pointed out by several of my colleagues across, that was when rail traffic was greatly increasing—yet the accident rate has gone down.
To me, I think that would indicate, to a certain extent at least, that the SMS system has been working. Whether it's perfect or not is not the judgment, but it seems to be the case, based on those absolute numbers, that accidents are going down when volumes have increased dramatically over the same period of time, and thus that there is something to be said for the SMS system actually working.