Today there was reporting of the Coronial inquiry into the 2008 level crossing crash in Tyabb, Victoria.
The big news is whether the train driver braked in time, were they looking, etc.
If you take a look at the satellite images you can start to draw your own conclusions about the view distance that a North bound train would have of a car approaching from the West especially given the new housing development and the tree on the corner.
However my main point is that the aspect that seems to be of interest to public discussions about safety, and too often the analyses themselves is, unhelpfully targeted at those simply closest to the occurrence.
If you were not there it seems that you have no involvement. However the risk of train collisions be they at a platform (like this photo) or at an intersection with a road is little to do with the decisions of the actors with physical immediacy and more to do with those who have broader influence.
In 2006 I proposed the following model to visualize the spheres of influence for the analysis of accidents. We need to get past the first circle. Those in the centre simply operate within the system created by others. Analyzing their decisions is only of use if we say “Ok, how can those with broader influence, change the system so that the decisions of those in the centre are not so critical”. Until we do this nothing changes.