Will a system that tells passengers where to stand on the platform be fully automated, or run by some sort of operator? In reality, it’s likely to be a little of both.
A base system will probably have an algorithm which optimises where people stand on the platform, determined by all data collected by the sensors in the area. This sort of system could run with no intervention.
But it’s not necessarily the ‘best’ system in all situations. Say someone falls down some stairs at one end of the platform – this base system might continue to send people that way, which isn’t ideal. This is where the operator could come in – someone overseeing the system could alter it to block off that entrance to the platform, or guide people away from that area. This has two benefits – by steering customers away from the trouble they get on their trains more quickly, and the problem can be more quickly solved because of a lack of congestion in the area at issue.
Another thing to consider is the expertise station managers and staff might bring to the system. An operator who uses it day in and day out will have a good understanding of when the station’s busy and where people stand (and don’t stand), and can alter the system accordingly. The way things work might be completely different in the evening rush as compared to the lunchtime lull, and having a human operator can improve the system’s operation by having some degree of control over it – perhaps even with different “modes” for different times of day.