I think it’s worth mentioning that while the IT system couldn’t coordinate the trains safely, it was a management decision to stop them from running. Coordination can happen myriad ways. For example, trains could operate according to their normal schedule, and if one needed to deviate they could radio in, and an operator could let downstream trains know to proceed accordingly. Maybe simple cell-phone GPS pings could suffice. Perhaps the state of the system was in such chaos that such orchestration wasn’t possible with the staff on hand. That said, a truly resilient system should have multiple coordination channels to fall back upon.