I don't see much technical difficulties They really are there though, even though they might not be immediately obvious! The thing is, this is a project which

I don't see much technical difficulties They really are there though, even though they might not be immediately obvious! The thing is, this is a project which essentially ties together two existing regional railway lines through a new underground core. Those two systems have existing rail traffic on them and different signalling systems on each. It's already been a nightmare to implement three signalling systems and transitions between them – a project to provide fully automated operation on two other railways lines in addition to this would ramp up the complexity even further. Fully automated lines are totally viable – the DLR system in London works in the way you suggest and has for some time. And I'd bet that if the Crossrail project had been a new and isolated line it probably would have been using full automatic train operation throughout. But it's a hybrid between a regional railway and a metro line – ATO is way harder.