It's easy to implement on a railway that's non complex (something like c2c) and uses the same (modern) rolling stock.
Actually the layout and rolling stock homogeniety is completely irrelevant for implementing ATO on an ETCS type line.
Essentially the controller reads the relevant information [current line speed, current train speed, next speed restriction, distance to next speed restriction] from the ETCS console.
It can optionally use an elevation map of the route the train is travelling over to determine stopping distance, or simply assume a worst case initial figure for beginning braking (assume a downhill gradient of the maximum on the route initially).
It would then, when the speed curve thus generated determined that braking must commence apply the brakes at the programmed initial value and continually redetermine speed and required braking effort, varying it on the fly to achieve a stop at the required position.
This will occur in the same way no matter how complex the layout it, all that matters is the distance to the next speed restriction, the gradients of the line (and even that can be excluded using a closed loop control system) and the initial and final desired speeds.
While the greatest benefits are only arrived at when the stock on the route is uniform it is not necessary for this to be the case for the system to be implemented.
It is not even necessary for the route to be entirely ATO operated.
This is what sets ETCS based systems apart from earlier systems such as those used on metro lines in the mid 20th Century which would just directly control each train using a central computer system - which did require those restrictions you mention to be practical.
--- old post above --- --- new post below ---
They won't be running ATO on those lines by the end of the decade no matter how easy you think it is to implement it.
Probably not, more down to our friends at ASLEF than any actually technical issues.
--- old post above --- --- new post below ---
Which level of ATO are you referring to though? As far as I'm concerned proper ATO is where the train does all the driving itself and has no requirement for the driver. I see no reason to implement ATO on the mainlines for the forseeable future. You won't save any money and it's not going to be that great a benefit.
This is often referred to as 'UTO' or 'Unattended Train Operation'.
I am talking about the type of ATO used on the Victoria line, where the driver starts the train with a 'Go' button but otherwise just waits to smash the emergency brake trigger if the train goes around a corner and sees an obstruction or whatnot.
If pattern recognition improves a bit more - based on the tech of those Google cars - then you might be able to take over the obstruction detection from the driver, leaving them with not much else to do.
Would probably mutate into a Guard/Driver hybrid where they remain in the cab as a final safeguard but spend most of their time doing a variety of Guard-esque roles such as announcements and helping with evacuations.
Either way, interesting times are ahead.
