As someone currently working in the Tech industry, I fully agree with this. If anything, I can see an AI-based automation system for the railways that would actually work for a mixed traffic railway within the next 10 years.
People have been saying that since the DLR opened 35 years ago, and probably since the Victoria line opened over half a century ago.
They have completely unstaffed trains ( i.e. no on board staff at all ) in Paris on Metro lines 1, 4 and 14 and the first of these has been thus for 25 years ( Line 14 opened in October 1998 ) - but these are totally self contained lines with no junctions and everything calling at every station; the difficult bit with full automation would be to avoid issues with junctions and differential service patterns.
Yes, I know the DLR has junctions but the service patterns are still much simpler than on the main line - there is no differential performance, no looping, no mixture of stopping patterns.
Incidentally, all the highly publicised incidents with the DLR in its early years occurred while the trains were being manually driven.
I'd agree the self driving cars are proving to be a bit of a disaster.
Thing is, as per the article I linked to above, the question is what does “driverless” mean, and what is it intended to actually achieve?
Only a tiny number of legacy metro systems have been converted to full unattended train operation (the Paris lines are an example of this), which is where the staff savings would be. This is likely to require massive investment in universal platform edge doors etc. and is many decades away on the tube, let alone the mainline.
If your version of “driverless” means you still end up with a member of staff on the train (either in a cab, a la LU, or roving through the train as per DLR), you have incurred all the expense of conversion only to wind up with the same number of staff required as a traditional DOO mainline train. Those staff will expect to be paid a salary, will need to be rostered, to have breaks and may well be unionised etc. So the savings are unlikely to justify the (massive) cost of conversion.
There are also issues with ATO still not necessarily being able to beat a driver in mainline conditions, as opposed to LU type situations.
@bramling has usefully expanded on these matters before, in particular in the thread
@yorkie linked to above:
This is not a discussion about the rights and wrongs of driverless trains but the technology which they require It is nearly 50 years since driverless trains were introduced on the Victoria Line and over time they have slowly been introduced on metro lines around the world including systems...
www.railforums.co.uk
So it’s pretty easy to see why there’s no real impetus to push ahead with “driverless” - least of all at a time when the railway is being starved of funding.