I'm unconvinced, simply because wherever it has been done many to many demand responsive transport (DRT) has been an expensive failure because you simply don't actually get the journey synergies. Sure, you save your staffing cost, but with the lack of staff comes the likes of vandalism, littering, vomiting and other antisocial misuse.
Errrr have you noticed the multiple cities where autonomous taxis are running? All of the above issues are easily handled because the person hiring has to use ID and it's pretty easy to do a before and after image of the vehicle then use AI to detect vandalism or soiling and then bill them. If that doesn't work, then the next user just reports it.
You also have the disadvantage of there being no guarantee a car will be available at the required time. I can quite happily leave the house 15 minutes before a train because the car is just there if I'm driving (20 if cycling, 30 if walking). But I need to go with 30 if using a taxi as there might or might not be one available, so I only ever do that if I have luggage, not as a time saver.
I covered this in my post, you cannot use anecdotes related to your current usage of human driven taxis to work out what the response to ubiquitous autonomous vehicles will be. The two are very different beasts in their final form.
At the moment a tiny fraction of cars are taxis and those are still driven by people who have operating hours. The first impact of proper self driving will be to make taxis dramatically cheaper which will massively induce demand on its own. The more taxis there are the closer any one will be. When you are at the point where a double digit percentage of car journeys are done by shared autonomous vehicles then the number of cars is massive. Let's do a thought experiment?
Let's say 10% of car journeys are done by a shared autonomous fleet. This fleet of cars has a utilisation rate around 10 times higher than the current fleet. This would mean that in the 400m around where you currently are 10% of the cars currently moving would be a shared autonomous car and 1% of the parked fleet would be one. How many cars are currently within 400m of you?
For me in a suburb of a medium sized city the population density of 3600/km2, there would be just over 1000 cars parked up in the 0.5km2 that a 400m circle around me represents. So at this 10% market penetration and 10x higher utilisation there would be 10 cars currently sat within 1 minutes drive of me.
In your specific use case of a taxi to the station within an autonomous taxi service I'm sure there would be a book and wait function.
I thus cannot see the model of owning or leasing a private car changing, nor can I see the model of large buses and rail changing very much even if we might end up moving to a "guard only operation" type model and perhaps lose a few branch lines. While potentially cheaper than ownership it just has too many disadvantages. I doubt it will get much further than the removal of some second family runabouts.
See above, the autonomous taxi has been here for a few years, it is gradually rolling out, all of the issues are being worked through. It's probably where the internet was in the late 90's where people with similar viewpoints to yourself failed to see any usage cases for it in their lives but non the less tens of thousands of people were still working on it, they aren't going to give up, there isn't an issue that hasn't got a solution to it and the solution will be found because of the inherent economics of autonomous cars.
The biggest cost of motoring is buying the vehicle, the biggest cost of a taxi is paying the driver. Autonomous cars will be able to drive at least 10 times further in a year than a regular car, the cost per mile will be dramatically lower. Owning a car is either the biggest or second biggest cost in most houses, people aren't going to keep paying thousands of pounds a year paying finance and insurance on cars that they then have to park inconveniently in car parks away from wherever it is they are going.
It doesn't matter that most people wouldn't give up their car today its a rolling process. The taxis arrive, people get used to them, the cost comes down people start using them a lot, some people start getting rid of second cars, they stop buying new cars, they notice they haven't used their own car for a week or two. It's like land lines, most people didn't make a conscious choice to get rid of them they just stopped using them at different rates.
And for intercity rail, it is simply more efficient in every way to move 1000 people on one train with one driver/guard than even say 200 autonomous vehicles. In reality the latter will just replace taxis to get to/from the station at each end but everything else is likely to look largely the same.
Point in principle, for all this talk of intercity train's efficiency an EV with one person on board has a lower cost per mile. An autonomous EV that is utilised at an even higher rate will beat it by an even greater margin. The real threat for intercity is the point when we integrate dynamic pricing into the model and all long distance car journeys are conducted without congestion. Again we don't needed buses to do this, just a portion of the most cost sensitive users to share vehicles.
At this point rail needs to be able to go much faster than cars door to door and needs to have the level of utilisation whereby the cost per passenger mile can be remotely comparable to driving. We aren't going to be able to operate a mixed traffic railway, everything is going to need to be high performance and high frequency.
If we aren''t talking about autonomy it's like discussing whether Blue Ray vs HD DVD would win in 2007, when it was pretty obvious that the answer was internet video. The same sort of arguments came up there about ownership of media, unreliability of the internet or the roll out of broadband.
