Many of these people would be better served by having better funded public transit, especially in densely populated areas.

This would 1. help the elderly have more freedom. 3. help people who already use public transport. 4,5. It would make roads safer and allow for less space being used for private vehicles which makes the remaining spaces safer.

Improving intercity and freight rail can also help many people in category 6 who can now do something productive or fun on the train rather than risk driving long distance, and can take some long distance freight off of roads.

The only people who need to have private vehicles are those would live in rural locations, who would still benefit from traveling to a more distant urban area, and people who's jobs demand it such as tradies (trades people). These would all be served by having less people on the roads when they don't need to be.

Public transit doesn't run enough. In Japan, which has one of the most functional train systems in existence, staying out late (and not even very late!) means paying for a hotel room or post-drinking room (they serve you snacks and wash your clothes, etc).

This is in Tokyo.

Public transportation can't replace private vehicles. It can help. It can be useful. It can't replace them. It's like trying to replace traditional powerplants with wind - it doesn't work.

Do you think this problem really matters? There is a whole industry around capsule hotels anyway, I've spent a bit of time in Tokyo and no one is highly inconvenienced from staying out till 3am having a wild time on the town.

If you do get stuck, you do what everyone else does, get one of the probably millions of taxis to drive you home.

Let's face it the taxi driver will at least help you get out of the car :)

Lastly, many people who stay out that late are hammered and probably going to spew in the car, that won't be fun.

It would take a non-trivial increase in funding, I bet. Like maybe 5-10 times the funding we have now. Take Portland for example, which has relatively decent public transit for an American city of that size. For any meaningful trip within the city, it takes 2-3x as long on the bus compared to driving a car. They'd have to run a lot more buses to make it competitive. That's expensive.

Replacing the entire fleet of cars with self-driving cars isn't cheap either; so both require a significant investment. So does owning a car for that matter: both in direct car ownership, but also public cost in road maintenance, health care costs, parking lots, etc.

With fewer cars it's a lot easier and cheaper to get quicker and better public transit going. I don't know anything about Portland, but in general going by bus anywhere is mostly waiting in traffic (i.e. cars) instead of actually driving. This is why some places have bus lanes, which do help, but in a limited fashion.

I don't know how it compares exactly, but it's quite complex and there are a lot of factors and I'm not so sure public transport ends up being significantly more expensive in the end.

Trip length has more to do with stops and average speed rather than frequency, that is unless you are making transfers.

Trimet is competitive in downtown once you account for time spent looking for parking. The max is also pretty good when compared to rush hour traffic on the freeways.

Ok, so whilst I have some experience with US cities, this post is about the British regulation on autonomous driving. I can assure you that any substantial increase in local public transport (substantial being 25% or more, this is not a huge amount due to how much has been cut in the past decade.

Intercity transport really needs HS2 pushing through and increase electrification of other routes. This will speed up commuter routes and increase freight capacity.

The traffic benefits of autonomous vehicles are dubious.

I don’t see how there’s be a significant benefit for people reliant on public transport. If people are stuck on public transport, it’s typically because of cost, and there’s no way a small passenger vehicle is going to beat out a bus or train on that metric.

Such brigading is unlikely to have a meaningful impact on the capacity of non-limited-access roads (e.g., the street grid of Manhattan). It might help the capacity of highways leading into a city from suburbs or exurbs, but I don't think you'd even practically achieve a 2x throughput bonus at best, and any extra capacity you might add would be meaningless so long as the capacity of the surface streets at the destination is insufficient to handle the extra load.

Add in the extra load of deadheading of self-driving cars, and I strongly doubt that there is any meaningful traffic benefit and in fact there could well be extra traffic costs incurred... and that's before anyone accounts for the effects of induced demand!

They can hypothetically be packed closer together, however:

- What level of reliability is actually feasible for such a scheme? If you don’t have 5 9s, are you going to be causing catastrophic accidents? Is that level even enough? Even with near instantaneous reaction times and perfect judgement vs humans, you’re still limited by basic physics/stopping distance.

- cars still take up physical space, no matter how close the cars get you eventually run into a flow rate problem. With the addition of self driving cars deadheading to pickups, you’ll also have added traffic volumes from entirely empty vehicles.

7. People wanting to get home from the pub legally and safely at 3am.

You can't do that now?