Comment by xnx

Single passenger cars still have a problem with density even if there's no need for a driver. Combining multiple people into one trip can help, but also lessons the utility of Waymo if riders have to go out of their way to pick up additional passengers.

Getting 1000 people downtown could be up to 2,000 Waymo trips (one trip to drop off the worker, another trip for the car to go back out to pick up another passenger). While one of these 56 passenger very light rail cars can do it in 18 trips. A light rail vehicle like the Siemens trains used in San Francisco can carry up to 200 people at crush loads, so that's 5 trips.

You mean it's very difficult to compete with a company that is massively subsidised by public infrastructure? That's what really killed freight rail in most of Europe, make the train companies pay for track maintance (often the rail companies even want this because it keeps competition out as well), while trucks atpapy very little of the cost they impose on the public (i.e. much higher road usage, causing most of the traffic issues).

If this were true, cities would have abandoned mass transit for taxi system decades ago. The requirement for there to be a warm body driving a taxi isn't among the prime causes of its inefficiency.

The big AV transit efficiency gains* can only happen when nearly all human drivers are removed from the road. Alas, that's at least 20 years ahead of now**, or more (e.g. if tech stalls for some reason, though I consider it unlikely). Otherwise they'd be limited by having to account for human drivers and that limits speed and throughput enough so other solutions are a must.

* Think having much higher speed limits (as far as humans are concerned, nonexistent), or mass coordinating movement over the entire traffic.

** We can reasonably estimate the minimum without bothering to ask how fast the tech will improve: Even if the tech were available now, think about fleet replacement costs which no one group would be too eager to pay. Best case, it's the typical 'make a concentrated pressure group lose for societal benefit' and we know how that politics goes. It will happen, but slowly.

*** Another thing to account for is that there's no good reason to design an AV car like a normal car, and there'll be some iteration time over that too.

Each Waymo vehicle is probably close to half a million USD in just hardware cost.

I don’t think fixed route transport infrastructure is going to have trouble competing on efficiency.

Public transit isn’t supposed to have a financial “ROI.” Are highways expected to turn a profit from tolls?

It is really efficient, at sitting in traffic with all the other cars 8-/