Hacker News Reader

Comment by DiscourseFan | Hacker News Reader

Ok but an autonomous bus, or autonomous vehicles designed to fit different numbers of people, would not be better than a one size fits all bus?

View on HN

The one size fits all bus would be better. A tiny car cannot share rides without adding big delays to the person already on it. That means you cannot count on getting there on time with a shared ride. The bus has defined delays so you can predict when you will get there and this arrive on time.

If a root is only serving a smaller number of people, why have an entire bus? Busses make sense for thoroughfares, but I often board nearly empty express busses that are only running because of state/city appropriations. If we could modulate form to meet demand precisely, like Uber but for all transit, it would be far quicker and more efficient.

Even in the suburbs it is not hard to get a bus that is half full during off peak if the service is good. Most service in the US is not good so perhaps you are only looking at one of the many bad examples and incorrectly thinking it must be that way - it doesn't.

> . If we could modulate form to meet demand precisely, like Uber but for all transit, it would be far quicker and more efficient.

Efficiency is not the only or even the primary goal. The primary goals are to "Get people from where they are, to where they want to be, when they want to go, in a reasonable amount of time, for a reasonable price.

Your modulate demand fails at that. A fixed route bus is predictable. I know in advance when it will be there to pick me up and where/when it will drop me off. That means I can plan my day. With Uber or any other demand/response system you don't get that and have to hope every time that it will get you there in time.

Quicker is much harder to quantify. A bus can connect to a faster rail (which potentially can go 300 km/h - only a race car can do that). A bus can be given signal priority while the uber needs to stop and wait more. However a bus will stop to pick up others on the way. (but if your "modulate to meet demand" does that it is worse than either a bus or uber because you can no longer predict when you will get there). A bus network done right is also likely to drive changes in your city such that most of the places you go are where the bus is going and that makes it just as fast.

You must not live in a place suffuse with Waymos. At 3am anywhere in the coverage area you will get a car coming within 2 minutes, often for at least 20% less than an uber/lyft, and sometimes up to 50% cheaper. There is no way a bus can be that efficient, be that quick, or, in a large network, as coordinated. You could even set up basically a series of pods that merge to share energy and split up depending on where they need to go. Every person could have their own autonomous modular form of transport that is always ready at hand. How is that slower/better than traditional public transport? As I said, the only technology that supersedes it, and only because you cannot overcome friction, is maglev trains.

If only we could build a network of high speed, maglev trains covering the span of the entire planet, from the southern tip of Chile, through the Darien Gap, across the Bering Strait, all the way up to Paddington Station and down past the Suez Canal all the way to Johannesburg; that, plus airplanes and modular autonomous transit, should cover 90% of transit needs, and anything more than that will be served by older technologies, off-road vehicles, horseback/camel, and of course for the most treacherous environs, on foot is still best. But I cannot imagine a world where a traditional bus system would be superior to a highly coordinated system of autonomous modular transportation.

Reply on news.ycombinator.com