Comment by meesterdude | Hacker News Reader

Comment by meesterdude | Hacker News Reader

Parent Full thread

> What are we left with? Not much really. In terms of time savings and safety, autonomous vehicles will not be a disruptive innovation to urban transport systems

Strongly disagree with the professor here. I'm not saying they will fix all our problems, and wont bring some new ones in addition, but to insinuate that driverless cars will not have a disruptive impact on safety and time saving leads me to question the authors understanding of autonomous vehicles.

Actually there are several other "facts" that author states that I disagree with, or that are separate from autonomous vehicles.

I think this stuff is certainly important to evaluate and weigh the pros and cons, and I agree its not going to be a silver bullet for all our problems, but this reads like a wonk who doesn't believe the internet will ever go anywhere.

In addition, who knows what additional technological or social changes take place that would impact how, when and why we travel.

TL;DR: author's magic 8-ball needs another shake.

> Actually there are several other "facts" that author states that I disagree with

Well, you could perhaps do as the article does, and state what your arguments are, and put some facts and figures to them. In particular, if you're going to use scare quotes on a word like "facts", you really need to say why those "facts" are, in your opinion, false. So far, you've told us you disagree but not made any effort to say why.

Peter Newman's bio is here: https://theconversation.com/profiles/peter-newman-1858

I'm not trying to make an argument from authority, but that looks to me like someone who knows his subject and probably isn't in the habit of shaking a "magic 8-ball". You've just made a series of assertions with nothing to back them up.

Unfortunately this seem to have become par for the course on HN. Disagree, but don't say why so no one can counteract your arguments. Anyone who looks at self-driving cars from an engineering perspective know that there are still many challenges.

Congestion is more a factor of tolerance than technology. As demand and tolerance for long commutes go up we might actually see less efficiency. [0]

It's still hard to match the human mind, especially without significant overhead. [1]

Cars are a compromise and aren't as efficient as pods or as fast as trains. [2]

Current electric vehicle technology can only drive you for about an hour at relativly slow train speed, at which point you'll have to charge for about as long. [3]

etc.

The car really is the weakest link in the "self-driving car". Which isn't unexpected because it's engineered around the driver.

[0] http://journalistsresource.org/studies/environment/transport...

[1] https://www.youtube.com/watch?v=kp3ik5f3-2c

[2] https://en.wikipedia.org/wiki/ULTra_(rapid_transit)

[3] http://imagizer.imageshack.us/v2/1024x768q90/203/modelsrange...

> Anyone who looks at self-driving cars from an engineering perspective know that there are still many challenges.

I'm not an engineer but (having started as an optimist about autonomous vehicles) I detect the whiff of vapor-ware about them. It seems everyone's saying how amazing they are (which is true if they work) but nobody has yet made them work, at least as a consumer option.

Shame more people don't quote references as you have. Makes for a more interesting conversation and gives an opportunity to modify a viewpoint or two.

And, if one looks at it from an engineering perspective, even self-driving cars don’t improve the passenger per m² per hour ratio of cars.

Self-driving cars can’t disrupt urban transport systems, because they don’t solve the issue.

The issue is that everyone wants their own, big, car, and everyone wants to use them at the same time.

During the morning commute you have over a hundred times more cars on the road than in the late evening, at 10pm, for example.

The only solution that allows the streets not to be completely overwhelmed by cars is to put the cars underground, or on elevated roads, or use more efficient systems.

40 cars vs. one bus. Self-Driving cars don’t help in urban environments at all.

Many cities already suggest to just use P+R instead of using cars downtown.

Firstly, you assert that everyone wants their own big car. This is not true. I don't want one, and I'm not alone in that amongst people living in cities with good public transport.

I choose my mode of travel (including cars) as a function of the time it's going to take to get there, and how much stuff I'm lugging. If I am carrying a lot or travelling out of peak hour, I take a car. If I'm under time pressure, I'll probably take the metro.

In Paris we have Autolib, which is a wonderful car service where you can reserve a car at one station and drop it off at another. There are over a thousand if these stations in the city, so there's normally one near you and your destination. But the brilliance of this is that I can go to work using public transport (because peak hour), gi to the gym after work and then drive home. Or catch public transport to a restaurant in a Saturday evening, and drive home, or whatever. Self driving cars will make this type of system even better because I can use one without having a station near me, I just call a car.

At the moment city centres get clogged by cars because public transport in the suburbs is horrible. But if you can get a car to a transport hub and then commute into the centre, I think a lot of people would prefer that to being stuck in heavy traffic for hours. Maybe not you, but many people.

I personally would prefer it, too – I actually don’t own a car, and live in a european city (Kiel, Germany) ;)

I’m trying to look at it from the american perspective, because I’ve long given up the hope that any non-american perspective is even considered relevant on here.

Actually, self-driving cars, while nice, won’t really improve your abilities inside the city – you’ll still use bus, metro, etc.

The "transport hub at edge of the city" is already a common concept over here, called "Park+Ride", with parking spots usually at places which have bus stations and train stations or metro stops.

>At the moment city centres get clogged by cars because public transport in the suburbs is horrible. But if you can get a car to a transport hub and then commute into the centre, I think a lot of people would prefer that to being stuck in heavy traffic for hours. Maybe not you, but many people.

That is actually the "first mile, last mile" concept the author of the article puts on the table as being an interesting problem to solve and considers autonomous cars to be, maybe, part of the solution.

https://en.wikipedia.org/wiki/Last_mile_%28transportation%29

They might help build more urban areas. Low cost per mile rides can shift a bunch of people that are currently making car first choices over to making car second choices.

This is what happens in much of the US, it is utterly impractical to not have a car, so people make their choices in the context of having a car. Make the car less important and there is at least a chance they will make different decisions.

Driverless cars will create less unnecessary queues due to faster reaction times, but yeah, that won't help all that much.

The author compares the safety and throughput of self-driving cars to trains, where they should have compared against regular cars. The problem with trains is the infrastructure isn't there and will not happen because it's impractical in an already developed metro area, and even if it magically happened the longer travel time would cause most people to still opt for cars. Self-driving cars will reuse the infrastructure that is already there and use it much more efficiently, so they are definitely a significant improvement.

You sort of have a point with the infrastructure argument, but I'm still getting a bit tired of seeing it. The roads didn't just spring into being. A great deal of money is spent on new roads and widening existing roads every year. If half that money went to rails, we would very quickly have an extensive and efficient rail system. As for space, one set of rails takes up about the same space as one lane of traffic. There are plenty of places where a lane could be taken off each direction of a freeway, replaced with trains, and the total throughout would be increased drastically.

Self driving taxis could be used for the last mile in those cases where transit really is too far from home or work.

Reply on news.ycombinator.com