Hacker News Reader

Comment by J0_k3r | Hacker News Reader

we have self-driving cars already, they're called trains

Show me a train that can show up at my grandmother's door and drop her at the doctor's.

Swiss somehow manage to use trains in massive numbers even if grandmothers need car assistance occasionally. You don't need to poison the atmosphere every day with a car on your commute just because your grandmother might need a pick her up for her doctors checkup (having a medical facility come to pick someone up is not an uncommon service either).

Why do some of you think that finding a wierd edge case is enough to invalidate a whole infrastructure? This isn't derivation of a mathematical proof where counter-example invalidates the hypothesis.

"Why do some of you think that finding a wierd edge case is enough to invalidate a whole infrastructure?"

Last year I was nearly going broke spending money on Ubers when my then girlfriend's kidney's failed. She wasn't considered immobile enough to qualify for the transport to and from Dialysis, but would be drained enough that walking to and navigating public transit home after would be impossible for her. So this is not just weird edge case for me, this scenario was a constant part of my life.

Not really a good example, Swiss rails are supremely expensive to state and travellers too. Most countries wouldn't be able to afford such density to have good coverage. They definitely don't run very precisely in many places due to many factors, as a local I can attest to that with many late arrivals to work in the past.

These days they run almost empty at huge losses. You still need to haul 300 tonne of steel through whole country, even if there are 5 passengers.

Maybe its the best implementation of the train network for civil travel, but still needs to be heavily complemented by network of buses to all small places.

And that grandma drive to the door would still have to use a taxi every single time.

> Why do some of you think that finding a wierd edge case is enough to invalidate a whole infrastructure?

Why do you feel you need to talk about a different infrastructure when people talk about the one you yourself acknowledge is needed to transport people to/from trains?

>poison the atmosphere every day with a car

So you won't mind electric cars then, which fix that problem ?

They don't fix the problem. Nowhere on earth is there a sustainable electrical grid that doesn't pollute the atmosphere, let alone one that supports electric cars for every household. Not only that but electric cars create local pollution in the form of particular matter from brakes and tyres and they are worse than internal combustion engine cars in that regard.

> and they are worse than internal combustion engine cars in that regard.

Regenerative braking means less wear on the brakes, and there's no particulate matter from the exhaust pipe. In what way are electric cars worse?

Regenerative braking does not replace traditional brakes. In particular it is only really efficient at high speeds and doesn't do much in cities. EVs have no tailpipe emissions, but particulate emissions from the tailpipe of modern ICEs is very low anyway. But EVs do produce significantly more emissions from the tyres and brakes compared with ICE vehicles because they are heavier.

https://www.emissionsanalytics.com/news/2020/1/28/tyres-not-...

It doesn't have to be 100% regenerative and an electric motor in generator mode it's very efficient at braking even at high speeds. One just needs to redirect the excess energy that cannot be absorbed by the battery to a controlled resistive load and trigger both mechanisms if the brake pedal is pushed more than 30% in which case you definitely need regular brakes. Trains already use these mechanisms.

The figures in the linked article are overblown: 5,8 g/km means one throws away a pair of used tires every 1000 km, which is not the case since the average tire lifecycle is 20k to 30k km.

> doesn't do much in cities

Incorrect. I drive a lot in the city and I use the friction brakes so little that they start malfunctioning after a week or so if I don't do some hard stops once a week or so to clean up the pad surface. I only use the 'regular' brakes if I am stopping from high speed, quickly, or at the bottom of a steep hill. Probably use regenerative breaking for >90% of all normal stops.

Trains use lots of metal in their construction, and even electric ones might use fossil fuel generated electricity.

I’ve seen plenty of elderly people use trains.

There’s a train outside my front door that will drop me off at the front of UCSF. But that’s about 10 stops away and requires a transfer. If one doesn’t mind the bus, there’s one that picks up in front of my door and drops off at the hospital (the bus stop is even closer to the main entrance than the parking garage!)

That's great! How soon can that train get to Talanes' grandmother's house?

What's always been amusing to me is that trains should be self-driving, but they always have a human in charge. I mean it's not like they have a steering wheel.

They don't always. There are self-driving trains operating full time in London and Paris as well as many other cities around the world. But they are low-speed trains operating at only around 30mph. High-speed trains run in excess of 180mph and heavy goods trains are often a mile long and weigh more than 10,000 tons. While much of the signalling is automatic these days, I think it will be a long time before we could think about running something so dangerous without a skilled human in the front ready to override the controls if necessary.

If there's between 100 and 1000 people in a train, getting rid of the driver saves between 1% and 0.1% of an hourly salary by ticket. The economic pressure to do so is very low, compared to the difficulty and legal liabilities.

Automating out a taxi/Über driver saves 50× to 1000× more money per traveler. And the infrastructure for a car ride cost way less than train + track + catenary lines + stations + antiquated ticketing system etc., so salaries represent a higher percentage of total costs.

Roads cost $2 million per mile. Tracks are half that. Maybe we should make car users pay for their own roads, and see how the economics work out.

I believe the main reason is the long tail of unexpected events. For instance: I regularly use the tram. Every few months or so, the driver has to get out to manually operate a railroad switch that does not react to the wireless control signal. If you had a driverless tram, the tram would have to wait for personnel to arrive at the switch for the manual override. Extrapolating from my experience manual overrides are probably necessary a few times a day across the entire city's tram network, so we'd be looking at massive interruptions multiple times a day that cause cascading delays throughout the whole system.

So unless you put in a lot of engineering effort to get rid of these unlikely-but-still-common problems, running a driverless operation is going to be more disruptive and less cost-efficient than paying the driver salaries.

It's also a cautionary tale for self-driving car advocates. Running a driverless train on segregated track with signalling is far easier than running a driverless car on a mixed street with lidar.

Yet it still has enough edge cases to mean it's not worthwhile

Single-occupancy trains are a terrible idea though..

Jokes aside, the dependence on self-driving OP was alluding to is wildly exaggerated in my eyes, because the other kind of self-driving (user drives a short-term rental) works just as well. And with short term rentals, physical footprint is making an actual difference because their availability is bounded by storage.

There used to be a wave of microcar concepts that were trying to solve parking by folding up a little. Back then those never made sense because they were still intended as personal vehicles and those end up taking up a one-size-fits-all spot anyways. But combined with the innovation of micro-rentals (modeled after docked bike-share) those fold-up tricks could turn from a gimmick into a key enabling technology. All you need is self-parking to store them in a compact FIFO and self-parking is beyond solved. Once you have that, all minor extension of the unparking range (from unparking just enough to unfolding at the storage dock to dedicated pickup locations nearby to pickup anywhere) would be incremental improvements to service quality that would require much less than full autonomy.

> because the other kind of self-driving (user drives a short-term rental) works just as well

User drives short term rental gave me all kinds worries which I don't have driving my own car. Being responsible for checking damage when picking it up, worrying about minor (as in accidentally kerbing a wheel) damage while I'm driving, having someone hit it while it is parked somewhere, getting caught in traffic while returning it and being fined for going over the allotted time.

Writing this down has just made me realise how nice it would be if there was a car rental company which worked on a first come first served basis with rental charged per minute. If they also had some kind of automatic photo booth to take detailed before and after photos that would be almost perfect.

i think you mean taxis?

To the end customer, the only difference would be cost, and if you are rich enough, or the service is subsidized enough by investor cash, or the cost of living in your area is low enough, then it all ends up roughly the same.

Yep, that's what I post every single time. Taxis are already self driving cars, so any argument for a self driving car should apply to them as well, yet people still prefer owning their own car than being "automatically" driven everywhere. It would already be cheaper for me to take a taxi to work every day than pay for my own car, yet.....I would rather pay a bit more and have a car that is mine and that I can do whatever I want with. The self driving aspect is just not appealing in the slightest.

I think most Americans will choose the subscription service over ownership. Subscribers will have instant access to any type of vehicle at any time. The subscription cost will be lower than the cost of ownership. The whole fleet will always be new and incorporate the latest technology. Robotocized cleaning centers and user ratings will mitigate “tragedy of the commons.” I think even at the high end people will use luxury subscription services more often than whatever car they may own as a backup.

It’s not cheaper vs. better. Subscriptions will be both cheaper and better.

I don't believe that for a second. Free market will always do what it does best, which means that whoever can get away with running their robo taxis cheapest will win. So the most common ones will rarely if ever be cleaned inside(cleaning costs money after all, even if it's automatic). Once the technology is common, they will run the cheapest version of it, just like regular taxis today don't all drive Mercedes S class cars, they drive 15 year old priuses because those are the cheapest cars that do the job. Yes, there will be market for higher end, clean and high tech cars. But the most common service will be nothing like this. A good example is something like Ryanair - yes, the planes are shit, they are dirty, the service is poor, but they are immune to ratings because.....what else are you going to pick? On several European routes that I travel on, they are the only choice.

The cars will certainly be designed for robo-cleaning by the time we have self-driving fleets. In a free market with free choice, people will obviously pick the more hygienic fleet (which I don’t believe will cost much more, controlling for other factors).

The airline industry doesn’t work very well as an analogy. It’s not a completely free market, and costs are so high that it’s genuinely hard for airlines to provide a better experience without charging a lot.

>subsidized enough by investor cash

Investors have no reason to subsidize costlier transportation modalities that customers may not even prefer half the time. As for the verbiage, I don’t think self-driving car subscription services will brand themselves as taxi fleets. They will replace the vehicle that you own and that sits on your driveway 97% of the time (that’s a general “you”). However, Americans like the feeling of ownership over their vehicle. The name of the service should take that into account.

the trains and other forms of public transportation may have to go the way of the other things impacted by covid. The new reality calls for individual, ideally autonomous, vehicles. Especially in the dense cities. Small, electric, may be even 3-wheelers.

Sidenote: i wonder whether the lockdown may work as a selection pressure on the covid to produce the more viral, like long living on surfaces, etc., mutation of the virus.

It's a bit early to say, but in my opinion (as a non-specialist, based on reading them), there will be no "new reality". Covid is a big thing while it spreads, but it won't be a large concern in 2-3 years even if it becomes endemic, especially with better hygiene practices and treatments.

Not even bicycles can even approach the density of a train/metro/bus, which is required to have cities with the density we currently have, and I don't think covid-19 is nearly enough of a threat to make us give up on that.

Reply on news.ycombinator.com