Hacker News Reader

Comment by njoro | Hacker News Reader

While I hope the technology succeeds I don't think it is going to be as big of a change as people think. If you are upper middle class in a large city where labour is cheap, like in some Asian countries, driving is already close to free. And while that is somewhat nice from a quality of life perspective, it also means that rush hour traffic is horrible and sitting in a car isn't much more fun just because it is cheap. Basically a car is still a car.

View on HN

"“Our experiments show that with as few as 5 percent of vehicles being automated and carefully controlled, we can eliminate stop-and-go waves caused by human driving behavior,” Daniel Work, assistant professor at the University of Illinois at Urbana-Champaign and a lead researcher in the study, said in a statement."

https://www.rdmag.com/news/2017/05/study-shows-self-driving-...

Per the article: > “Before we carried out these experiments, I did not know how straightforward it could be to positively affect the flow of traffic,” Jonathan Sprinkle, the Litton Industries John M. Leonis Distinguished Associate Professor in Electrical and Computer Engineering at the University of Arizona, said in a statement. “I assumed we would need sophisticated control techniques, but what we showed was that controllers which are staples of undergraduate control theory will do the trick.”

If it really is that simple shouldn't it be possible to put this into an app that if 1/20 drivers had installed would also solve the problem?

An app can't convey near the granularity of control to the driver compared to what a fully automated solution could do, and you run the risk of making the situation worse due to the driver splitting focus between paying attention to the road and following the app instructions.

To some extent we already have "the app" (google maps) and I wouldn't be surprised if it's running for close to 1/20 of drivers actively in traffic, and a whole lot more passively collecting and reporting location info that is trivial to map to major roads. To that argument Waymo somewhat belongs to Maps (or Maps to Waymo if you'd like).

> An app can't convey near the granularity of control to the driver compared to what a fully automated solution could do

Yes I agree, however it might not require that level of granularity, that's what I was wondering given that the researchers were surprised by it working as easily as it did in their experiment. Since, That's why I am wondering, maybe it's something a human could be taught or directed to do by an app.

Open questions to me: Is central coordination required which observes the current state of the system and doles out specific instructions depending on that? If it's simpler is possible to craft a simple set of heuristics we could be taught to follow that would make things better if enough people knew them and followed them? Could they perhaps be implemented with simple road signaling targeting know problem area where jams frequently occur?

I'm pretty sure central coordination is not required for at least some significant benefit.

If you drive at the average travel speed steadily instead of filling gaps in front of you as soon as they open, that smooths traffic out for the car behind you as well. Traffic waves of stops and starts can still form back a bit behind you, but it's unlikely to happen immediately behind you. If even a small fraction of drivers drove like this, there would be much less opportunity for spontaneous jams to form.

I saw an impressive video demonstrating this a few years ago, but just spent a few minutes failing to find it. It showed a guy driving at the average speed of the car in front, so that the gap opened and closed as traffic in front slowed down, sped up, stopped, etc. There are articles about doing this, and animations, but what made this a compelling video was where it was filmed: a long straight concave-up road full of cars that was coming down a hill, so you could look out the back and see an long line of cars moving smoothly and not jammed, while in front there was a long line of cars stopping and starting chaotically.

I think this simulation[1] explains why even a small change in traffic patterns can have huge consequences to the stop-and-go behavior.

(experiment holding one of the buses for just a couple of seconds...)

[1] http://setosa.io/bus/

Uber went from zero to essentially killing off taxis, despite the fact that "a car is a car".

Primarily because they don't pay as much for their fleet as a taxi service does.

What's coming next is going to be that, except you don't have to pay your drivers anything because you don't have drivers!

(yes, there will be huge technical overhead of running this system and maintaining the cars/hardware, but over time this will become mass produced, and become cheap).

Uber went from zero to killing off a pedestrian, despite having LIDAR tech.

Downvoted by Uber devs no doubt.

"Uber went from zero to essentially killing off taxis, despite the fact that "a car is a car".

Primarily because they don't pay as much for their fleet as a taxi service does."

Uber (and possibly other similar companies) is a huge charity for passengers funded with investment money. It's definitely premature to declare they've demonstrated or accomplished anything before they operate on a sustainable basis. I do think using a credit card an app is far superior to the way taxi businesses are traditionally run, regardless of price, but I also don't see why traditional taxis can't adopt the mechanism eventually. At some point, the bubble will burst and capital will be scarcer for a while, and things will realign.

However, I think you should consider how autonomous vehicles enable better utilization of existing cars on the road. If you own a car, what % of the time is it sitting in your driveway idle?

Rather consider how with autonomous driving, when paired with services like Uber/Lyft enable a car to be active 80-95% of the time - driving folks around. This means that your cost “of the trip” can be significantly less once you factor how his affects vehicular depreciation.

Peak demand at rush hour determines the number of vehicles necessary to achieve a given level of service. Self driving vehicles don't change the underlying mathematics of the job scheduling problem.

This would only be true if everyone commuted at exactly the same time or along the same route.

You basically just described 'traffic.'

The goal I think is a complete ban on human driving on public roads. With that in place, maybe we can make significant improvements in traffic. I agree that the last mile will always be a problem though.

While self-driving cars are cool, I’d imagine self-driving subways would be easier. In New York, the f train in Queens hits a standstill because of problems in the d line. The air train at JFK has no driver. Why can’t we do this for MTA?

While we are at it, transit in the city should be entirely tax funded. N

We also need self-driving buses. The buses in Queens sometimes don’t show up for over half an hour and three show up at the same time. Why?

Agreed. While self driving cars are 'cool', they are still massively impractical. We need a better, rethought transit system in general.

>While self-driving cars are cool, I’d imagine self-driving subways would be easier. In New York, the f train in Queens hits a standstill because of problems in the d line. The air train at JFK has no driver. Why can’t we do this for MTA?

Technical debt. The system is 100 years old and MTA can't contain its contractors. At the current pace work to install communications-based train control (CBTC) on every line (it's on the L) will cost $20 billion and conclude in 175 years.

Oh, and rules require having two drivers on the self-driving train anyway, who occasionally have to drive to keep proficient.

https://www.theatlantic.com/technology/archive/2015/11/why-d...

The 3 buses can go faster through their route if they move together as on average each one has to stop less times. It's possibly a case of a misguided metric (average route time or delay) driving behavior to undesirable local optima.

Even so, on my way to/from work I'd estimate at least 50% of cars are still parked. That's 50% of space that could be gained.

That is what I am considering when I say that it is already close to free. Those city, at least for those people, essentially already have self-driving cars because it is so relatively cheap to be driven.

Optimistically, self-driving vehicles may allow the west to finally discover the joys of share taxis, which would increase people/vehicle and (hopefully) decrease congestion.

Autonomous vehicles will make deliveries become almost free. This in turn will make food delivery explode, to the point of kitchen becoming an option in houses. In general, owning stuff won't be as important as before because you'll be able to simply borrow anything you need.

Reducing driving accident deaths (>37,000/year in the US alone) will be a pretty amazing change.

37,000 lives per year wasted due to fatalities.

400,000 lives per year wasted mind numbingly staring at a road and keeping a car driving along it.

Currently glad I did not go into trauma surgery. Most of their work is car accidents.

Sure, I just don't think that is as much of a paradigm shift as gradual improvement over time. It is not self-driving so much as the deployment of self-driving that will make that change if that makes sense. If everyone drove XC90s on barrier separated roads with automatic traffic cameras that enforced things like speed and distance between cars, we would also be much safer. But it would still be largely the same thing as today.

The idea is that with autonomous driving nobody needs to own a car and traffic jams are severely reduced because robots don’t drive like assholes (or at least by design shouldn’t...)

Not having to own a car to travel by car will of course increase driving, or being driven really. A situation where there is only self-driving cars is probably decades away at he minimum. And that doesn't really solve inherent throughput or speed issues.

Not only "doesn't really solve", but "makes vastly worse".

Not necessarily

>... technology move people from their homes or work to existing public infrastructure hubs

could be a thing. Also maybe some sort of road pricing to stop them being clogged with empty cars going to get pizza.

Not having to own a car to travel by car will of course increase driving, or being driven really.

I'm not sure. When you own a car, you pay large fixed costs but your marginal cost per mile is low, whereas with rented cars it's the reverse.

Sure, people who don't own cars currently will probably ride more, but everyone who does have a car now will have an incentive to ride less.

Your visible marginal cost (i.e. mostly your fuel) may be low but the vast bulk of car costs are proportional to mileage--especially in areas where salt in various forms (winter roads, marine) isn't an issue.

It doesn't really matter if it's lower than the fixed costs, it matters whether it's lower than these hypothetical rented cars, and most likely, it is.

Even if self-driving cars are 50% of the fleet, these cars drive better and can collectively work around traffic jams, so even human-driven cars benefit from automation.

They may also lower the costs (time, effort, money) of driving. That will almost certainly lead to more driving and likely greater urban sprawl.

And mostly because they can drive when you're not in it. I'd be very curious to see city simulations where people just get off 200m from their job and wave their car goodbye as they can park somewhere else smoothly.

In the techies dream world, the car goes off and, gives someone else who starts work a little later than you a ride.

Yeah sure, in the fully shared/non-owned system that would be it. A fleet of non polluting transportation mules.

In America people will get in the car to drive 200m instead of walking.

Because that's always what happens when you make something better: people use less of it. What?

That's clearly not what I said.

Reply on news.ycombinator.com