Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Let’s Talk About Self-Driving Cars

medium.com

Post reply on HN
Print view View original

91–100 of 341 posts

Re: Let’s Talk About Self-Driving Cars

#91

Post by dahart »

losteverything wrote: post #23
dahart wrote: post #10

> Level by level or straight to level five? This strikes me as a slightly strange question. Maybe it's just an excuse to present the six level system as a framework for thinking about it, but a huge chunk of commodity new cars on the market this year are level 2. A small chunk of higher end cars are level 3. Teslas are arguably level 4, and we all know there are level 4 & 5 cars on the road around the bay and few oth…

Do you think It's too late to bring them "back to port?" I see one endgame where state referrendums put ballot questions to citizens that ban driverless vehicles or require drivers on non-Federally funded roadways. Yeah, sensors will be added to cars (like seat belts were) and cars will be aware of other vehicles.

> Do you think It's too late to bring them "back to port?"

Probably, yes.

> Yeah, sensors will be added to cars (like seat belts were) and cars will be aware of other vehicles.

High end cars have had auto steering and auto throttle for at least a couple of years. One of my friends took me in a ride two years ago in a Mercedes that keeps a safe distance between the nearest car, and steers itself on curvy roads. You can set cruise control and go from Sacramento to Reno on I-80 with little intervention.

I drove most of the new compact SUVs on the market over the weekend, and bought a 2017 Honda CRV. All of them have sensors to warn about nearby traffic. Half of them have some form of automated steering, usually called "Lane departure mitigation". The CRV will try to steer itself to stay in the lane. (It's not great, I wouldn't trust it for very long, but it is there.) It has "Active Cruise Control" where it follows the car in front with a preset time gap you can set. I let it drive me through the city yesterday, it comes to stop at a light as long as there's a car in front of me. (Which, incidentally makes it seem dangerous to use this way since it doesn't read stoplights.) The car has anti-collision automation, it will slam on the brakes and possibly steer if a crash is imminent, and it has "Vehicle stability assist" mean anti-roll and anti-slide measures that will take over brakes or steering or both.

The CRV is a mass market car, not something high end or experimental, so yeah I think a high and growing level of automation is on the road already and will not go away.

Fully driverless, I'd agree, that seems unlikely right now, and I wouldn't be surprised either if lawmakers or the public decided to make that illegal for now.

Re: Let’s Talk About Self-Driving Cars

#92

Post by ForrestN »

CPLX wrote: post #19

I find it odd the way people talk about moving through the levels sequentially without really acknowledging there's a massive phase change that happens somewhere between level two and four, where a self-driving car could easily be way more dangerous than the average human driver. That comes as soon as you have a situation where the automation has to be able to handle all foreseeable scenarios without failing because…

There are cultural and legal dimensions here, though. Will some countries prefer partial autonomy/incrementalism in legalizing autonomous cars? Will some people feel more comfortable getting used to autonomous cars while they are still able to drive them when they feel like it?

Re: Let’s Talk About Self-Driving Cars

#93

Post by rsync »

maxerickson wrote: post #52
rsync wrote: post #49

Earlier quoted context omitted.

No, I would not. We, as a society, are about to learn some hard lessons about software. As in, deuteronomy-level lessons - things that will be (figuratively) carved into stone. We will gain a new appreciation for the value of human decision making.

We will gain a new appreciation for the value of human decision making. A pretty big chunk of automobile collisions are directly attributable to poor human decision making. For example, nearly 1/3 of fatal collisions in the US involve a driver with a BAC of 0.08 or higher.

"For example, nearly 1/3 of fatal collisions in the US involve a driver with a BAC of 0.08 or higher."

This is a flawed comparison because failing to update your software or keep your tires inflated or clean off your sensors would also be human decisions that could lead to adverse outcomes with self-driving cars.

Non-driving-operation decisions like that are applicable to both regular and self-driving cars.

Re: Let’s Talk About Self-Driving Cars

#94

Post by maxerickson »

Retric wrote: post #67
maxerickson wrote: post #59

Earlier quoted context omitted.

So only 1/4 of traffic fatalities are attributable to a single category of poor decisions. Oh boy. Do you have a good source that tracks fault in alcohol involved collisions? (I'm not trying to argue, I'm curious to look at it)

I don't want to defend driving while impaired in any way. Sadly, it's still really common to drive while impaired. 121 million self-reported episodes of alcohol-impaired driving among U.S. adults each year Just remember, even if you are sober, driving home from a party at 3AM is still significantly more dangerous than average.

So 1/4 is just a seat of the pants estimate?

I sloppily implied causation above, but I didn't actually say 1/3 of crashes are attributable to alcohol, I said that a "big chunk" are attributable to poor decisions and separately pointed out the large percentage of fatal crashes involving alcohol.

Re: Let’s Talk About Self-Driving Cars

#95

Post by Fricken »

sgustard wrote: post #89
Fricken wrote: post #83

Earlier quoted context omitted.

I would much rather you didn't place so emphasis on the control premium. The cumulative effect of everybody pursuing their self interest with no regard for the sanctity of our shared spaces has been resoundingly bad. It's the tragedy of the commons in a nutshell. With personal automobiles we've already completely jumped the shark. People travel from all over the world to visit cities like Paris, but I can't think of…

Hm, to me the "on demand" economy, which Uber and the like are part of, is the ultimate pursuit of self-interest. The entire model is built on getting exactly what you want when you want it, with no downsides. Once there are robot cars there is no longer a limit to how many cars can be on the road. Imagine one car zipping around for every Amazon delivery, 24 by 7.

I don't necessarily disagree with you. While autonomous vehicles can be tremendously valuable tool, if we let narrow self interest and unfettered capitalism determine how they're implemented, they'll cause as many problems as they solve.

Re: Let’s Talk About Self-Driving Cars

#96

Post by saosebastiao »

In an article that has obviously been well thought out and reasoned, it's disappointing to see the author using the current accident rates as reported by self driving car developers as if they were a representative sample and subsequently used as a lemma for further arguments.

Simply put, the current rates reported aren't a representative sample. In fact they aren't even naively semi-representative. They are deliberately skewed by development patterns, and that happens to give them an optimistic accident rate.

SDCs in testing do not drive the same miles that humans drive. They don't drive in the same conditions, in the same traffic, and not on the same roads...and they have engineers ready to take over at any time, which is a luxury that the baseline does not have. Engineers carefully select the scenarios where they do drive by taking the scenarios that they know the cars can handle well and then slowly pushing the upper bound of that range. In other words, they choose the easy and work up to the hard. It is yet to be demonstrated how well they will perform once they are closer to driving the miles that humans drive.

Re: Let’s Talk About Self-Driving Cars

#97

Post by notheguyouthink »

aetherson wrote: post #41
bluGill wrote: post #37

Earlier quoted context omitted.

Just remember that self driving cars do not have to be perfect, they just need to be better than humans. The use is at just over 1 death per million miles. If self driving cars can reach .5 they are better already. If the cars can detect the times they are bad and pull over (screw your appointment) that might be enough.

Would you get into a self-driving car that you knew had fewer fatalities per mile than the overall human average, but more fatalities per mile than the average of non-drunk, non-sleepy human drivers below the age of 70 and above the age of 20 with more than 5 years of driving experience? Serious question.

I think in that example, i would move the bar to exactly what you cited. I think there's an implicit understanding in the OPs statement that means "better than average competent humans".

Although, another way to look at is: If you fall below the bar for the average human, including drunks/etc, aren't you by definition more safe using this AI system?

If your question/argument is: What about those who are safer than average - then i think the answer is, they shouldn't use self driving, only the "poor" drivers should. Of course the next question is how do you determine that.

Perhaps infractions on your driving record would inhibit you from driving. If you get caught for drunk driving, that's it, no more license, purely automated driving (even if that means no rain driving, etc). If you're above 70, or etc.

Re: Let’s Talk About Self-Driving Cars

#98

Post by maxerickson »

rsync wrote: post #93
maxerickson wrote: post #52

Earlier quoted context omitted.

We will gain a new appreciation for the value of human decision making. A pretty big chunk of automobile collisions are directly attributable to poor human decision making. For example, nearly 1/3 of fatal collisions in the US involve a driver with a BAC of 0.08 or higher.

"For example, nearly 1/3 of fatal collisions in the US involve a driver with a BAC of 0.08 or higher." This is a flawed comparison because failing to update your software or keep your tires inflated or clean off your sensors would also be human decisions that could lead to adverse outcomes with self-driving cars. Non-driving-operation decisions like that are applicable to both regular and self-driving cars.

Driving while impaired is not a non-driving-operation decision. It's an ongoing driving decision. The driver is poorly assessing their ability to safely complete upcoming operations. Over and over.

I also expect that automatic systems will greatly improve the level of maintenance of vehicles (for example, mandatory tire pressure monitoring already does...).

Re: Let’s Talk About Self-Driving Cars

#99

Post by MR4D »

CPLX wrote: post #19

I find it odd the way people talk about moving through the levels sequentially without really acknowledging there's a massive phase change that happens somewhere between level two and four, where a self-driving car could easily be way more dangerous than the average human driver. That comes as soon as you have a situation where the automation has to be able to handle all foreseeable scenarios without failing because…

Imagine some drunk guy hops into his car in bad weather. He taps the auto-cruise button and then passes out. Snow, rain, construction, - something - hinders the ride somewhere along the way home. Car cannot process the issue correctly, and slows to a stop, alerting the driver. Driver, being sound asleep, cannot hear the annoying alert tone. Alert continues, and possibly after some time, alerts authorities of it's disabled status.

I'll bet that in the near future (next decade or so), there will be rising incidents of traffic being blocked due to a stopped automated car who's driver is not responsive, but the computer couldn't handle some construction or weather, and went into fail-safe mode, simply slowing down and stopping (but not pulling over due to lack of space).

Heck, we're halfway there with people checking FB on their phones at stoplights, so anticipating this seems reasonable. We haven't solved that problem, and I don't think we'll solve the one you raise (not for a much longer time, at least). Valid problem, but a solution is likely far off.

Re: Let’s Talk About Self-Driving Cars

#100

Post by nradov »

bluGill wrote: post #37
CPLX wrote: post #19

I find it odd the way people talk about moving through the levels sequentially without really acknowledging there's a massive phase change that happens somewhere between level two and four, where a self-driving car could easily be way more dangerous than the average human driver. That comes as soon as you have a situation where the automation has to be able to handle all foreseeable scenarios without failing because…

Just remember that self driving cars do not have to be perfect, they just need to be better than humans. The use is at just over 1 death per million miles. If self driving cars can reach .5 they are better already. If the cars can detect the times they are bad and pull over (screw your appointment) that might be enough.

In terms of improving safety, I think most people are looking at it backwards. Instead of having the computer drive and expecting the human to take control in dangerous situations we should have the human drive and allow the computer to take control temporarily when it detects a high probability of an imminent collision. This could prevent most collisions and is technically easier to accomplish.

In fact some vehicles already have front crash prevention systems and stability control systems that automatically brake and cut the throttle (2 control axes).

http://www.iihs.org/iihs/topics/t/automation-and-crash-avoid...

So manufacturers could build upon that baseline by adding more sensor inputs and control axes for collision avoidance. Instead of just braking when a large obstacle appears in front the vehicle could steer around obstacles, exert force on the steering wheel to prevent running off the road, limit throttle inputs when approaching curves, etc. Some aircraft autopilots already have similar functionality by automatically climbing to avoid terrain.

Post reply on HN