Hacker News new | comments | show | ask | jobs | submit login
Self-driving cars could save more than 21,700 lives, $450B a year (computerworld.com)
26 points by dsleno 1490 days ago | hide | past | web | favorite | 42 comments

  Of course, many such benefits may not be realized until high 
  [numbers of self-driving vehicles] are present
Insurance will drive the adoption of self-driving vehicles. If you can buy a vehicle that statistically won't get into an accident, it will be cheaper to insure.

Even better, Google and the automaker's could get into the market of managing fleets and their own insurance directly or via wholly owned subsidiaries. That's what I would do if I were in their position. Basically, this is going to completely disrupt the auto insurance industry as well and no one has more and better data to price the insurance than Google that now has hundreds of thousands of miles logged and will always be in possession of the sum total of all data on all miles ever driven. They are no longer performing statistics on a sample size, but are instead doing so based on an entire population. Basically Google can do the same for auto insurance that Amazon has done for online retail. i.e. they can operate on razor thin margins that no one else will ever be able to match because they'll be operating on sample-based statistics and it will take them a while before they have a significantly large sample to properly price their product.

FWIW I posted the above comment in the other story submitted on this topic today on insurance. [0]


It will have some effect, but insurance is a fairly small piece of the total cost of ownership of a car. It will be interesting to see if the cost of all the extra mechanisms and electronics to make a self-driving car are offset by lower premiums over the expected period of ownership.

Small? I pay over $1000 a year for insurance up here in Canada, and keep my vehicles for more than 10 years. That's a good chunk of the cost of a vehicle.

Interesting, I picked what I thought would be a "typical" car (Ford Focus sedan) and it shows the insurance getting more expensive as the car gets older. That's exactly the opposite of my experience with any car I've ever owned.

I'm also reminded why I drive old cars that are cheap to insure. My insurance costs are a fraction of what that page shows.

This assumes the insurance cost of a non-self-driving car doesn't go up as the pool of people using them shrinks and the liability shifts to them. Insurance companies like people who probably won't have to take their money more than people who will eventually have to.

Is it?

Let's say I buy a new 2014 Corolla for $16,800 and keep it for 20 years.

My insurance (in BC) would be about $100/month or $24,000

Even according to the Edmunds' calculator, insurance is the largest single cost.

We may also see initial adoption by people who buy 0 cars. With self-driving cars it becomes much more convenient to use a car rental service that has a car pick you up when you need to leave.

1. Car design, interiors: get rid of pedals and steering wheel, have a touch screen (possibly your phone) for communication. Cross an elevator with a tiny lounge, have a single central sliding door or whatever.

2. Insurance companies: basically, they operate on a margin over the volume of business. Which will go drastically down.

3. Switch over: I'll bet it will be more sudden than anticipated. Think "cash for clunkers", "buckle up your safety belt" and "dont drink and drive" programs all rolled up in one and magnified tenfold. Anything to get the murderous unsafe machines off the roads.

4. The smug, withering, contempt of news writers in twenty years will be something to behold: how could those primitives allow themselves that awful road carnage for so long?

I am more worried about pollution.

Hopefully the self driving cars will be re-used like driverless Uber cabs. That would reduce the unsightly clutter on so many American streets, colloquially known as "parked cars" and alleviate congestion on city streets, much of which comes from cars circling around the block "looking for a spot to park".

Sure, it would cut into the revenues of parking lots and car manufacturers. But IMHO no one has done more to pollute the environment than the latter. The number of people on the planet has increased and - even as Europe cuta down on its car population thanks to the high price of petrol - we US folk continue to purchase our gas guzzlers and have a car culture in many cities. Some like LA are perpetually covered in smog, so we can see the effects of this car culture.

Thankfully, after decades of polluting the environment for our individual joyrides, things are changing - the younger generation isn't into owning cars as much, and self drivi g cars are on the way. An unfortunate victim of this trend will be the venerable macho tradition of the man picking the lady up in his OWN shiny ride, which he drives. Now it will be about renting the car for the occasion, and both occupants can interact with each other while the car is in motion.

In short:

  Less pollution
  Less congestion
  Less wasting time looking for parking
  More equality between sexes in terms of driving
  Less unsightly clutter on streets
  Less revenues from car manufacturers from producing CARS,
  so they can focus on recurring revenues (eg renting stuff
  inside their fleet)
Hopefully this will lead to a more sustainable form of transportation.

I think "not owning" a car is going to be a big thing. Car sharing even more so.

I used to be one of those people that would never ever give up his car. Giving up my freedom and pinning me to this city? Never.

After having my car broken into and vandalized a couple of times added to the insane cost of parking where I live, I did eventually give up my car 1 year ago.

I now use a car "sharing" service where a company owns 100's of car throughout the country, all bound to their own (extremely convenient) parking spot. I reserve the car 1 minute in advance, open it up with my phone and drive.

The best thing is: It's not my car.

I don't care what happens to it. I just leave it at its little spot after I'm done and I forget about it. No maintenance, hardly ever have to fill it up with gas, billing takes place at the end of the month.

Now imagine this car not having that parking spot but being automated and just picking me up from my doorstep. Not having to put it back into it's designated spot. I would feel pretty awesome about that.

My friend recently sold his car in central London and is moving to Zip car and the alternatives when he and his wife need one. The running costs, maintenance, insurance and parking involved with it didn't add up. We are seeing drastic changes in the transport industry.

The article interchangeably uses "self-driving car" and "autonomous car", but it doesn't clarify whether the benefit calculations include completely unmanned usage.

A car that drives by itself with passengers in it, is one thing. But an unmanned vehicle takes the concept to a whole new level.

It will have an incredibly deflationary effect on the economy. Trucking is a $600 billion business. It employs 3 million people. Over a million of them are long-haul truckers. And that's not counting non-driving support staff. With self-driving trucks, 30% reduction in headcount isn't unrealistic.

That's just one example.

It's a scary prospect for many people.

This is yet another example of how we are /very/ rapidly plunging (like it or not) into a future where our amazing advances in technology make most of what we consider "work" obsolete and unnecessary. In a sense, we (as a species) kind of won that "life" game.

Right now a large percentage of society is doing "work" that isn't particularly complicated or difficult. I's necessary, of course, but it's the kind of thing that we are - with increasing frequency - starting to think we really should take the very small shell script[1] option.

Many obvious references could be made to Marx at this point, of course. If you want to. I prefer, on the other hand, the idea that our economy is simply taking on some Star Trek-style, post-scarcity traits.

Technology can never replace /all/ jobs, of course. More precisely, Gödel showed us there's always more complicated, more interesting tasks out there. The tasked involved in maintaining the baseline needs a person has in life (food; shelter, basic health-care) and even a few of the "luxuries" (transportation, communication, the general-purpose computer), though, are going to simply be solved problems.

Autonomous cards are indeed going to be a very scary prospect, and it isn't (primarily) because of any safety concern. The changes to society this technology will cause (such as those 3M truck drivers) is pure, distilled terror for anybody that wants to hold on to current social model.

Worse yet - the self-driving car is just the first of many technologies that are happening right now. Millions of unemployed drivers is nothing compared to /billions/ of bureaucrats, low-level factory workers, fast-food/walmart-type workers, and middlemen of all sorts and the like being made obsolete.

Hmm... what was it that Marx said about revolution?

We better stat getting ready for these changes in the meantime. "The avalanche has already started, it is too late for the pebbles to vote."[2]

[1] http://www.nerdyshirts.com/shell-script-funny-t-shirt No store endorsement (good or bad) intended; the link was an arbitrary google pick for that t-shirt.

[2] http://www.youtube.com/watch?v=60loeoblu0M

edit: link formating

It would be a massive boon to companies like Amazon, as the cost of anything delivered by road would decrease dramatically.

Actually, everything would drop in price, including your basic weekly groceries.

The questions the article asks are extremely poor.

> Could automakers prevent hackers from getting into onboard computers?

"Hackers." Modifying their own cars. Terrible.

> And who would be liable in the event of an accident in a self-driving car?

The insurance company.

Insurance companies are not liable, the owner of the car would be liable, just as he is if he lets someone borrow his car and that person crashes it. He carries insurance against that risk, but the liability is his.

What makes you think the owner of the vehicle is liable for a crash caused by some third party driver?

In any event, the point is that the question of liability is both irrelevant (because insurance company will be the one paying for the loss in the large majority of cases in any event) and obvious (because these are for the most part well-established questions). It's not like a self-driving car is the first instance in the history of time that some piece of property could cause damage in the absence of human action.

What makes you think the owner of the vehicle is liable for a crash caused by some third party driver?

Because they are? While there may be some variation by state, if we look at California law[1]: Every owner of a motor vehicle is liable and responsible for death or injury to person or property resulting from a negligent or wrongful act or omission in the operation of the motor vehicle, in the business of the owner or otherwise, by any person using or operating the same with the permission, express or implied, of the owner.

So if I lend you my car, and you crash it causing damage, then I am liable. Likewise, it seems to me, if I allow my self-driving car to drive home and it crashes into another car causing damage or injury, then I am liable. In general, the owners of things are liable for damages caused by those things.

The question of liability is not irrelevant, because insurance does not transfer liability it transfers a quantified, limited amount of risk. If I cause damage in excess of my insurance coverage, I am still liable for the damage and I'll have to make up the difference.

1. http://www.dmv.ca.gov/pubs/vctop/d09/vc17150.htm

The California law strikes me as unreasonable. It's an example of the law being expedient rather than just: The person who should be responsible is the person who perpetrated the "negligent or wrongful act or omission in the operation of the motor vehicle." But if that person has had to borrow a car then perhaps they don't have any money, so plaintiffs want to take it out of the pocket of someone who we know has at least enough money to buy a car notwithstanding their apparent lack of any wrongdoing. The just outcome when the owner lends a vehicle to someone who is not known to be irresponsible is for the driver to be liable and, if the driver may not have enough money to pay the damages and we want the injured party to be made whole, the government should establish a requirement to carry sufficient insurance.

> In general, the owners of things are liable for damages caused by those things.

In general the owners have to have been at least negligent to be liable. And the cases to the contrary are similarly contemptible instances of injustice where the lawmakers see sympathetic plaintiffs and nearby deep pockets from which to filch coins and for political reasons (in the case of legislatures) or because they are restricted to deciding a single case (in the case of judges) are more inclined to impose liability in innocent bystanders than set out a requirement for parties to insure against losses incurred through the acts of culpable but insolvent third parties.

> The question of liability is not irrelevant, because insurance does not transfer liability it transfers a quantified, limited amount of risk. If I cause damage in excess of my insurance coverage, I am still liable for the damage and I'll have to make up the difference.

But that isn't the common outcome. The common outcome for a self driving car is for there not to be a collision. In the rare case of a collision, the common outcome is for there to be insurance which covers the damages. In the rare case the insurance is insufficient, the common outcome is for the vehicle owner to be insolvent (because owners with wealth to protect will tend to carry larger amounts of insurance).

If you want to be pedantic, the question of who has liability is complicated. But in the very large majority of cases the relevant question of who pays for the damages is simple: Cars are required by law to be insured and the insurance company pays. And if that isn't the common case then it's a strong argument for increasing the legally mandated minimum amount of insurance.

The state mandated minimums are laughably low: http://personalinsure.about.com/cs/vehicleratings/a/blautomi...

They are scaled to the assets most people need to protect, not the nominal liability they might incur.

Car makers fear and will do whatever they can to stop autonomous cars as they represent an existential threat to their business. We will need an order of magnitude fewer cars in a world of cars-aas, on demand cars... it blows away the entire industry. .. People are, I think, underestimating just how sweeping the shift would be once it actually gets fully underway. It is not an incremental change... It's like the shift from horse and Carriage to autos... That big or bigger

So far they don't seem to be doing much to stop it, but in the end they won't have a choice either way.

Anti-car lobbyists in the dawn of the car era got ridiculous laws about having someone walk in front of the car with a flag put on the books in many places and it still didn't stop car adoption.

So far there's nothing to stop!

At the same time, individual car makers will want to be on the front line of adopting autonomous cars, as it will give them an advantage over their competitors.

At some point in the future, someone who decides to drive manually and gets in an accident may be liable for much more than they are now.

How so? Why would they ever be liable for more than actual damages?

Because they had a presumably safer alternative available. It might be construed as negligence. I'm not making a judgement one way or the other about it. I just think it's likely to happen.

I see a couple of extremely powerful advantages to having a self driving car.

Your family only needs one car. You go to work at 8, and your kids go to school on the way so you drop them off. You get to work at 8:15, the car drops you off and heads back home. Your wife heads to work when the car gets back at 8:35, she gets there at 8:45. Your kids get off school at 3, but they have an after school activity so they let the car know it can charge up until 3:30, when it comes to pick them up. It drives there, picks the kids up at 3:45, gets back at 4, drives back to pick you up, brings you back home, drives to pick your SO up at 5, and by the time your SO gets home you have dinner ready.

If you're a single person you could probably split the car 5 ways with your friends. You could probably make a lot of money if you make an app that says "Pick me up here, my current GPS location, and drive me to work 15 minutes away at 8:00 AM every weekday. Bring me from work at 4:30 to a destination of my choosing after work". You could extremely accurately predict the price of transit, you know the probability of accidents. It's cheaper than any taxi, it's convenient.

Few accidents, no traffic jams, courteous to drivers, no speeding or parking tickets (unless you 'opt in' to speed), leisure, take a nap on the way to and from where you're going. The cars return to 'base' which is a parking lot that has a roof outfitted with solar energy, allowing the cars to store energy during the day. When cars weren't being used, if they still have juice, they either share the energy with the other cars that returned, or they compensate for the lack of renewable production at night. Some cars are 'regimented', where they follow the same pattern every day of the week. How much energy they use is known, they know how much energy they need to start the day with, and how much they can contribute back to the grid. There's other cars that are free runners that are always kept charged in case they need to drive to somewhere 4 hours away.

It'd be a little bit of an inconvenience to move things, but If you wanted to go from city A to city B, but needed to recharge halfway, you could just jump out and jump into another car at the charging station. NFC to verify who you are, grab your things and go. When you're driving towards the recharging station, it knows what music you have playing, what temperature everything is at, it preheats or precools the car for you - to ensure that swapping cars doesn't feel like swapping cars. Maybe you go to the bathroom at the rest stop, and when you come out the car is ready to go with all your belongings swapped. Put a little container in the middle that you can stash stuff in, just grab that when you change cars.

And carpooling - if you have 10 people that go to to work 15 minutes away, get some sort of carpool system going. First person gets on, the car drives to the next house, waits 2 minutes for the person to come, if they do great, if not they take off towards the next house. You say what time you want to be picked up, and a time within 10,20,30 minutes that you'd like to get there. Cheaper service if you have a bigger window of availability. Car texts you when it's outside (you can follow it on the map once it engages your current location).

I like thinking about electric cars.

I think it's even better than this. A family doesn't need one car, it needs zero cars, it just needs to be able to use one or more cars on demand.

The average family uses their car probably less than 10% of the time. The rest of that time could be used by others.

Once self-driving cars are ubiquitous, I expect that an Uber-like service will be the norm. You either subscribe or rent per trip and schedule when and where you want to be picked up.

The service you subscribe to has a fleet of cars and knows when people want to be picked up and where they want to be dropped off, as well as whether you're willing to carpool on the way (for a discount). They can optimize the scheduling of their fleet in the same way that FedEx and UPS optimize their driving routes now.

The future is going to feel a little strange and I can't wait.

Yeah, the optimal solution for society involves not owning cars at all. An individually owned car moving around between a small web of trust such as a family, is more likely to make empty trips than a car that drops people off and can pick up the closest person that needs its services.

I hope we one day see a future where you will only be allowed to own a vehicle for inter-city and rural travel. Suburbs and cities should all be serviced by one collectively fleet for all citizens of that city.

I never really got the huge advantage of self-driving cars until I read something similar a few years ago talking about one parent in a family taking the car to work, then the car driving itself home, so that the other parent could drive the kids to school, etc.

Before then I had never thought about the biggest game changer of the self-driving car: The car driving around with NO ONE in it!!

That is huge and once the technology is good enough to be safer than our current state of affairs with human drivers, it is a complete and utter game changer. You would no longer need drivers for 18-wheelers and would not have issues with tired drivers pushing their limits to drive 12, 14, 16 hours...instead a truck could drive all day and all night at 55mph or whatever speeds were safe, only stopping for refueling or maintenance issues.

You could also have driverless taxis, school buses (maybe repurpose the drivers into chaperones who actually pay attention to what is going on in the bus?!), Greyhound-type bus lines, etc, etc.

Plus, as you pointed out, lots and lots of families would only need one car. Or several people could share a car together. Or services such as Zipcar could really take off because you don't have to go to where the car is — you could just find the nearest one and request that it come to you!

On top of all that, a huge benefit is that you can do other things while getting to your destination. Let the car drive while you read a book, catch up on email, text your friends, watch a movie — whatever.

I don't think the biggest impediments will be technological—they will be governmental and legal. Who is at fault in an accident of two driver-less cars? Do you still need a "license" to operate? What happens if they get hacked? Etc.

Totally agree with you that it is interesting and amazing to think about!

All of this, and more, is probably why Google Ventures invested heavily in Uber in their latest round. (Note that they got a lower valuation than TPG was paying in the same round, so they must be bringing some tech to to table...)

Self-driving cars could save more than 21,700 lives, $450B a year

Until they get spoofed. That will be fun.

link to paper: https://www.enotrans.org/wp-content/uploads/wpsc/downloadabl...

If by spoofing you mean holding up a stop sign, I don't see this as a problem. Google's cars use Bayesian reasoning. There are many additional pieces of information for the car to draw on that should outweigh the evidence of the stop sign:

* a Google car will have been on that road before (many times) and never seen a stop sign (and Google can probably tap into the government data for new stop sign installments anyway)

* looking at a map the car can establish what the prior probability of a stop signing being there is (probably low if it's not an intersection)

* the car can look at cars in front of it and observe what they are doing

Beyond this a simple look ahead will solve the problem: the car can extrapolate what will happen if it does or does not stop. If stopping suddenly at the stop sign means causing a rear end accident, but not stopping at the sign has no negative effects (because there is no cross traffic), then the car can safely ignore the sign, especially if it has assigned a non-negligible chance to the sign being fake.

I think you're over-nerding it here.

Google's cars use Bayesian reasoning. — How does that change anything? What's an alternative to "bayesian reasoning?"

Google can probably tap into the government data for new stop sign installments anyway — such a thing (an accurate government database of new stop signs) does not exist.

the prior probability of a stop signing being there — scenario: a construction worker with a rotating STOP/SLOW hand sign during road work. Or, manual traffic overrides with police standing in the road with hands up or hands waving to manually control traffic.

the car can look at cars in front — assumes said cars exist and the motion of forwardness is more important than any potential cross traffic being stopped for.

I think he meant that instead of following deterministic rule (ie stop sign == stop), the car can take into account other factors, and make the actions with the best probability to not cause a disaster.

And keep in mind that the world in which self-driving cars are dominant is very and much different than the world we're living right now (as far as transportation are concern). Stop signs are meant to be read for human, car don't actually have to care about it. How about instead of stop sign at crossroad, we have some devices that can monitor the incoming traffic, and relay that information to the car? (I'd say that the car can even communicate with each other by themselves - but that approach has a host of other problems too).

Honestly, I'd say the main problem in a world of self-driving car are other human-driven cars: the self-driving car has no idea what the human would do, and the human has no idea what the automated car would do.

FROM the report:

RISK: Since each AV in the fleet represents an access point into such systems, it may be infeasible to create a system that is completely secure.

MITIGANT: "Fortunately, robust defenses should make attacks even more difficult to stage. The <U.S.> has demonstrated that it is possi ble to maintain and secure large, critical, national infrastructure systems, including power grids and air traffic controlsystems."

This is a combination of an assumption and non-sequiter.

Besides being an apples to oranges comparison.

Basically a "trust me"...

Unless you dress up as a construction worker and have the sign that says "slow" on the other side.

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | DMCA | Apply to YC | Contact