Just press go: designing a self-driving vehicle | Hacker News Reader

Just press go: designing a self-driving vehicle

googleblog.blogspot.com

540 pointscloudwalking307 comments

The comments here are rather disappointing. It doesn't have manual controls? Well thats the point! I want to leave my crap in it and paint it some ugly color? Get with the times, young people already opt to not own cars. I can't double park in the city? Why are you imposing your steel box on people living in cities, many of which do not own cars and want walkable streets?

I mean, this is the news website for a startup accelerator. If you are on here, have some appreciation for new ways of doing things. To disrupt the status quo is the very goal. If you ask for perfection from day one, we will never get anywhere.

These cars have the potential to massively reduce traffic fatalities (one of the biggest remaining killers) and make cities useful again to the people that actually live in them.

Is a car parked on a street really an imposition on you?

It is (indirectly) to me. I often visit the US, and I'm always amazed that I can't seem to go anywhere walking. The distances to go anywhere are huge. It seems to be that way because cities in the US where designed / build with cars in mind.

In Europe cities were obviously not build with cars in mind, and as a result I enjoy here a much higher density of amenities (and suffer much worse driving conditions / parking time).

Having all this space available for cars is not transparent to pedestrians.

> It is (indirectly) to me. I often visit the US, and I'm always amazed that I can't seem to go anywhere walking. The distances to go anywhere are huge. It seems to be that way because cities in the US where designed / build with cars in mind.

Huh, why would you think that? I'm not sure there's a major city in the U.S. that wasn't already a major city pre-car. And cities weren't redesigned post-car, that's why a lot of them have road problems. U.S. cities are spread out because of suburbanization. It's only going to increase as the workforce becomes more remote-based.

> I'm not sure there's a major city in the U.S. that wasn't already a major city pre-car.

Los Angeles. Possibly more.

> And cities weren't redesigned post-car, that's why a lot of them have road problems.

Sure they were. Not from scratch, but incrementally. Its true that because it was incremental and not ground-up, and mostly affected new development and re-development, that cities that were major cities pre-car (like NYC) look very different than newer cities (Los Angeles).

> U.S. cities are spread out because of suburbanization.

"Suburbanization" is an effect, not a cause. Designing for cars is a (but not the only) cause.

New York City is the most obvious example, but most large east-coast cities were large pre-car. And many cities made major infrastructural changes in light of the car, mostly in the form of the addition of elevated highways, some of which are now being torn out again because of their negative effects on neighborhood continuity, urban fabric, etc. Examples of the former would include the I-395 underpass/mixing bowl into DC and chunk of I-5 that separates downtown Seattle from Capitol Hill, and examples of the latter include the ripping out of the Embarcadero freeway in San Francisco and (likely) the Alaska Way viaduct, also in Seattle.

Note, however, that suburbanization is a post-car phenomenon.

Las Vegas. I know it isn't a typical case but it is definitely pedestrian unfriendly and largely if not entirely developed post-car.

Edit: Even along the strip itself it is hard to take a direct line and you need to go over bridges etc. Walking from the main part of the strip e.g. Bellagio to the convention centre is only a couple of miles but is a pretty unpleasant walk. Whether it is better than waiting for a bus/taxi from CES probably depends on the condition of your feet and the shoes you are wearing.

I do understand that in the Summer it would be a pretty unappealing walking any distance at all.

I think you could make a reasonable case that at least 7 of the 10 largest US cities were not "major cities pre-car". Depending greatly, of course, on exactly what "major" means. Note that none of these cities resemble older, denser cities like NY in their layout.

LA Houston Phoenix San Antonio San Diego Dallas San Jose

Source: https://en.wikipedia.org/wiki/List_of_United_States_cities_b...

Have you been outside the east coast?

I visited orlando last year, and it was impossible to get anywhere on foot, mostly due to distance but often also because pedestrians had not been taken into account in the road planning process. In fact, there was a large shopping mall visible across the road from our hotel, but no allowances had been made for crossing that road in anything but a vehicle, so we had to take a taxi to get there, or 2 buses.

(I say this as someone who drives to work each day and parks on the street.)

Yes, think about the space taken up by parking, and the visual barrier it presents.

There's a shopping and restaurant strip between me and my office. The shop owners campaign for maximum street parking for the convenience of their customers, on top of the side-street and off-street parking behind most of their buildings. Yet at almost all hours, the entire strip is lined on both sides with cars to the point that anyone passing can't easily see shop windows, foot traffic, the social proof of people eating in a popular restaurant, etc. You just drive on through because it lacks appeal.

Think about a shopping centre and how open-air parking means that the complex takes up a huge amount of space and sets it all back so far from the road for anyone walking. If we took out parking allocations (or went with undercroft parking) and collapsed cities or filled in those areas with more "living" spaces, I think the result would be more appealing. Blocks could be more lively or smaller and more walkable.

There's some interesting stuff about parking here: http://www.cato-unbound.org/2011/04/04/donald-shoup/free-par...

Real estate in most cities is extremely expensive. Now consider all the space we give away for free so people can momentarily store their cars.

Some people have realized the extreme bargain this is:

http://www.nytimes.com/2004/05/08/nyregion/a-rent-free-place...

A whole lane for parked cars each side of a street often means no dedicated lanes for public transit like buses or trams. Letting them get stuck in traffic is a huge loss for the many people that commute on them.

Assuming this is an actual rather than rhetorical question, I'd recommend Donald Shoup's The High Cost of Free Parking: http://www.amazon.com/High-Cost-Parking-Updated-Edition/dp/1...

> The comments here are rather disappointing.

Most of them remind me of Clifford Stoll's famous rant:

http://www.newsweek.com/clifford-stoll-why-web-wont-be-nirva...

"This new thing isn't like the old thing, so it could never work."

>The truth in no online database will replace your daily newspaper, no CD-ROM can take the place of a competent teacher and no computer network will change the way government works.

Sure called that wrong.

No kidding... change.org such a letdown.

Stoll was one of the first to hop on the Internet bandwagon, and he thought he could get lucky twice by hopping off the fad before it became passe.

As for the crazy from the other side, read anything by Nicholas Negroponte at the same time.

A lot of young people, especially young hackers, think like authors of science fiction novels, where there is some grand society that they have designed, and how everything is happy. Why don't people get on board?

First, they don't trust you to be right about all the trade-offs. This mistrust only increases when you tell them to get with the times; it's clear distaste for their preferences, so it only signals that you don't value their input.

Second, even if your new society model is better, you have to get people to use it. And if the way you transition is "everyone give up your old thing, then we will all move to the new thing, trust me," they will do the opposite of trust you.

Self-driving cars are a very nice transition, because no one is forcing them into use. You can have one if you want. You can hear your friend talking about how useful they are. You can try it out, and if you don't like it you don't have to use it.

"No one owns a car" is not a nice transition, no matter how awesome you think it is in the latest novel you read about able-bodied and pretty 20-somethings. Public transportation really sucks in a lot of places precisely because making public transportation awesome gets less priority than a bunch of other things, like demanding that people get to bring their pets or bikes on board, or the homeless people using it as a place to sleep.

You can make a very good case that social justice requires letting the homeless sleep on the buses. That's fine. It may even be the right call. But it means that no one else will want to use the buses. (NYC has cops every block to keep the homeless moving along, and their public transport is pretty nice.)

Buses could be awesome, if they were managed like Google buses. I would love to take one of those to work. But look at all the grief they get from the left. Grand ideas for re-imagining society will face opposition from places you would never have imagined.

Not every change is an improvement, but every improvement is a change.

> "No one owns a car" is not a nice transition, no matter how awesome you think it is in the latest novel you read about able-bodied and pretty 20-somethings. Public transportation really sucks in a lot of places precisely because making public transportation awesome gets less priority than a bunch of other things, like demanding that people get to bring their pets or bikes on board, or the homeless people using it as a place to sleep.

...

> Buses could be awesome, if they were managed like Google buses. I would love to take one of those to work. But look at all the grief they get from the left. Grand ideas for re-imagining society will face opposition from places you would never have imagined.

I think this is all a bit over the top. Many cities already exist that function very well substantially without private cars. Central London, although rather a pleasant place, is not a science fiction utopia exclusively filled with nubile twenty somethings, but nevertheless it functions overwhelmingly on other means of transport - buses, cycling, the underground, and taxis. I'm not aware of it being particularly unique, either. What you're suggesting as difficult or intractable problems are frankly rather minor. Maybe I'm just lucky, but I've only ever come across the occasional over-chatty drunk.

Even taking into account the whole of London, including the much more suburban outer sections, where you really do need a car to get around, 40% of households do not own cars [1]

The real issue with public transport isn't the homeless, it's that it tends towards a hub and spokes model which works very well for commuting, but quickly becomes awkward for non-standard journeys. Driverless cars would add an excellent complementary element to the combined transport system. I expect that in a couple of decades the number of households that don't own cars will have gone up to 60%, meaning pretty much universal in the centre, and I don't anticipate that will create a dystopia.

[1] http://www.uncsbrp.org/driving.htm

As someone who has ridden those buses, I spent a long time trying to figure out why the left was giving them so much grief as you astutely point out. I can't see it. I put myself in their shoes, read the rants, and the platform just doesn't make any sense to me.

Which is too bad, really, I have a feeling there is an important issue there that I should see.

I think the idea is that San Fran should be for people who work in San Fran. If you work in Mountain View, go live in Mountain View. Rent is high enough in San Fran without people who commute out of the city for work.

Yes, I realize that, and live in Mountain View. I've also lived on the east coast, where it's common to live in north Jersey and commute to NYC. People live where they want, and attacking a supply and demand issue by starting with the demand seems like the wrong approach.

I'm not prepared for this debate, just pointing out why I can't understand the position.

It's not my debate. I was just giving my opinion on what the argument was.

it's mostly envy.

> young people already opt to not own cars.

Really? Any study on this matter?

Yes. This has been discussed on the Internet quite a bit. Not sure if exact reason has been determined. I'm not young but I sold my car years ago because it's expensive, inconvenient, and seldom used where I now live on the US east coast.

Here's one article, you can Google hundreds more.

http://www.digitaltrends.com/cars/5-reasons-young-people-are...

I would love to see an actual study on this. There are only a few cities in the U.S. where public transportation is good enough to get by without one and while those are typically big cities they aren't big enough to compensate for the number of cities with poor public transportation.

I came from a car-dependent city to NYC, which is one of those cities where being car-free is less of a hassle than owning a car. However, I still want a car. No, I do not want to have one car for each person in my family. No, I do not want a massive SUV. No, I don't want to commute by car (hate traffic). I want a simple hatchback or wagon that I can use on weekends to haul stuff, I want to be able to drive to the beach or the mountains whenever I want. Renting / sharing isn't the answer. ZipCar sucks, to be frank, and is too expensive. Renting from the majors in NYC is a complete nightmare and is prohibitively expensive. No, I'm not going to use Uber or Lyft to get my groceries or go to Home Depot or to take a road trip.

This is why cars will never disappear. The solution is self-driving cars. Cars are a menace when operated by humans. Reducing the danger, traffic problems, and parking issue with cars is the answer, not removing them entirely (although I do support banning them in dense urban areas and business districts -- people can always park on the outskirts and take transit in).

If it's not too long you could try a (electric) cargobike to get groceries etc. :)

But yes, cars are necessary sometimes. We need a solution for people who only need a car a couple of times a month or less. A fleet of self driving cars you can rent perhaps? Like a taxi service without the driver.

and seldom used where I now live on the US east coast.

Seldom used? Even New York City has lots of cars.

I think people are assuming everyone else is just like them.

As for universal car-sharing, it starts with the idea that "cars are idle 90% of the time" and ignores the fact that 80% of people want to use them at the same time: on their morning and evening commutes.

I live in New Jersey. Sure, lots of people spend $400/month to park so they can drive on the weekends. Anyway, so we can stop talking in generalities, here are some numbers.

http://en.wikipedia.org/wiki/List_of_U.S._cities_with_most_h...

It would appear that many people do live without cars so there is an untapped market.

When I googled, this PDF was the first result: http://www.uspirg.org/sites/pirg/files/reports/Transportatio...

I believe this is skewed to young people in urban areas, but I think this has always been the case. I lived in Chicago when I was a "young person" 25 years ago and I did own a car because I moved there with it. I soon wanted to get rid of it and had I stayed there longer I would have. In that setting it just wasn't worth the hassle and expense.

Are you saying that if my opinions are unlike yours, then I should not be visiting HN, or even voice my opinions?

Not everyone on HN is from the US - US has a massive car problem, because Americans had a very strong car culture for 7 decades now - so for them, not having a car is "new".

I come from a country where 25 years ago pretty much no one had a car, because you had to have a permission from the government to own one, and even if you did, there was a 5 year wait to be allowed to purchase it. For us, having a car and embracing the car culture is "new". Not having cars is a sign of the older, poorer country, which we want to have little to do with now.

For me personally, driving is like the second favourite activity I do during the day, even if it's just going to work. I dread the day when self-driving cars become mandatory - I will be incredibly unhappy and I do not want that to happen anytime soon.

Is it ok to voice my opinion here? Or will you tell me that I should go away from the "startup accelerator; new way of doing things" website that I have grown to love? Can you not like both startups AND cars? Is that too complicated?

This isn't even about cars (self driving or not) per se. It's about someone doing something different and all the comments amounting to this isn't how I'm used to things working, and I don't like change anyway. As Elon Musk puts it, you can't necessarily ask people what they want because the answers will be incremental updates to the status quo.

Since you insisted on cars: the short answer is that theres lots of evidence humans can't drive cars. It goes well 99.9% of the time and the rest ends up being 30000 people dead every year, many of them not in a car at their death. So they died for somebody elses convenience, and that turns out to be a rather common occurence with cars and all the things they bring along (just look for the free parking debate below). So the sentiment in countries that have had cars for a long a time and have seen the effects of that is shifting to no longer put up with it, certainly in cities where suburbia descends every morning.

> 30000 people dead every year,

It's worse than that: the annual global estimates are 1.2 million dead, 50 million injured. [1]

[1] http://en.wikipedia.org/wiki/Epidemiology_of_motor_vehicle_c...

You are blowing the post you are responding to far out of proportion. (Who said anything about self-driving cars becoming mandatory?) Your whole comment seems to be spoiling for a fight where none was sought, which is much more disagreeable to me as an HN reader than any of the other behaviors you mention.

You can bet that as soon as they solve all the technical problems, Google will be lobbying to make them mandatory.

Even if they don't lobby for them, insurance will very quickly make them the only option. Eventually, insurance cost for manual cars will be many times greater than the one for self-driving cars, making all manual cars prohibitively expensive to drive - and then after another decade or two of that, the percentage of manual cars left on the market will be so tiny,that they will be banned altogether, restricted to race tracks and private property.

What is your reasoning for why insuring manual cars would be any more expensive than it is now? If anything, it would be less expensive since those cars are sharing the road with safer self-driving cars.

If manual cars were comparatively more expensive then psychology might come into play and people would be reluctant to go for the more-expensive manual option.

"I dread the day when self-driving cars become mandatory"

The upshot of this is that this probably won't happen until after you become unable to drive yourself. I'm sure there will always be a market for enclosed driving courses/racetracks anyway.

I propose that the polarization of the opinions on this thread are due to the relationship between the relevance of the announcement and the time it'll take to become a mainstream polarizing issue. I tend to lurk on HN, mostly because I'm not on obsessively enough to have any relevant comments before a thread is mature. I think many lurkers share my interest in the subject matter, but are storing it away as a data point to be considered in the future rather than a policy discussion generator now.

I've broken out of my normal mode in case you actually believe that HN's general opinion on this technology is a consensus that entirely disagrees with you. Personally, I'm excited for the technology to become mature enough to become a consumer product, but I'm unlikely to remand control of my personal vehicle to automation if I don't plan on drinking.

I know that retaining my vain control obsession over my vehicle will eventually make me one of the more dangerous vehicles on the road (as the share of quality autonomous vehicles increases in the future), and I know that I'll buy one for my family (whenever that happens). Until my body fails me or I don't have enough time to drive, I'm going to ride my motorcycle and drive my car myself, while sincerely encouraging literally everyone else to get an autonomous vehicle. I know that there will be others like me, and I know that these people will either evolve into very good drivers or angry teenagers. Regardless, the road will be safer for all, and our collective time will be less wasted on piloting tasks.

You're not alone.

*EDIT I should note that I'm a U.S. resident. My daily commute is >50 mi, though I'm looking to relocate. Public transit doesn't exist for my home and work destinations without significant (>4 hours a day) layover and transit times.

While I do love this and think it is going to mae for a fantastic future, why wait also, for this and ignore fixing two of the biggest blights in the SFBay area; train and bus service.

BART is a cancer, Caltrain and Amtrak have service schedules designed by morons and make, I think, literally TWO farking connections to BART.

There is a hell of a lot of improvement that could be done to fix public transportation in addition to working on this.

> this is the news website for a startup accelerator.

Something people forget waay too often..

What about driving for fun? Please note, fun has no correlation to speed when it comes to driving.

I suspect people who like to drive for fun will continue doing so, and that eventually it will be a fairly uncommon novelty, like horseback riding.

It's funny you should mention horseback riding as the fun driving I was thinking of involves driving a a truck with a horse trailer attached to it. :-)

It seems like a mistake to remove manual controls entirely, not necessarily because of safety concerns, but because there are a fair number of times when you need the car to go somewhere without having a route. I'm thinking of things like:

Moving your car ten feet so somebody else can get out of the driveway; handling double parking during alternate side parking hours; moving the car because you were waiting to pick someone up in a no-standing zone and a cop just pulled up behind you; moving the car to just the right spot so grandma doesn't have to walk any more steps than absolutely necessary; parking on a lawn for a party or concert; backing up to just the right spot so you can hitch up a trailer.

The list could go on forever, and that's the point -- you can surely automate away some of these tasks, but there is an infinite multiplicity of things that need to be done with a car, and having to point them out on a map in order to get the car to do them, or even use a joystick to do it, sounds like an obnoxious chore.

Edit: to those who disagree with me by pointing to a utopian future of car sharing for everyone and no one even being able to drive, that's fine I guess, but this seems to be suggesting that the very first generation of these will be without a steering wheel, when very few people will have these cars and all of the problems I listed will still exist. Even if your ultimate goal is for passengers to be completely dependent on the AI for navigation, that's just not going to be an option when these things roll off the lot as soon as 2016.

I agree with you, perhaps in the context of a personal vehicle there does need to be manual control.

However I can see these more as a form of private public transit, a fleet of driverless google cars cruising around and anyone with a subscription a la zipcar can hail them and go.

Perhaps in this context, Google could operate parking garages where they can be recharged and cleaned while keeping them off the street and out of people's ways.

That sounds cool, although most cities already have something similar in the form of buses, which are cheaper and more efficient than what you're talking about.

In any case, they will probably not start by having a huge fleet, but by selling them as personal vehicles.

I wouldn't compare it to buses, I'd compare it to taxis combined with zipcar. Buses don't have the destination/pickup resolution that a small 2-4 seat vehicle has.

Buses don't need to be on fixed route. In Helsinki some buses are using dynamic routing based on demand. So basically those are shared taxis with drivers. When the driver can be dropped out of the equation service can be cheaper and there can be a lot more of these vehicles.

https://kutsuplus.fi/home

You're going to need to translate some of this for me. How long in advance do I have to signal my intent to get the bus? How frequent are the busses/how long will I wait on average? How close will busses come to me, will they come to my house in a court where it needs to three point turn to get out? Will the bus wait 10 minutes while I load my things in it?

I don't think the google car will replace mass transit like busses, trains and trams. They're a far more blunt instrument. I do however see it as a more flexible hybrid of car sharing and taxis. A taxi fleet that doesn't have the overhead of keeping enough drivers active for small demand spikes and can be sent directly to the user.

Edit: Weird, was not showing up in English on my phone, does on my laptop.

I assume they're using accept-language. Because it's in Finnish on my work laptop, but on my home desktop and mobile it's in English. Anyway, here's link that goes always to English version. https://kutsuplus.fi/home?lang=en_GB

All the signs right now point to Google's self-driving cars being a fleet, not personal vehicles. A few points:

* Google has expertise in "big data" and algorithms: running a fleet would fit their expertise.

* On the other hand, Google has no expertise in designing or marketing pretty cars. They don't have any dealership or repair network. To release a personal vehicle, they'd have to team up with a car company (and that partnership would have to be in place by now given Google's aggressive timeline).

* A fleet lets Google control the deployment carefully. Nobody is going to buy a self-driving car that only works in Mountain View. (Or, worse, they would buy the car, and then complain loudly.) But with a fleet, Google can start on a corporate campus, expand to Stanford, say, then Mountain View, etc. No worries about driving in the snow. No worries about self-driving cars getting in well-publicized accidents after an owner ignored the maintenance schedule.

These would be buses on steroids.

Imagine if you could fire up an app on your phone that would make a driverless car magically appear to take you somewhere. Once you get to that somewhere, it would scurry off to take someone else somewhere or recharge or go towards where Google's analytics show that it's likely to be needed next.

I think the eventual goal will be that most cars will be self-driving and that will negate most of those problems.

I imagine that the vision of the future is a giant car-sharing program made up of self-driving cars all coordinated by Google. Think about it - you fire up an app to request a car, one appears at your door within a couple of minutes, takes you to work, then disappears. After work, a different car picks you up, whisks you home, then scurries off to recharge/refuel/drive someone else. I suspect that in the long term the thinking is that things like "parking" and "owning a car" will be less common than they are today.

There will still be a place for manually driven cars, like ones pulling a trailer, etc.

Correct. That's one reason I think that Chicago's decision to privatize parking was an inspired one. The city got a great deal of criticism for giving a decades-long revenue concession to a private operator for a relatively low up-front cost, but all the critics assume a constant or increasing demand for parking based on historical usage patterns. I think usage patterns are going to change radically and that we'll see far lower traffic densities in the not-too-distant future.

It was a desperation move, not an inspired one. Chicago and the entire state of Illinois are in terrible financial condition.

AAAS - Automobiles as a Service.

That could actually be pretty cool. Basically, a ZipCar + Uber that doesn't suck. That would be the dream.

See http://techcrunch.com/2013/08/25/uberauto/

I miss the days when you could read science fiction and think, "none of this exists, I wonder what the future will end up looking like?" We are way past the conceptual stage for the future of transport, and I think it will look exactly like the mashup you are describing.

It could even be improved with printable self-driving cars, areas would have a base load of self-driving cars, but in case of need/rush the area center would just print new cars.

My biggest fear re. cars as a service is degradations, it's been an issue in at least some of the bicycle sharing systems.

I can't see that working. You'll require a ton of cars taking individuals or small groups to and from work in a big city. That's not very efficient, and the reason we have a metro system in the first place.

1. That's a lot of pollution. 2. You have twice as much traffic, since these cars need to come to your house every morning, and drive back to their charging station in the evening. They can take those trips off-peak hours, but you still have a lot more cars on the road. They could stay at your house overnight to save on wasting energy, but at that point you're just owning a car, or renting one full-time. 3. Do many people take a taxi to and from work daily? No, and this is quite similar. It's too expensive, people take public transportation, or buy a cheap vehicle (this could be that vehicle).

This is a replacement for the taxi, and for car rentals. If you live in a city and need to rent a car for a day, this would be a great situation. You don't even need a driver's license, so you just push a button, car appears, and you're taking it to your destination. I've lived in a dozen different cities, and this is something I'd use. I don't rent cars often because it's kind of a hassle, a little scary driving in a new city, and a little annoying searching for parking, but the convenience of pushing a button would encourage me to use it for certain trips. It should also be slightly cheaper than a taxi since you're not paying or tipping a driver.

They could probably get creative and offer monthly plans, where you get up to X trips, and Y kilometers, for a slightly better rate than renting on a per trip basis.

1. Less pollution than what we have now given that many people spend time hunting for a park when they arrive, or drive five-seat vehicles even though they're alone or with 1-2 others. I read something saying that in a city centre, often 30% of traffic is simply people hunting for a park. I know when I arrive near my office, I can spend 5-10 minutes driving around trying to get an all-day park. Three times a week I'm driving an SUV that can seat five and has a large boot (passengers are my wife and toddler), the other days I'm in a five seat sedan (and solo).

2. I'd say that depots will be interspersed around the city and suburbs, for one thing. Cars might move from suburban depots over night to densely-arranged city parking/depots during the day, waiting for the evening rush. A car that dropped you at work after your run in from the South is then immediately free to take someone from the city centre down South again, or be available for a courier job, or a tourist leaving a city hotel for a day trip. The mesh of cars communicating will determine the most economical use of the vehicles and anticipate demand. Further, cars communicating with each other will mean that typical causes of congestion (like traffic lights) will be better handled and less of an issue.

3. Taxis are expensive partly because there are salaried humans driving them. If these are electric vehicles and solar charged, fuel and driver are largely out of the equation.

1. Not sure I agree with the first point. People hunting for a parking space once a day likely causes less pollution then a car driving itself to a charging station or depot twice a day. Plus, if these are an affordable method of commuting, you're going to be pulling people away from public transportation, and getting more vehicles on the road. If they're more expensive, then you'll just have people buying their own, and then they're in your situation, where they buy an SUV model since it's more versatile.

2. Moving some depots outside the city will definitely help. Yes, some of the cars going into the city will be used for courier jobs, or tourist day trips, but that's an incredibly small percentage. You might have 500,000 automated cars taking people to work. 450,000 sit in a depot waiting for the evening commute, and 50,000 are used for people active in the day. Once again 500,000 people take their car home, and about 450,000 sit idle again waiting for the next day. How does this work as a business? If you use one of these cars for work, you need to pay for 90% of the cost, because that car rarely gets used outside of taking you to and from work. Sure, you could do a ride share with a couple of other people, but still, the few of you are renting that car full-time. It doesn't work.

3. If these can be solar charged, then surely a taxi can be solar charged, so there's no difference in fuel cost. You're paying a driver, and someone in an office answering calls, so there is an extra cost, and I think they have a use case here.

In short, these cars need to be on the road all day long, that's how they lower their prices and raise efficiency. That doesn't work for daily commuting. However, they would be an incredible taxi service. Imagine you have a fleet of 100 automated cars, people use an app to set their pickup/destination, and the cars find the most efficient way of organizing the routes and schedule. A car breaks down, and another one automatically reroutes itself. You could give people the option of carpooling with a checkbox. Let people register on the app, and upload a photo. So, you set your destination, and it says it'll be $10 and 15 minutes, or you can carpool with John (25/male), and it'll cost your $7.50, and 20 minutes. This could even happen while you're in transit. You get a pop-up on the dash. Want to lower your rate and share a ride with the person in this photo? Say yes, we'll credit $5 to your account, and take a 5 minute detour to pick them up. Worried about safety? You could review passengers, and only car pool with people that have a high rating and good feedback. I think that's exciting, it should offer lower rates to consumers, and you get to socialize with new people at the same time, so you might make some friends on your next taxi ride.

1. Public transport will still be cheaper, so it will still have its place. And I think we'll see small charging spaces that autonomous vehicles can use every block or so. Plus parking for these vehicles will be dense and save loads of space - it is obscene how much of our space is dedicated to parking.

I think we'll also see something in between a typical commuting vehicle and a bus. If 10 people in a specific area have regular bookings to go from a two block area to the CBD in a fixed time, then a van will do pick-ups and drop-offs either from houses or from fixed corners (think unmarked bus stops). It will be like demand-based bus routes.

I think we might also eventually see pod-based vans for people that don't want to socialise in any way with others in the vehicle. That said, people cope with the subway OK.

I don't think pollution from driving will be a significant factor with a serious influx in electric vehicles.

2. That's worse than what's happening now. I think we'll see peak pricing motivate changes in schedules and a decrease in the full complement of cars. I also think we'll see companies selling time-share in vehicles so you might buy one of these cars to have priority access, but then earn money from whenever it's available to work for you as part of a network. Those that want to keep one dedicated to themselves could.

3. Staff costs are significant in Australia so here and in similar places, I think taxis will be increasingly popular. Already it's not far off the case that commuting via taxi outside peak times could be more affordable for me with a second car I use twice a day, twice a week. Four taxi trips would cost me $12-17 each ($2500-3500/year). My second car, ignoring the cost to buy it would reach $2500 annually from registration, insurance and maintenance. That ignores depreciation, cleaning, hassle of parking, walking to my parking spot in the rain four blocks away, etc. Take away salary costs of taxis and make them easy to hail and pay by phone and I'd switch so quickly. And I think many people would prefer to travel by autonomous vehicle than with a driver who smells, listens to crappy radio stations or answers their phone the entire trip.

A bus is possible, but if this was a viable business, it should exist already. The only difference is the cost of the driver, so the automated bus might save passengers $5/day. Is that $5/day extra cost keeping this business non-existent at the moment? I'm a little skeptical.

Keep in mind your taxi service is cheap because it's on the road all the time. A taxi might make $250/day, and pay the driver half of that cost. That means the company is making $125/day per taxi. You're one of 15 or 20 people taking that taxi in a day, so the $15 trip is possible. If everyone was taking an automated car to work, they need to pay the majority of that $125/day price. Split it with 4 people, and that's just over $30/day each. Kind of expensive. Add more people and we're turning into a bus or public transit again.

Work schedules offset would help everything all around. If some people worked 7-3, 8-4, 9-5, 10-6, 11-7, etc, then we could decrease rush hour traffic, and use the cars multiple times to lower the price to $10/day, which starts to become interesting.

I like your idea of buying a car, adding it to a network, and getting money back when you're not using it.

Everything relating to public transportation will be the major pain point. It is heavily regulated, subsidised, unionised, etc.

London isn't able to replace ticket sellers with machines yet, so you can imagine how long it will take to replace the drivers. Alternatively, they will see this a profit eating activity and resist until the people wouldn't take it anymore.

When I was in London in 2009 I bought metro and surface train tickets from a machine.

Humans are still an option. And there are few other roles that are redundant, but every two months workers strike to keep them.

"A bus is possible, but if this was a viable business, it should exist already. The only difference is the cost of the driver, so the automated bus might save passengers $5/day. Is that $5/day extra cost keeping this business non-existent at the moment? I'm a little skeptical."

$5/day could be enough, especially if the parking situation changed for drivers at the other end. Take away street parking (no longer required when most cars are autonomous) near my office and I'd likely catch the bus. Have a bus that took me virtually door to door (rather than walking 3-4 blocks at each end) and I'd likely switch already.

I think lack of imagination might be one thing stopping a service like this existing now, and volume of passengers. If you wanted to start it now, you'd need coverage to justify the advertising campaign to build customers and to make it worthwhile. You'd need a fleet and then to hire and train drivers.

In the future, it will be APIs/software.

Imagine it's almost worth my while to pay $15/trip now. Then take out the driver and make it $10/trip plus pain-free bookings - I could read/work without feeling like I was being rude to the driver. Then imagine my booking app offers me $5/trip if I'm willing to car-pool, with the software automatically picking my ride-mates and most efficient pick-up and drop-off route. I'm pretty likely to do that. In dense areas and with time, manoeuvrable vans could be even more effective and the upgrade could be done with a software update.

Not to mention the bonus of taking cars off the street.

Autonomous vans on demand are going to be incredibly useful for wine tasting trips...

> A bus is possible, but if this was a viable business, it should exist already. The only difference is the cost of the driver, so the automated bus might save passengers $5/day. Is that $5/day extra cost keeping this business non-existent at the moment? I'm a little skeptical.

You're already seeing the much-maligned google (and other tech companies) buses in SF and area, so they're just-about-viable for some cases. I've seen a similar business model for hotel-airport shuttle services, and these kind of vehicles are everywhere in denser cities (Hong Kong, Firenze).

You can price the car based on the time of the ride, to spread out rush hour and reuse vehicles for multiple trips. It will always be more efficient than privately owned vehicles due to the ability to choose the appropriate vehicle for the ride, instead of having to buy a vehicle for the most intensive purpose and underutilizing it a majority of the time.

I think you're underestimating the amount of pollution/resources that go into making a car. Reducing the overall number of cars produced would offset a huge number of those extra trips.

I think people forget about this point, especially tyres. They take a ton of effort to make and you cant really resuse them for much afterwords.

I keep hearing the "there would be fewer cars" claim, but my counter-claim is that the mileage would be the same, so wear and tear would be the same. Look at how quickly taxis decay to see this.

Furthermore, I looked up reuses for tires. There is already a large, multi-stage chain of reuses and recycling for them, for all kinds of applications, from burning for cement kilns to civil engineering and agricultural products.[0] [1] [2] As one article put it "Those same characteristics, which make waste tires such a problem, also make them one of the most re-used waste materials"

[0] http://en.wikipedia.org/wiki/Tire_recycling

[1] http://www.designrulz.com/product-design/2012/12/20-ideas-of...

[2] http://www.earth911.com/living/travel-living/recycling-myste...

It will presumably be cheaper than taxis, that is kind of the point.

I think most people don't even realize how expensive their car really is (is there such a thing as a "cheap car", if you take into account taxes, insurance, gas prices, maintenance, time wasted looking for parking and waiting in traffic jams). Convenience of autonomous cars might quickly win them over.

Charging: maybe you have a point for people who commute from suburbs. But you could have charging stations in the suburbs. I think there will be algorithms for determining optimal distribution of cars and charging stations. Apparently the Tesla can be charged at a normal power socket. If that goes for other cars as well, there would be no reason not to have lots and lots of charging stations (contrary to gas which is explosive and toxic, so you don't want to have a gas station near your house).

Anybody with a house front could perhaps offer a charging station to earn some money on the side.

You're right, public transport isn't going anywhere and still has huge efficiency benefits. But these cars can make it better, by solving the first/last mile problem.

I imagine them ferrying people frantically to and from mass transit hubs, especially between low-density suburban housing and trains.

You are going to have to, first, make the mass transit so awesome that people voluntarily leave their cars behind to get on the train.

As you get more cars off the road, other people will use the open roads to travel into the city.

Note that "you have to make mass transit awesome enough that people use it" is not "make driving suck." You will never make any headway in a democracy by telling the majority of the people they should re-arrange their lives to follow your utopian model.

I imagine this working a lot like car2go, which I've used to commute in Austin. The cars are rarely at a central depot except for maintenance. They're left wherever you last used them until someone else gets in. This could only be better than that, as users won't have to walk to find one.

There's no reason for the cars to drive back into the city every evening. They could charge themselves while you're at work if they had to charge daily.

If you have automated cars at either end of a rail system, you get rail efficiency in the middle. This alleviates one big reason not to use trains to commute - the stations are farther than walking distance at one end. If you get an automated ride to cover the gap between the station and your work - it complements both systems.

Why not go the whole hog and just shuttle people to whichever house has an open bed for the night? With some smart scheduling software, you could ensure that families get shuttled to the same location.

I realize you're being facetious, but these already exist.

They're called “hotels.”

It doesn't take a huge stretch of imagination to think that if the marginal costs of operation were similarly reduced by automation that it could become economically feasible to reside solely in hotels.

Google will be unable to centrally coordinate the transportation needs of everybody in any way approaching the ability of individuals to autonomously manage their own needs. What would actually happen in this scenario is persistent shortages of cars in times of high demand and long waits.

In the long run, if self driving cars work, people will no longer be qualified to operate their vehicles. Would you let your toddler move the car 10 feet to pick up grandma?

That said they will almost certainly have a manual override in these things, it will just be tucked away out of view.

With approximately 30 thousand road fatalities every year, people today are barely qualified to drive. But with so few viable transportation options in most areas, we make do with a bad situation.

I think there's room for a control system that gives a human a reasonable degree of direct control over the car's movements, while still having it be able to refuse potentially unsafe commands.

Compare it to a microwave oven. The microwave flux inside the oven is high enough to cause severe burns if you were exposed to it, but the system is designed so that it won't run with the door open -- and it's fail-safe, in the sense that any failure (or combination of failures) is overwhelmingly more likely to result in a safe shutdown than in unsafe exposure.

It's still possible to burn yourself on hot food, of course, but my point is that nobody thinks of a microwave as a dangerous piece of equipment that you need to be "qualified" to use, in the way that they think of a car or a table saw.

Just download the Google Steering Wheel app.

If the car's display shows a map of the surrounding area, and that map includes not just static objects like buildings, but also includes other cars, people, etc. (based on sensor information), is touching a spot on the map to make the car go there really a chore?

Touch-screen interfaces are inherently imprecise. Ever tried to select a destination on a GPS map?

Ok, so how about if you can drag a representation of the car to the place on the map where you want it to go? This doesn't seem like a very difficult UI issue to me.

That depends entirely on the quality of the touch screen interface. Most GPS navigators are just terrible. The precision you get with a normal tablet or smartphone is just astonishing in comparison. Sometimes I'm pissed because I have troubles aiming a link which is 1/10 the size of my finger (a desktop site rendered on my phone) ...

Ever used the equivalent system built into a 2019 model car where the absence of a steering wheel and other driving instruments mean that the interface could be two feet wide and easily zoomed for precise navigation?

Current GPS devices are built to be cheap, compact and work with limited connectivity. I think we'll see amazing things in 5-10 years.

Have you ever played any 3D game on a 2D screen, where you had to click on a position to tell the on-screen character to go somewhere? Did they always go where you wanted, or did you encounter problems created by projecting 2D points into 3D space?

People seem quite comfortable with the joysticks on those electric mobility scooters that are all over the place, and I certainly see more with joysticks than with handlebar steering nowadays so the market seems to disagree with you.

I'm not sure why you and a few others think a joystick will be problematic. Of course, there will be a bunch of tasks/use cases for which these cars won't be suitable, but so what? there's many use cases for which bicycles and motorcycles are unsuitable, but lots of people still use them.

I can certainly see using one of these things for local get-around driving, because I hate to drive and my wife can't.

For manual controls your could just plug in something like a PlayStation controller into a USB port or even a Phone app via Wifi/Bluetooth to control the car.

No need to have permanent wheels, levers and pedals cluttering the "passenger" compartment.

It's always nice to have an actual mechanical linkage to the steering and motion, if the electronics had issues it could just as well affect the BT link from the phone or controller. I'm not familiar enough with all makes and models out today and perhaps there are plenty of examples where wheels are already just "controllers".

Essentially every new consumer grade car made today comes with power steering and electronically controlled or assisted "mechanical linkage". To say nothing of the recent high profile cases of large auto makers having faulty engine control software or mechanically unsound ignition systems.

It would almost certainly not be worth it for a robocar, but having an assisted mechanical linkage is valuable and very different from a pure electronic linkage (i.e. drive by wire). As emgee points out, drive by wire is essentially non-existent in consumer cars.

The technical term you're looking for is "drive by wire" and it actually only came onto the market in the last year on the Infinity Q50.

Sure, Steer by wire is new. DBW generally refers to the connection between the accelerator pedal and the throttle. There, DBW has been the norm for many years.

Surely not the brake, though. Right? Right?

Edit: http://en.wikipedia.org/wiki/Brake-by-wire

Mechanical failure was always possible with brakes.

Even if the pedal is physically connected to the brake cylinder ABS can do whatever it wishes to: ABS can apply and unapply brakes at any time, also asymetrically.

This years F1 cars are brake by wire. It let's them control the kinetic energy recovery better.

Haven't recent cars and even aeroplanes shifted entirely to drive by wire? The mechanical linkage is already gone for the most part.

Kinect.

Note that all those examples are very low speed uses (< 5 mph). It would be completely acceptable to control the car with a handful of dashboard buttons no larger than the buttons to roll the windows up and down. (As mentioned by other, pedals and a steering wheel would be overkill for these rare situations.) And there's really no need for Google to have mentioned it in this blog post, nor would you have seen it in their rendering, so the post is completely consistent with your concerns being addressed.

Not to mention snow/ice. The software might be able to handle the nasty Minnesota roads, but not for a while (and what if people don't trust the cars in those conditions?). As far as I know, they've only been tested in sunny California.

That statement was probably a bit of a jump.

What is it that you think makes handling snow and ice more difficult? Feedback control systems can respond an order of magnitude quicker to loss of traction/unexpected behaviour than people can. Icy conditions are different, but not necessarily more complex than "ideal" road conditions. From the point of view of a driving control system it just requires a slightly different control law. Certainly there will be testing required, but I'd be surprised if it presents a major issue. Adverse conditions are one of the more obvious applications of driverless cars.

Another option might be to use a collapsible steering wheel.

It could neatly and automatically retract into the dash, when not needed.

How about something simple like telling the car, "back up 10 feet" or "move to the spot across the street". This is dead simple short range routes...

How does one "tell" the car what to do without controls ?

I hope you're not suggesting voice recognition because anything less than 100% accuracy will guarantee accidents and possibly deaths.

The car needs to know what's safe regardless of whether or not the instructions are via voice or another method, so inaccurate voice control does not guarantee accidents.

yep, http://en.wikipedia.org/wiki/Three_Laws_of_Robotics

A smartphone app would probably suffice, given that I can control a drone aircraft with one already.

The same way you tell it drive anywhere else without controls.

I'm pretty sure current cars with controls guarantee accidents and deaths

How sure are you that cars without controls guarantee neither?

Because it's provably possible to make a car that drives better than the average human. It's not about guarantees, it's about improvement.

It's not provably possible yet - the average human is quite capable of not killing themselves or others in any number of hazardous and random driving conditions which autonomous cars can't even manage.

I should probably mention since I typically bleed karma in these threads, that i'm not at all against autonomous vehicles in principle. I can see a lot of ways in which they could make driving safer, and transportation costs cheaper, and do a lot of good.

What I am against is the premise that people are so dangerous and terrible at something they've been, mostly, able to do safely for about a century that their freedoms need to be taken away. Autonomous cars are fine - trying to take for granted that personal freedom is dangerous and stupid and that people need to be controlled like herd animals is not necessarily a good thing. Where is this absolute certainty and conviction that I see in these threads coming from? Certainly not from having observed several million autonomous vehicles on the roads.

>What I am against is the premise that people are so dangerous and terrible at something they've been, mostly, able to do safely for about a century that their freedoms need to be taken away.

People are dangerous and terrible drivers. There are 30k vehicle accident deaths each year in the US (1.3 million world wide), and it used to be much higher (especially per mile driven). Number of deaths per mile driven has been going down drastically in large part because technology has limited a driver's direct control of the vehicle.

> Autonomous cars are fine - trying to take for granted that personal freedom is dangerous and stupid and that people need to be controlled like herd animals is not necessarily a good thing.

ABS/traction control/drive-by-wire takes direct control away from the driver and gives it to a computer. Self driving cars are just the next layer of abstraction above this. You're still in control of the car, it's just that the interface is a touchscreen with a map instead of a steering wheel.

Because it would have to immediately take the action without any sort of verification check?

I wouldn't want to ride in that car either, but that's not how anyone remotely intelligent would design it. We're not talking about unscoped directions either which is a much less harder problem then general recognition.

There also needs to be a 'pause for others' button so that a motorist in a side street can be let in to a congested road or for a motorist on the opposite side of the road wanting to turn can make their move. This is the usual scenario where a motorist flashes their lights to let someone through.

As a cyclist/pedestrian I am always amazed by this altruistic side of motorists and it would be a shame if 'the computer said no'. A little 'be nice' button could be a cool feature.

Why does it have to be a button? Can't the computer just be programmed to be nice by default? Or then we'd need a "oh c'mon" button?

Actually, if as you approach a traffic signal, you cede control of the car to the signal, you can avoid red lights altogether as the smart signal shuffles the cars through without any of them hitting each other, handing off control back to the car after it passes through the intersection. A similar technique could be used on on-ramps.

I don't think the parent was talking about traffic signals, just a side road with a give way/yield.

If both roads are congested then "always let everybody in" and "never let anybody in" will both result in deadlock.

Another example is the pedestrian crossing outside my house. There's no traffic lights. Usually cars will drive right past and you (as a pedestrian) have to wait for a gap in the traffic, but sometimes they'll be nice and stop (especially for children).

It's always possible that the end result will be a hybrid one that's somewhere in-between the two visions (particularly with other variants, as it seems like Google's primarily targeting an urban environment). There are already concepts for retractable steering wheels [1] that could be combined with retractable pedals to handle situations that require manual control. Hell, the movie Demolition Man had them way back in the 90s alongside with plenty of cheesy acting.

[1] http://www.webdev.trw.com/sites/default/files/TRW_OSS_SWS%20...

These sort of low speed manoeuvers are in my opinion quite awkward with traditional control and viewing systems, which is presumably why there has been so much development of parking sensors and cameras, and now automatic parking. I think many people would prefer being able to set where they want the car to go on a top down schematic/plan and let the automated control systems wrangle the car into place.

you could just plug in a joystick

When you try to paint this as a beta of a marketable vehicle, yeah, it doesn't make much sense. However, this seems much closer to the long history of automotive concept cars, which have traditionally been impractical and designed to definitively show off certain features that may or may not make it to the final cut. By building a vehicle from the ground up without a steering wheel, they're much more readily able to get people asking the questions as to what's really needed in a car of the future.

No one said that there won't be cars with manual controls in the future. There are certainly use cases for both. The market will decide which cars are more popular. There's no need to play pundit on the Internet.

I would not associate the interface with the functionality. Even without a keyboard, a smartphone with a touchscreen allows one to use a virtual keyboard to type in a similar fashion to an actual keyboard with some definite compromises. With cars when the manual controls are removed, we have a chance to have even more robust functionality with opportunities to map a whole new set of controls to a fallback interface.

Some alternative input systems can be useful for this type of maneuver, such as a gamepad or smartphone via app. Perhaps this is done from inside the vehicle, but in the case of attaching a trailer, this could be done from outside the rear of the vehicle to guide it precisely.

Given that they're capped at 25 mph, these are legally "low-speed vehicles", not cars, which means they're way less regulated.

https://en.wikipedia.org/wiki/Low-speed_vehicle

I think, too, that there's a logical disconnect between "we focused on learning, not luxury", and "we want to hear about what you want in a vehicle where you sit back and enjoy the ride" (probably, to most, -not- a dearth of comforts).

Think of this as a bike-commute substitute with a roof.

If only my bike powers and steers itself

Without the exercise capability.

You could remove the seats and install an exercise bike.

that's a terrible analogy :) ! more like a taxi without having to speak to the driver or a personal bus.

An interesting thought. How small can self-driving vehicles be? Would you ride a self-driving motorcycle? A self-driving wheelchair? How about a self-driving Segway? You get on, hold on, and it does the rest. The technology for getting this working at all (let alone safely) is a long way away, but perhaps it could happen.

As long as it's safe (duh) and moderately comfortable, I'd be just as happy to read or suchlike. Chances are these cars have good internet connectivity as they're probably going to be used in cities and you can stick an arbitrarily large antenna into the car body, so work, gaming or video consumption are practical too (on your tablet or other personal device). what else do you need - a cup holder and easy-to-clean interior?

Lets be clear here. The aim here is to let you sit back, browse the web, and click on Google ads.

this meme of everything google has to relate back to ads is wrong and needs to be retired. They already have non-ad revenues above $1B/year. They are clearly investing in non-ad businesses across the board.

It's because of the 55B they made last year, more than 50 of it was from advertising. I'm sure Coca-Cola is more than a soft drink company, but that's the main thing. The only real threat to google is a threat to their ad revenue. So, it makes sense to think of google in terms of ad sales.

I think when cars themselves came about, speeds of 10mph were considered breakneck and dangerous. I remember reading an old crash report where the author claimed that the driver must have been going at least 10mph and thought it was lucky nobody was hurt.

I guess to gain the trust of the public and politicians, self-driving cars must go through a similar process. I have met people who have said self-driving cars will never take off, never match the skill of a well trained manual driver and that they dont trust them. I guess it's slightly biased in that they have been driving for many years, one was my driving instructor and all are not considered technologists.

I'm looking forward to the experience, personally. I do, however, enjoy driving :)

Early cars also had wooden wheels, no tyres (well, strips of rubber, but no air), and no suspension to speak of. Going 10mph in one of them probably would've felt like 50mph in a modern car!

People would have had experience travelling at speed because of trains. Even horses can travel faster than that. I do remember reading about the first train and there was worry that people couldn't survive at speeds greater than 25 mph.

That's very handy when you're building a bunch of barebones prototypes. Clever Google!

I wonder whether this might be a category with long-term viability in certain city environments - light, small, padded cars that travel at 15-25 mph would be able to mix in a much more natural way with pedestrians and bikes than traditional road vehicles.

I'm reminded of those Dutch-style shared space schemes:

https://en.wikipedia.org/wiki/Shared_space https://en.wikipedia.org/wiki/Exhibition_Road

When the man with glasses gives his perspective, the enablement hits you. If you've any humanity, you feel the potential this technology brings to so many.

I can relate to this. Anyone who has ever had to take a family members car away from them as they get older can as well. When someone has driven most of their life, losing the ability to drive can be a very difficult. To them its losing their independence. Self driving technology is very much about enablement.

It's exciting to think that for the younger people now those sorts of issues are likely to be "fixed" when they get older.

The man with the glasses is Steve Mahan, the blind man from their previous video:

https://www.youtube.com/watch?v=cdgQpa1pUUE

I think it's interesting, also, to think not only of the freedom this would bring to people with disabilities, but also just to children, and their parents. Kids could have a prepaid card/ticket which takes them only within 5 miles of home, or only to certain preset destinations. It would make it far easier to meet up with friends, go to youth clubs or social clubs, or come back from school, in places where the school buses are not up to scratch. And parents in turn don't have to spend hours joylessly ferrying kids around.

I think the biggest news was this sentence:

    If all goes well, we’d like to run a small pilot program here in California in the next couple of years.

This is the next step in their very slow but, apparently, very steady march towards real self-driving cars that anyone can buy.

Why buy a self driving car when you could rent one as needed?

I used public transit, taxis and car sharing services exclusively until I started a family. At which point setting up the carseat every time became prohibitively cumbersome and I bought my own car. Children's car seats are very personal in terms of size and configuration and are required by law.

I think things like that will make people want their own self driving car. You want the one that fits your wife and two kids and your dog. Or you're rich and you want to be driven around in a really nice one that hasn't been used by other people.

The more I think about it, the more I imagine rich people increasing their car ownership due to self driving cars. Take your fleet with you to the cottage or across the country, why not?

I fully agree with all of the above. I also feel that people who have the means to purchase a self-driving car will do so. The compromises necessary to make a vehicle appropriate for use by multiple users conflict with the comfort and luxury many people expect in a vehicle they spend hours each day in. Would you like to spend your hour commute each day in a bus or taxi seat when you can afford the comfort of a Mercedes, BMW or even Toyota seat?

(I have a toddler, so I drive a car with a child seat.)

I think we'll get over that attachment to specific car seats and instead more versatile seats will come into play.

In the future, I imagine your car booking app will recognise your usual bookings (e.g., solo/couple, family with one toddler, moving cargo) and you will tap a preset to book whichever car you need.

If your partner and kid are heading to a family engagement, up rolls a car with a baby seat pre-fitted and cleaned. If you're needing to move something large, a van shows up.

Those that want a premium service, will book the platinum tier, either all the time or maybe just for date night.

In NYC, at least, taxis and for-hire vehicles are exempt from car seat laws: http://www.nyc.gov/html/tlc/html/faq/faq_pass.shtml#9

I wish I could use car sharing and the baby seat issue is a problem with that. It seems easy to solve, though - I don't see why those couldn't be integrated into the cars by default. A lot of taxis already have that, they are just sometimes too lazy to take them out.

I don't know what you mean by personalized car seats - I think there are merely some variations dependent on size of the child? But car seats can grow along with the child, usually you can leave more and more stuff out the older the child gets?

Not sure why car sharing services overlook that atm - maybe because most families tend to own a car, so it is not a big market for them?

For most people who own cars, they are an extension of all their other personal space.

I want to be able to leave things in my car and not have to take them out or replace them each time I drive. My sunglasses, the car seat, old-fashioned maps. Maybe I'd like to be able to leave my potato chip wrappers in there without worrying about it, or maybe I'd like to own something where I'm guaranteed others haven't left their potato chip wrappers behind.

Cleanliness. Who else has been in this thing? Did they leave lice behind? Did they jam the window so that when I use the car, there's an annoying breeze?

Style. My car says something about me. Maybe I want rims, or to paint it with flames, or maybe just pink. What about that stereo system?

Status. My car is mine and I can afford it. Like watches, nice clothes, or big houses, it's something people desire.

Practicality. If I don't live in a city, how long is it going to take a rental car to get to me? Why not own one that sits in my driveway, so I don't have to wait for it?

There are plenty of great reasons to want to own a self driving vehicle, instead of renting them.

I think the price of a self driving car will be very high, and the price of a taxi/rental service (which will no longer have a driver) will go down, making a taxi/rental much more attractive monetarily than it is at present. To your arguments:

- storing necessities - this is probably the strongest argument that will have to be worked around

- cleanliness - is this currently a problem with taxis and rental cars? I haven't had this issue.

- style - I think this is true in certain places (I'm thinking of places like LA where there is a strong car culture) and I think is something that will change as owning a car becomes an unneeded expense. Think of it like owning a plane. Sure, it could say something about who you are, but there are cheaper ways to do that.

- status - same argument as style

- practicality - especially at a larger scale, wait times could be greatly reduced by having more taxis/rentals on the road at any given time, including in the suburbs. There still would be a lack of convenience, but I think the idea of having an expensive piece of equipment sit idle all day in your driveway for the off chance you need to use it suddenly will seem wasteful. In fact it's wasteful now, it's just there isn't a good, cheap, and convenient enough alternative to use instead. A driverless taxi/rental could be that alternative.

I predict a maximum price of $9k, but suspect the target price is $5k. For reference a SmartCar in this market (California) is $15k. Low power and a minimal interior means the main costs are the roll cage/safety features and the drivetrain. The marginal cost of the self-driving functionality is low, and Google can amortize the fixed costs very easily.

About cleanliness, taxis and rental cars are clean because there is someone there to clean it between each customer. That doesn't work for carsharing, where cars go directly from a customer to the next.

My experience with these systems is that you don't find anything really dirty in the cars, but you tend to find empty bottles and things like that in them. Not a deal breaker if you just want to get somewhere, but not very family-friendly either.

Car sharing services in Japan ask you immediately (via the navi) what the state of the car is. I imagine it negatively impacts the previous rider in some way. A simple social pressure like that can work wonders.

> A simple social pressure like that can work wonders.

So would variable pricing based on how often you leave the car a mess (with appropriate filters to mitigate against the trolls).

I live in a mega city where taxis are cheap so I take the daily and avoid owning my own car. The big deal will be in cities like mine where parking is rare and expensive if found at all, while roads are extremely congested. Self driving cars solve the parking problem (which admittedly, taxis do today) + can use road bandwidth much more efficiently (which taxis cannot). In the end, I see all of these factors destroying personal car ownership in big cities. Now:

Cleanliness: this is an issue in taxis already. Sometimes I get into a taxi without window handles, but that is the government's doing.

Style: expressing yourself through your car sounds like a very American thing. Much of the world is not America, to say the least.

Status: this is the biggie where I live: those with money have to show off their Audi TT's and BMWs. Perhaps they will still own their cars, just like they do today (whereas many people just rely on taxis).

Practicality: my city already does not allow outsiders to enter within the 5th ring road during working hours (7AM to 7PM), while their is a lottery for "owning" a car with 20-1 odds per cycle. I can imagine that there will just be different vehicles for the country-side vs. the cities: the latter will be automated while the former won't be. You can own your country-side car if you want, but you can't drive it inside the 5th ring road given traffic unless it has a self-driving mode (to optimize traffic).

I never want to own another car. That is a huge amount of capital I could put towards Uber, or this.

Personally, I don't drive; it freaks me the hell out. However, I've got a housemate who does drive, and he has several thousand dollars worth of ham radio equipment installed in his car. I also have friends who do volunteer fire work and other first response type duties. Doing things like that in a rental would be near impossible.

Tesla (or insert company willing to go big on electric cars) and Google really need to team up on this. Self driving cars have the potential to completely do away with owning personal cars. Imagine a new public/private service that allows people to travel where & when they want without having to deal with the hassles or cost of owning a car. With the added centralization, the headaches that electric cars cause can be eliminated with a bit of logistics planning (cars are interchangeable so low battery cars can go straight to a charging station instead of wait for their owners to take them to one).

This suggestion comes up whenever the topic of self-driving cars arises, but I wonder if access to a fleet of self-driving cars would be more akin to a cheaper taxi service than a replacement for car ownership. I suspect that many people would still prefer to own their own (self-driving) car rather than rent one from a fleet only when required. However the privately owned self-driving car could still drive itself to the nearest charging station when not in use so I agree with your premise that self-driving and electric car technology has some synergies. Being a high-tech automaker, Tesla would be an obvious licensee for Google's self-driving technology once it matures.

Fair observation, but I think this sort of centralization will come after maybe 2 or 3 generations (40 - 60 years). With a long enough timetable I believe the sentiment behind private car ownership will dwindle until it becomes a hobby akin to collecting telegraph machines.

I think Google wants to do a similar thing to Tesla. Maybe have a few of their own vehicles but bet big on the technology. They can then sell it to the big car manufacturers. With Tesla it's their battery technology, with Google it's their sensor technology.

The video suggests these cars are probably being built in Tesla's factory - the white-painted walls and red-painted machinery are pretty unique to Tesla. See https://www.youtube.com/watch?v=CqSDWoAhvLU#t=58

(Also note that "Rotary Lift", the brand seen at that point in the video, is the exclusive supplier of vehicle lifts to Tesla - http://www.prnewswire.com/news-releases/tesla-motors-chooses...)

The NYTimes reported that they'll be built in Detroit;

"Google is having 100 cars built by a manufacturer in the Detroit area, which it declined to name. Nor would it say how much the prototype vehicles cost. They will have a range of about 100 miles, powered by an electric motor that is roughly equivalent to the one used by Fiat's 500e, Dr. Urmson said. They should be road-ready by early next year, Google said."

They could have used Tesla's facilities for the first prototypes though. My money for the first production run is on a niche house like Roush, or potentially Chrysler since these little cars share components with the Fiat 500e.

I'm reminded of a talk I attended about entrepreneurship by the founder of zip cars. She talked about how everyone she talked to would suggest she partnered with car companies. So that's what she did. She would go around talking to all these car companies, and then tell potential investors that's exactly what she was doing. It made for a great story.

Of course the truth was, in the meetings the car companies had no interest in working with her. From their point of view, zip car was a threat to their business. The little bit of money they may make from an exclusive contract would be smaller then all the people who might have bought cars, but didn't.

But the story about the meetings was nice.

I feel as though they really havent embraced the driverless concept with their prototype.

Why have seats that are fixed and facing forward? What about a couch? Why is there a dash?

The people in the video are essentially watching the car drive itself, when they could be doing anything else.

There are probably a multitude of reasons why the things above are the way they are. However, the prototype seems like just another car to me.

Also, off-topic, but can we have a google/product launch video without the corny piano soundtrack.

>Why have seats that are fixed and facing forward?

Seats facing forward means that acceleration pushes you backward into your seat. Any other direction would increase motion sickness. And what would be the advantage of the seats facing a different way?

>What about a couch?

The seats do look like a couch to me...

Rear facing seats are safest, since acceleration is not the biggest problem, rapid deceleration is.

http://www.airspacemag.com/need-to-know/are-aft-facing-airpl...

Not when your top speed is 25mph. You're far more likely to get rear-ended than suffer a head-on collision.

>what would be the advantage of the seats facing a different way?

When you got into a head on collision, you would (perhaps) be spared the whiplash.

>And what would be the advantage of the seats facing a different way?

In a four seater, two seats forward and two seats back, for conversations. With only two seats, there is little benefit except for a head on collision.

They should add a table, or two sets of seats facing each other, and an optional table that pulls out between them. Then show 4 people in the car, having a little business meeting on the way to their office, or have a family of 4 in the car, eating sandwiches or something on the table as they go to school.

That should be one of the key points, you can use this time productively, and not just spend it looking out the windshield. You can visit with your children on the way to their school, and not have your back to them as you divide your attention between the road and their conversation. Show a side-by-side of a parent driving, scatterbrained and frustrated, then show one sitting next to their child in the automated car reading a book together as it drives.

People want to know where they are going. Very few, if any, people would be happy looking at where they've been. The dash is most likely two-fold: they have to cover up what's behind the dash and they need a place to put things like air bags. Finally, if I was designing this, I'd want it to feel enough like a car that people fully recognize all the ways it's not a normal car to increase their excitement about it. Just like they put a cartoon face on the front, complete with a nose: make people feel good about it.

Ever sat backwards on a train? Or first class in a plane? It doesn't matter if you're not driving.

I know a lot of folks who get car sick sitting backwards in busses.

As was already said, some people get sick. Also, many more people prefer to be facing forward on a train. I know I personally do. Sure, I can handle facing backward with no problems, but I just prefer facing forward.

I much prefer facing backwards, because I know in the rare case of crash that I am going to be much safer.

This is also true of aircraft and cars and buses and coaches. Sadly, most people do feel uncomfortable facing backwards and so the risky forwards-facing seating is used.

I really don’t think that kind of risk mitigation is of much utility.

It would save many lives in cars.

One of the cars my parents had when I was growing up was a VW Eurovan. The seat configuration in the car was a 'sofa' (fits 3) facing forwards and then 2 seats facing the sofa. On the left hand side of the car was a foldable table. As a kid it was fun to use it to play cards with the person in front of you or to eat, etc. etc.

0:50 over at https://www.youtube.com/watch?v=IBt-013d7Bw

I absolutely agree that cars can be much more fun when set up with different seating configurations.

That looks cool, and sitting facing rear is safer too!

> Also, off-topic, but can we have a google/product launch video without the corny piano soundtrack.

I think we can thank Apple for that.

It is easy to explain no manual control. Google will include insurance cost into the service itself. If there is any accident, Google's insurance plan will cover it. If there is manual control, the insurance cost will be a mess, since you have to factor in the driving record of the person who is in the car. This alone would be a show stopper.

I wonder whether they will first become available as a product that individual people can buy, or if Google will open some sort of service where you can sign up for a monthly "car plan" for X miles. Or maybe partner with a taxi company/Uber, where you pay per ride but the cars are driverless. I imagine all 3 models will be tried at some point, and it'll be interesting to watch their development and which one eventually dominates.

Also, commuting could be done so much more efficiently. Imagine a fleet of 1 person cars that pick you up in the morning, drop you off at a more central location (than your house), at which point you get on a higher capacity (driverless?) vehicle - bus or van maybe - for the trip into the city. And the bus unloads into another central area, with another bunch of small cars taking people to their offices. I'd use the bus if it never stopped, and that system would be almost as fast as driving yourself. Heck, it'd probably be much faster, because the hordes of people people sitting alone in their sedans would be consolidated and eliminate a lot of highway congestion. Not to mention cheaper, and better for the environment (less gas).

I haven't heard much about addressing the legal and regulatory issues driverless cars are going to have to overcome, does anyone have more information on that? Obviously something will have to be done for when the car wrecks or malfunctions and damages something.

I really think that self-driving buses will be common long before most cars are.

  * They cover the same routes day after day
  * They are already centrally owned by large companies or governments
  * Driver cost is a big reason you can't have 2 small buses running twice as often
    as a single big bus. Or why buses can't run longer hours.
  * You still have limited space on roads and you need to use PT to get the
    person density up. This will always be hard with just one person in
    each vehicle

I think the driver on the bus has more functions than just being a driver. He's also an authority figure and can provide assistance.

There have been self-driving trains for some time now (Nuremberg, London Docklands). There are more people to be responsible for on a train and I do not see why a bus would be a special case.

Trainstations are somewhat controlled, bus-stops are not, they can be just about anywhere.

Which is relevant to the difficulty of designing a self-driving bus, but says nothing about any driver responsibilities outside of driving.

Not really. In case of a problem on board a train (unrelated to train's motion; for example, a medical emergency or a rowdy passenger), the train can usually be met by someone at the next station. In case of a bus, much more often there is no one to meet it at the next bus stop.

The London docklands railway have a staff member on board who opens/closes the doors and starts the train moving. It's a less skilled (and therefore almost certainly cheaper) job than driving a train, but it's still a person onboard.

Apparently not all the trains in Nuremberg are staffed, so perhaps it does work? I suppose the train driver is often segregated from the passengers in a way which makes them almost absent anyway.

It will be interesting to see if completely unstaffed buses are adopted anywhere, and if that proves succesful. I'd imagine that security wise a staff member isn't really necessary during busy day time routes (but then conversely that's probably when they'd be the most cost effective and provide the highest utility).

While I sort of like the idea, there's still going to be lots of traffic in and around those bus/van depots. We're also going still need storage near urban downtowns for all those 1-person cars that spend 95% of their day idle waiting for the 5pm rush.

You could probably get a lot more density with a custom built car storage facility. No need to accommodate pedestrians, cars of different sizes, or random access.

In between peak times, these vehicles could do same-day goods delivery.

One pulls up outside your office/residence and a Boston Dynamics Cheetah leaps out with your package between it's teeth.

Or the car acts as a mothership for sidewalk to door airborne drones.

This will eventually make non-automated driving sort of like a manual transmission (in North America) - something you get because you either specifically want one or because it's all you can afford (I'm probably way over-generalizing). Eventually, just like the automatic transmission, the self-driving car will be the default for most people.

I don't imagine that we'll get away from people who drive because they find it enjoyable any time soon.

> I don't imagine that we'll get away from people who drive because they find it enjoyable any time soon.

I can see that there will be a rapid change in social attitudes, once automated driving reaches the point where it is generally accepted to kill/maim fewer people than automated driving.

Think back 30 years, when it was socially acceptable to light up cigarettes around small children. Now it is socially unacceptable to do so, to the point that it is illegal in many jurisdictions. Where I live, it is illegal to smoke in the vicinity of a school, within 10 metres of a playground or in a car with a child under 16 years [1].

Once a tipping point is reached, where manual driving is widely accepted to be more dangerous that automated driving, manual driving will quickly be banned in the vicinity of schools. There will be resistance and arguments from manual driving aficionados. Eventually, public manual driving will be confined to either remote areas or purpose built facilities, just as public smoking is restricted to open-areas and smoking rooms.

[1] http://en.wikipedia.org/wiki/List_of_smoking_bans_in_Austral...

Edit: On reading other comments, I'd extend the analogy to smoking (and other activities that carry a public risk) by pointing out that there will be a financial penalty for choosing to drive oneself. Just as smoking is a risk factor that increases health/life insurance premiums, manual driving will also be a factor in calculating insurance premiums.

Exactly, and the way I see it is that driverless cars will remove the parts of driving that people don't enjoy, things like commuting, dealing with traffic, and parking (and running over pedestrians). But people who enjoy driving, are still free to a leisurely drive up the coast, or to race on the local track.

My ideal for driverless cars would be that, eventually, the ability for the cars to communicate with one another and detect each other would permit increased speeds and higher vehicle density on the roads. At some point, a human behind the wheel of a car would be like a horse-drawn carriage or a farm vehicle on the road.

Most manufacturers don't even produce manual transmissions on the majority of their models in the U.S. any longer, and, in the models that do have them available, they're sometimes more expensive. More than likely, the manual driving controls would become something like the pseudo-manual shifting available on some automatic transmission vehicles: you get a little more control over how the vehicle operates, but you won't be permitted to do anything that will be likely to damage the vehicle. I think we're already seeing a lot of this in the safety features of some higher-end cars (collision avoidance/warning systems) and features like automatic parallel parking.

This is going to change the world.

No more car insurance. Vastly fewer driving fatalities. Shared cars. Less parking problems.

I can't wait. Just take my money already.

I bet you'll still need insurance (or the provider will need it, in a service model).

It should be much, much cheaper though. Most vehicle accidents are caused by human error -- the insurance would just need to cover accidents caused by mechanical or software failure.

I think you're both right. It'll probably be fixed into the price, and they'll probably just lease them to avoid legacy issues.

Also comprehensive damages - hail, tornado, vandalism, or vehicular burglary.

One of the few pleasures of modern world is now slowly being removed! :-D

I guess I will treasure my memories of driving Lamborghini Murcielago over 200mph forever and will tell my grandchildren there was a time when you actually could drive a car yourself!

And to those that argue young people don't want to own the car - it's mostly about not wanting to spend your life commuting, hunting parking spots or in traffic jams, and frankly, daily driving to work in the States is as boring as it can get.

I don't own a car, I rely on public transport which is usually faster in European cities than driving your own car, allowing me to work on a lot of stuff while traveling. Whenever I need I rather rent a car - it's newer, serviced, no hassles and I can try different cars at will. From that point of view self-driving vehicle is wonderful (if it becomes eco-nomical/logical), though as a car-geek and F1/CART fan I really love to drive for pleasure.

in the self driving future, there will be a place for off road trucks and fast cars on race tracks. We probably won't let those type of vehicles on the main roads.

Still, if you think about it, what will be then the difference between having all cars self-driving, and public transport? All cars will be queued automatically, possibly flow-"optimized" at junctions, behaving like a special type of public transport (if we manage to create a proper real-time highly intelligent scheduling).

When you look at how Google computes the fastest route in their maps/online navigation, they basically have some precomputed grid of "main" traffic centers, and they route all local traffic to the nearest of those centers. The main centers have pre-computed distances/times/routes between each other, so they can be stored in a distributed hashmap, and the lookup of the fastest route takes nanoseconds. The remaining "local" part is then computed on the fly.

I am curious if a similar algorithm will be used in our self-driving future - you'd have local part of the route where you'll board a self-driving car and that will lead you to one of those "main" centers where a super-fast transport will move you to another "main" center, either by having you to board an express-train like device, or putting the whole car or just its interior there. What do you think? I think Chinese can probably experiment with such approach in one of their experimental cities ;-)

> One of the few pleasures of modern world is now slowly being removed! :-D

For most people the driving commute is the most unpleasant part of their life.

> I guess I will treasure my memories of driving Lamborghini Murcielago over 200mph forever and will tell my grandchildren there was a time when you actually could drive a car yourself!

Driving over 200mph has never been safe (and rarely been legal) on the public roads; if that's where you did it you were recklessly endangering the lives of others (I'm sure you thought you were perfectly safe; everyone does right up until the point they crash). If you did it on a track or other controlled environment, guess what: those places will still exist. Just as people still ride horses for pleasure, people will still drive cars. But those of us who just want to get to work in one piece will be able to.

Well, I live in Germany so go figure.

And please stop with your condemning talk! The ladies returning from a supermarket go often over 120mph over here on their way back home (legally!), you can see Porsche test drivers going full speed at 2am on an Autobahn almost every day, yet the accident/fatality rate is actually lower than on interstates in your country (I assume you are from the US). People here are drilled like crazy to obtain a driving license, unlike driving around the block and doing simple parking, and they behave well on the highways, letting faster cars ahead and not blocking them going 65mph in the fast lane or hastily switching lanes.

Not far from my city is probably the most hardcore racing circuit in the world open for public - Nürburgring, and what other car would be suitable for reaching 200mph if not one like perfectly serviced Murcielago? Yes, it's risky. So is eating HFCS in all your food daily and sticking to high-carb diet, depriving oneself of sleep due to working long hours or not taking vacation to recharge due to constant stress.

I am not talking about stupid reckless street racing - that's very uncommon over here.

Wrong guess, I'm not from the US. I've turned down high-paying job offers because they expected unreasonable hours or didn't have enough holiday allowance, and would do so again.

120mph maybe, but 200mph simply can't be safe, sheer human reaction times dictate otherwise. We don't allow trains to go above 125mph without in-cab signalling because that's too fast for a driver to observe lineside signals - and that's as simple as observing a coloured light when you know exactly where it's going to be.

No-one's going to close the Nurburgring, you'll be able to continue driving 200mph there with other people who've chosen to drive on a circuit like that as long as you like. But not on the road I have to take to work; I shouldn't have to put up with that risk.

I hear ya. And I don't think it's going to be quite so bad in the future. I doubt we'll see a time anytime soon where "traditional" cars completely disappear, but I bet cars like a Murcielago will be relegated to an even-smaller niche than they already hold.

The best analogy I can come up with is this: Plenty of folks still ride horses, it's still possible to get a horse-drawn carriage in many places, and there are more than a handful (though not much more) of Model Ts still on the road. But in all cases, that's not the majority anymore, and those that do that do so as hobbyists. I bet that's the future we'll most likely see, at least in the near term (~50 years).

I agree with your assessment. It's probably inevitable and actually will do a lot of good allowing people that are unable to drive to reach their destinations, as well as avoiding the need to waste time on learning how to drive (I know in the US it's pretty easy, but in most of other countries it's a fairly difficult and non-trivial endeavor to acquire a driving license).

If we can really pull it off in all real-world conditions, it would be pretty cool. At that point I wouldn't see any future for old school technological marvels like Murcielago anymore, as they wouldn't be allowed to interfere with automated traffic (human factor is unpredictable and we probably won't crack the problem of safely handling all situations automatically, though maybe with a huge sensor coverage the risks could diminish).

Personally, I can't wait until we invent teleports capable of transferring living subjects, though that seems to be a task for a century far away ;-)

Many people also seem to be under the mistaken impression that GPS provides sane routes for 100% of the driving you'd ever need to do.

Every summer I visit a shooting range at the end of a handmade [grass/dirt] road that has no mailing address.

Even if I could program that route into a self-driving car: it can be a very difficult road for a skilled driver to navigate in a truck. You have to know where the ruts are, what angle to have the wheels at [lest you get stuck], where the soft spots are [after a rain], etc.

Attempting it in a 4-door sedan loaded down w/ firearms is another matter entirely [unless you like picking grass out of your radiator.]

For some reason: I can't imagine engineers from California have given much thought to this particular use-case.

---

Google Maps can't even figure out my morning commute. Google would have me take a _non-existent road_ up to a drainage ditch. Then I'd need to scale a 60% grade up to the chainlink fence that surrounds _the rear_ of my building.

I'll reiterate: I drive a 4 door sedan w/ a 4 cylinder engine... even if I managed the hill I don't know what Google expects me to do about the fence.

---

I'm not saying driverless cars can't overcome these challenges. After all the TerraMax[1] was built by a Wisconsin company.

Not everyone wants to own a brutish truck when a bit of human ingenuity can get a Ford Ka through the muds[2] just fine. ;-)

[1]: http://en.wikipedia.org/wiki/TerraMax_(vehicle) [2]: https://www.youtube.com/watch?v=66yISwURzE4

Looks like it has no trunk, which is a huge use case for self-driving cars. How many trips do you make just to move things around? How much of what you own would you rather rent if you could have it delivered within 15 minutes? Then there's the economics of prepared meal delivery... I'll be surprised if passengerless cars don't outnumber passenger cars in 20 years.

As for moving people, a significant part of city bus fares goes to driver salaries. And other vehicles tend to avoid buses / don't rely on making eye contact with bus drivers. And buses follow predefined routes. So it's surprising to me that Google is going for private cars instead of city buses as a first market.

Looks like there is a bag compartment for had luggage where the dashboard normally is. The size of the thing does not suggest it's built for heavy lifting, I'm sure there will be models for that too.. maybe even the same with the passenger seats ripped out. When the car self-drives, you don't need the driver anymore to get stuff around. This really opens possibilities.

The first prototype might not be the be-all-end-all thing, but it makes me confident that they don't try to design that on the first go.

The video shows a rotating camera at the top of the car that's scanning for objects (I guess). What's the advantage of a mechanical rotary movement over an array of cameras that give a continuous 360 degree view (e.g. the street view cars)?

Update : Found a link : http://www.pcmag.com/article2/0,2817,2402516,00.asp In an interview last week, Stuart Woods, the executive vice president of Velodyne, which manufactures the LIDAR (Light Detection And Ranging) rotating module that sits atop Google's autonomous vehicles, said that his company has started developing prototype security systems that could be trialed within the next year. The Velodyne LIDAR HDL-64E and HDL-32E modules use an array of either 64 or 32 lasers to electronically "see" the environment, Woods said. On Google's car, the module is set inside a rotating drum. Its lasers complement Google's own mapping software and GPS data, which help orient the car on the road. The LIDAR provides additional positional data, but also identifies other cars, bicycles, pedestrians, and road hazards.

If I'm not mistaken, the "camera" at the top sends a very thin, concentrated vertical beam to detect objects. To implement it in the way you described, you'd need quite a few cameras (100+) to get the same precision.

Don't think of it as a camera. Think of it as a radar (that uses laser light instead of radio waves, hence the LI instead of RA in LIDAR).

I don't know why they chose it, but off the top of my head :

1) possibly less production cost. an individual laser module is likely more costly than the mechanism to rotate a single one.

2) fewer channels of data often means less processing overhead.

https://maps.google.com/locationhistory

right now google already knows where we have been; with the car they will know where we are going to;

The car as a tracking device, pure genius.

No Dave, you can't go to that protest; really not possible....

So far Google's self-driving cars have been designed to have the driver take over in certain situations. I wonder how that works without a steering wheel and pedals. Are the cars manually drivable at all?

Looks like they are.

"We’re planning to build about a hundred prototype vehicles, and later this summer, our safety drivers will start testing early versions of these vehicles that have manual controls."

That implies that the prototypes will have manual controls, but the production versions will not.

From a UI perspective if they rarely need manual control you can use a less ergonomic option. Such as a joystick which would be a pain for regular use but would take up less space and still provide plenty of control for situations like parking in a field for a concert and or a quick emergency override in case something goes wrong.

> control for situations like parking in a field for a concert

Or the car just takes a panorama and converts it to an aerial view and you simply drag your car where you want it to be and press Go.

Yes, that's is a very interesting point. I would think that, long term, the only manual control a self driving car really needs is a "kill switch." How long will it be after self-driving tech goes mainstream until people simply don't know who to drive?

I can see my kids now, "I remember when I was a kid, my mom/dad actually had to control the automobile. Archaic!"

Not too long ago my friend's son asked me why my phone was attached to the wall. He had only ever known cordless phones. I image cars will go a similar route.

There's a funny mention of this problem in "I, Robot". When will smith takes the wheel he is scolded for taking such a completely irresponsible risk. Given how those robots drive versus how he drives, we likely will have to agree with this at some point.

I wonder if real life will play out the same way : will controlling a car the human way become an unacceptable risk, and how fast will this happen ? Because we're likely to see both effects play here : first, humans will become worse at controlling vehicles, and vehicles will become better at it, likely having superior senses and predictive ability. So there should come a point where humans become a risk.

So there will come a point where we get to choose between new policies and letting humans drive. If we had self-driving cars and we could eliminate speed limits on interstates entirely, in trade for not allowing human drivers at all, would we do it ?

I would not be at all surprised to learn that computers are better at driving now. One interesting (albeit expensive) way to test this out would be to put some computer controlled cars into demolition derby races and see how they fare in artificially hostile driving environments where the human participants are extremely risk-tolerant.

I'm very glad that California has legislation in place to allow this already, and think it's going to pay off in spades to be a leader in this area. My bet is that insurance will be baked into the price of the first generation of cars (ie subsidized by the manufacturers), liability rates will be extremely low, and there will be an avalanche of consumer switching over a fairly short period of 10-20 years.

They're citing >>300,000 miles under automatic controls with zero accidents. That's probably better than the human average. (Small sample, I know, but still.)

It looks like Americans average about 200,000 miles between accidents. I am not sure what the variance here is, e.g., depending on driver experience.

Still, the numbers are completely incomparable. Google's cars have always had a driver at the wheel, and we have no idea how many times a manual intervention has prevented an accident. Google's cars have also been driving in very limited conditions: they have not tried to drive in very bad weather, and they do not handle highway merges or difficult surface streets.

Finally, it isn't clear that "zero accidents" is completely correct. Google has said that their cars have been in accidents, but only while under manual control or while stopped at stoplights. But what does this really mean? Perhaps there was an accident caused by the self-driving car, where the human driver took control at the last minute. Or perhaps the car saw a yellow light up ahead, slammed on the brakes to stop, and was rear-ended by another car expecting them to go through. (Technically, the other car would be at fault in this situation, but it is still unsafe driving on Google's part.) There is no public information that lets us know the safety record of Google's self-driving cars.

This wouldn't be a very interesting story if the answer were "yes." Whether Google or Toyota manufactures a self-driving car with manual override doesn't matter much from the point of view of a typical citizen or someone who's interested in the architectural implications of this technology.

As long as there's a manual override and a responsibility to use it, the car requires a capable driver paying attention all the time. It's an interesting beginning and it might improve safety but it doesn't fundamentally change things.

But without manual override, the car must be very reliably capable on its own. I suppose there might be a network connection and the ability to dispatch an Uber to pick you up whenever the car decides it is unable to proceed. But nobody will choose to use these cars if that scenario is remotely common. So announcing that this thing could ship in 2 years indicates to me that either Google's tech is doing way better than has been suggested in other recent press, or Google has started talking up vaporware to an unprecedented degree.

From the stories that go into detail about google's self-driving cars, it really seems like at this point the driver-take-over situation is really more like a breakpoint than an actual safety valve. It's there for when the computer has multiple reasonable options and the driver's judgement and awareness is needed to help resolve the ambiguity not just for that ride, but for future rides as well (as a sort of feedback).

When they drive entirely on Google property (like this parking lot), they don't need controls. If they want to drive on public roads, the law requires manual control.

In Amsterdam, I'm a big user of car2go.

These are little electric smart cars which you can rent by the minute. For times when I don't have my bike nearby, or the public transport is not handy.

These are basically little smart cars which drive themselves, which would be perfect.

I use car2go in Austin, and already love the system. A car2go that could drive itself to my door would make my commute perfect.

Will there come a time when you can just cross the road whenever you choose knowing that all traffic will stop, without fail, automatically?

I think the driverless car will actually make more of a difference to the typical pedestrian than the typical car occupier.

> I think the driverless car will actually make more of a difference to the typical pedestrian than the typical car occupier.

Depends on the country really: in germany, if a pedestrian gets close to the curb all drivers will pile on the brakes.

Gets awkward when you just wanted to take a slightly better look at something on the other side of the road (I've more than once crossed only to cross back a minute later, because a driver had thought I wanted to get to the other side)

I'm thinking about this basic idea: We have, in this case, a machine that is doing the job that a human driver would do - A machine replacing a man.

My question is: Should we be happy about this? Or should we be sad?

This time, we have a machine replacing a driver. In the future we may have machines replacing teachers, doctors, carpenters, artists... even lovers and machine makers.

So, my question remains: Should we be happy? Or should we be sad?

Normally, in science fiction movies, when the machines become intelectually smarter than humans, they try to destroy humanity.

But if, in the real world, they just become smarter than us, and just replace us in every activity. Then, what would we do?

What would we do? Well, we might aspire to be more like our newfound superiors. We might attempt to become cyborgs.

What we always do - find more time and capital for more interesting projects.

What did housewives do when they didn't have to stoke a fire to make dinner or out dirty clothes through a wringer?

The relentless increase of technology improves everyone's lives even as it replaces manual labour. It's a non-stop process that shouldn't be messd with.

I've been thinking about this a lot. I'll answer it two ways.

Negative:

- As we've automated jobs we've also created new ones, but lots of these new jobs are BS. Inspectors of inspectors etc. to fulfil health and safety regulations. We're also creating a lot of things that improve our lives or entertain us but are really just created so we have something to spend our money on. We don't need them.

Positive:

- I think the idea of a basic income would fix this. If we can automate everything and reduce the number of people that need to work AND provide them with a basic income, they can live their lives and not have to work. If you aren't anti-government you could even have the government owning factories, producing goods through automation and selling them to other countries in order to finance a good basic income for its citizens.

It's going to be interesting and if you are young enough likely something you will see played out in your lifetime. Although the changes in the 20th century were massive and world changing I think the changes this century are going to be ones which change how we actually live on lives on a very basic level. What we do everyday and how we get money. Even if it all goes to hell it'll be interesting!

> My question is: Should we be happy about this? Or should we be sad?

We should have a machine be sad about it for us!

Am I the only who actually enjoys driving (good cars)?

Finally someone! I thought I'm the only one here who loves driving.. Some people go as far as suggesting that manual cars will be banned at some point in the future - to be honest it sounds like some kind of dystopian vision where you have no control over anything.

Personally, I find it disappointing that this is perpetuating the inherent inefficiencies of the car concept, carrying a small number of people, usually a single person. The ultimate end game where cities are full of autonomous taxi vehicles that aren't owned, but paid for by usage makes sense though. You fire up an app / web link or SMS and a car appears within 5 minutes, drops off someone else and takes you on your way. Provided folk don't leave too many empty beer cans in the back seat...

> Personally, I find it disappointing that this is perpetuating the inherent inefficiencies of the car concept, carrying a small number of people, usually a single person.

"Self-driving" systems for carrying big numbers of people have existed for centuries now: http://en.wikipedia.org/wiki/Public_transport

Without manual controls, what mechanism do they use to drive it into and out of the trailer as seen at the beginning of the video?

It seems like they would eventually have a four-seater version but the bench seats could face each other.

What a simple thing but so different. Maybe like suicide-doors used to be common in the US but they're such a novelty now.

Yes, one of my first thoughts was about how much I enjoy road trips. But realistically, the bits I don't enjoy are having my attention pegged to the road and my foot to the pedals. A driverless car still leaves that progression through the landscape, the scenery, the togetherness.

Taking your family on a road trip where the travel was in a capsule with two bench seats facing each other, and a table between, would be great. Read books, play cards, get some work done. Convert seats to beds and get some sleep while you cruise at 40 kmh to your destination overnight.

Sounds like you are waiting for the VW Camper Van version to come out. However, if you could sleep and drive there could be a whole new class of commuters cluttering up the roads on 4-8 hour commutes, or a whole new breed of tourists that set the car to drive through the night to the next destination so they wake up somewhere different, outside some cafe for brekkie.

> or a whole new breed of tourists that set the car to drive through the night to the next destination so they wake up somewhere different, outside some cafe for brekkie.

Or, maybe there'll be a 'random' button; the car surprises you about where it intends on going.

>> "It seems like they would eventually have a four-seater version but the bench seats could face each other."

I can think of two reasons against that.

1. People who get motion sickness get it 100x worse if they are facing the wrong way. There are also people I know who normally don't get motion sickness but get a bit queasy facing the wrong way.

2. If you are in an accident isn't it more dangerous for your neck if you're facing the wrong way. I know several cars currently (mostly people carriers) that have rotating seats but they can't be rotated when driving. I'm guessing that law is in place because rick of injury is higher facing the worn way.

To the person that down voted me: why? I believe I made two good points that add to the discussion. Down voting should be used on comments that do not add to the discussion, not just comments you disagree with. If you disagree with a comment ignore it and move on or reply.

Side note: What has happened to edit buttons on HN? This is the second time I've had a comment I'd like add an update to and the edit button isn't available.

You have a limited time after you post in which you can edit (2 hours I think).

I've upvoted your comment as you are right it wasn't downvote worthy. I may have downvoted your complaint about the voting but the side note is a real question).

Thanks. Is this a recent change? It's only a problem I've noticed myself having in the last few weeks.

No not recent; although it is possible the window has shortened.

London cabs have (optional) rear facing seats already.

I'm going to be sorry the day they outlaw manually driving a car.

Can't wait till these work on country roads!

Would really free me up from living in the city I work. Can get work and other things done on the commute to work. Or live miles away and telecommute work most days then travel overnight whilst sleeping for times I need to be in the office.

Best yet, I can go on nights out without worrying about driving home / paying taxi.

The other maniacs and safety of country roads has always put me off commuting from rural areas, but this really changes that!

They won’t have a steering wheel, accelerator pedal, or brake pedal… because they don’t need them.

I like the self-driving prototype. But this comment brings to my mind a situation where police operators or organized crime override the car's controls and reroute it to an alternative destination. A manual override, perhaps no more than an emergency brake, would be really nice if you thought something weird was going on.

Technically, our current cab drivers could do the same thing.

One question I'm curious about is whether or not we will shift to having seats in a different configuration such as having people sit facing each other. Furthermore I'm curious what factors will contribute to one configuration over another.

factors that may matter: (1) safety

(2) situational interaction (meeting, road trip with family)

(3) social norms (what are we used to. e.g. the french sit next to each other in cafes, but americans sit across from each other)

(4) motion sickness

(5) ???

It will be good if we can see these cars tested on the streets here in India. India is an amazing test bed for this kind of an application. Narrow roads, haphazard traffic, undisciplined crowd and driver behavior, kids playing on the roads, obstructions, potholes, animals on the roads etc etc.

This will make an amazing place to test this set up.

Plus India is an amazing business opportunity, given the overall scale of the economy here.

How many drivers in India properly obey driving laws? My understanding is that it can be pretty hectic driving India streets and breaking the law is required to make forward progress. I'm guessing a self driving car as designed for the US roads may be useless in India.

That's true.

Which is why to make a generally intelligent self driving car. You will need a global test bed.

Yes, I think the car would just stop until there are no hazards or so that it can navigate around them. In India that might just mean that it won't move anywhere.

>They won’t have a steering wheel, accelerator pedal, or brake pedal… because they don’t need them. Our software and sensors do all the work

This sounds amazing.

I think Google could reach their goal faster if they started with more controlled environments - such as baggage vehicles at airports or mining trucks. There are plenty of people in my part of the world who are paid six figures to drive a truck up and down the same road of a mine all day.

I know there are open-cast mines that already use remotely controlled/monitored and self-driving trucks. I think using it in an environment that they're used to, and which I presume would be most practical, makes sense. It is more helpful for them to research and build their technology where it will mostly be used, on the streets. There is more volume of traffic, which makes it possible to deploy 100 small cars cost-wise, and could lead to 'quicker' development as they won't always be slowed down by legal and other negotiations with airport companies and the like.

Just my thoughts anyways :)

The biggest problem I see here is that if all cars on the road were self driving then you have more control. The issue is when you mix human behaviour and error into the mix. The first fatal or near fatal accident with self driving cars could be the end before it's even started.

Anyone else wonder why a self-driving car needs side-view mirrors?

to get out safely avoiding incoming traffic once parked.

Best part of the embedded video, and a moment sure to bring a smile to the engineer's faces is at 1:45. No amount of data can give you that sort of useful feedback.

I think many people against Google self-driving vehicle are worried about privacy issue. Well I say please bring out the self-driving cars asap.

IMHO, big part of the first wave of customers to buy self-driving cars will be older drivers who are not allowed (or choose to) drive on their own. My parents are able to drive on their around now but are already getting worried about the day when they can't drive around on their own. They would appreciate a self-driving car.

Only issue for google would be that they won't be providing any useful data to google with their destinations and they are not really online but oh well...

They don't really add any extra privacy issues than carrying around an android device in your pocket. Also even if you dislike Google, they are fairly transparent and would likely tell us if they are tracking the data, how do you know GM/Ford aren't already tracking vehicle routes and using this data?

I can't believe I didn't realize this earlier: When you have nothing in front of you, there's no need for an airbag!

Hmmm, will Google go into manufacturing their own cars or will they stick with software and license it to other manufacturers?

I think they'll do what Tesla does. They'll do a few of their own but their long game will be selling the tech to big manufacturers.

White one looks better. Overall design not too bad, especially if they want them to be immediately recognizable.

This warmed my heart. And thinking about it from a business perspective, not everybody needs to own their own self-driving car. Ie. it can self-drive itself home for use by the next person. Which means economically, the car price can be many multiples of the cost of current cars (imagine a $100k car that is shared between five families rather than five families each with their own car).

I find it cute how most of the people still are not comfortable taking their eyes off the road.

To be fair I'm not comfortable taking my eyes off the road as a passenger with a human driver. I think people underestimate the fear of not having control. This will be a major issue to overcome and for some it will be like the fear of flying on a commercial aircraft. The fact they know it's safe doesn't really help much.

It's so cute! Look at it! It's even got a happy face! It's totally not a life-threatening robot of doom!

And it promotes family talking time! Just like fast food restaurants in recent episodes of Mad Men!

Sorry. It really is a slick ad for a neat product, I just can't resist noticing how they position it.

I wonder if they'll ever get the vehicle below $200k with the 64-beam 360-degree FOV LIDAR.

Please hire me!

Forgive my crankiness, but: Does this say anything that the casual follower of self driving cars doesn't already know? All I got out of this was:

- Self driving cars would be awesome, for well known reasons.

- We have to use a lot of good sensors to make this work.

- It's an interesting problem.

It says that Google are building the cars.

I'd argue that the entire propose is to expand beyond 'casual followers.' The challenge of self driving cars is ultimately a social (more than a technical) one. Distribution. PR. Legal.

Takeaways:

- They're working on physical prototypes.

- These cars don't have a steering wheel or pedals for human override.

- Speed-limited reveals a bit more of their strategy.

- Makes clear how useful and empowering these vehicles will be to someone visually or physically impaired, the elderly, etc.

Reply on news.ycombinator.com