I'm interested to see where all the major players end up in 5yrs:
• Tesla bootstrapping a ride service on the backs of buyers
• Waymo directly rolling their own fleet
• Uber trying to get a self-driving fleet up, burning mountains of money to maintain their "monopoly" on Uber for rides
• Lyft working with GM to get a fleet up
• All of the other car manufacturers trying to get autonomous vehicles going, presumably hoping for consumers to still want to own a vehicle rather than just pay $1 to get a ride
So: How much is Uber's market share worth? I suspect it'll evaporate overnight in every market where another service has autonomous vehicles and they don't.
Also: Private car ownership is going to fall off a cliff shortly after autonomous ride services arrive. Which probably means general demand for vehicles will fall off a cliff.
> Also: Private car ownership is going to fall off a cliff shortly after autonomous ride services arrive. Which probably means general demand for vehicles will fall off a cliff.
I've seen this sentiment so many times now that I have to ask a serious question to anyone who echoes this sentiment:
Do you own a personal car right now? If so, why?
All of this... excitement over the transformations that self-driving vehicles will bring to the world of transit basically boils down to a model that's a mixture between taxi service and (really crappy) bus service, both of which exist today and have existed for several decades--and basically predate the automobile when you think about it. Yet somehow self-driving vehicles are supposed to completely change the mix of transit use when it's not really adding any capability changes. So I'm genuinely curious as to what the magic sauce is that makes people so giddy for this future.
I can't add anything profound, but I will say that 40+ years ago I camped at Disney's Ft. Wilderness campground, with all my many brothers and sisters. The electric trams had a routine schedule that came around and stopped almost in front of your campsite, all of kids would simply jump on, go to anywhere in the park like the stores, the showers, the outdoor theater, and hop off again. It was incredibly liberating and fun, our parents didn't have to be involved, we could go where and when we wanted. It was fun. I know, childishly silly, right? Just a way of getting people around a big place? Also called a Bus? No, not really as buses don't go where all the people live. Transportation was a critical part of Disney's attempt at showing how his futuristic communities would work. Now whether those specific trams were directly part of that, I dunno. But it did make your community activities, events, places, much larger and much more accessible, efficiently, and no worries on parking, maintenance. I have never forgotten that feeling of ease and efficiency of movement and freedom, and I saw that promise echoed in the Waymo video.
> No, not really as buses don't go where all the people live.
They do, though, you just haven't been to places where this is standard. Try finding a place that's more than 10 minutes walk from public transit: https://www.google.com/maps/place/Warsaw,+Poland/
No, I have been to many places in Europe and was amazed at the places they did go; schoolkids into the Alps for their skiing lessons, for example. I meant here in the US where bus service is mostly an urban experience, as well as underground/trains.
Yes. Having moved from a country that lacks good public transport to one that does some years ago I had exactly the same experience. People tend to judge from experience (and why would you not) and I'm guessing many Americans simply have not experienced the freedom that you get by being car-free.
Yes, but Disney never solved the problem of how expensive it is to run that service -- they just got people to pay. Will self-driving cars be cheap enough? They save money on the driver, but cost money in the automation hardware and security against vandalism/theft.
Most of the cost of a taxi service is in wages to the human driver. With an autonomous car, you don't have the driver, so the service can be cheaper (especially with electric cars). I lived for about a year without a car, using Uber instead. But when I tallied my Uber usage, it ended up being about $600 per month. So I got a car instead and saved a lot of money. If this could be reduced to $200 - $300 per month, it would have been a different matter entirely.
I think you're basically agreeing with the parent. Like Zipcar and Uber, autonomous vehicle technology when it's eventually available will almost certainly be one more incremental step toward enabling reduced personal car ownership in edge cases.
Except that they are not really cheap, safe, or convenient. They are not door to door. Public bus/trains run at their own scheduled time, not at my desired time. Even services like Uber involve humans who are more concerned about their profit than my schedule or preferences, so they cancel rides when I need them the most because some other ride is more profitable for them. Moreover in countries like India where road accidents are the highest in the world, rash and unsafe driving is a big problem (government run bus, car drivers or private ones). This can just wipe that out! So personally, I'm super excited about this future.
I think that people who believe that private car ownership will fall are those who live in the urban centres and have little understanding of the needs and living styles of rural or suburban population. It could just as well be that the opposite will happen, because cars will suddenly become much more accessible for people who can't or don't know how to drive.
Or even those who live in urban centers but like to get out of them on weekends and vacations. I've tried living for years in western US cities both with and without a car. Going without was much cheaper, and I managed to find entertainment and not starve, but getting a car vastly improved my quality of life. Trains and an occasional rental would probably have worked nearly as well in parts of Europe, but the US just has too much empty space between interesting places.
> It could just as well be that the opposite will happen, because cars will suddenly become much more accessible for people who can't or don't know how to drive.
This is important to note. I'm on those margins, and have been saving for my first automatic car for ages. I live in a mildly rural community, and don't wanna wait for uber or whatever to deploy autonomous cars here.
I think the point was that the "urban area" definition in the census was quite broad. I live in an "urban area" on a decent plot of land with a Christmas tree farm on one side and a working farm/apple orchard on the other. Probably not what most people think of as urban.
My parent's house has an orange orchard in front of it, a wilderness area behind it, but Uber rides from downtown are still only $11, and everyone I know who lives in that town now uses Uber instead of driving. DUIs are way down. Nightlife is better, etc. etc. This is a very real 'urban' situation.
It is relatively broad, but the point is that it's based on population size and density. The fact that you live next to a farm and an orchard doesn't change the fact that you live near lots of other people.
I suppose it depends on your definition of near. I'm about an hour drive from a major city. Less so to smaller urban areas that have no significant public transit. No it's not the wilds of Wyoming but most people here probably wouldn't consider being on a 100 acres or so with a couple of other houses being near lots of other people.
From a personal perspective, if you could easily pre-schedule a self-driving car, and the reliability was very good (it's basically there when you need it), it would increase the convenience factor by a lot. From my perspective I'd be encouraged to reduce the cars in my household from 2 to 1 if this was available -- I would probably use this type of service for any sort of situation where I have a pretty good idea of when and where I'll use the car.
Taxi service exists, but the reliability factor is very poor right now (in that you call and you have to wait n minutes, where n sometimes is a very large number). I see Uber is rolling out a pre-scheduled service and I wonder how good this has been so far for the markets that have it. This would basically be the same thing only with human drivers.
I don't think this pattern works for rural areas and all situations. I also envision some nostalgia and/or tech discomfort keeping the car ownership pattern going for a while. Still, I am wondering if there are a lot of 2+ car families that would be in the same boat. Even drastically reducing the 2+ car families will drop vehicle demand.
With a huge fleet of self-driving cars, there would be enough capacity to have permanently configured cars for child seats, bike racks, dog areas, etc.
eg. "I need a 4wd vehicle for 2 adults and 2 bikes to take me to X park at 10am tomorrow."
Not without seriously increasing the costs for car rental.
The cost savings for car rental are predicated on increased utilization. Increased utilization is predicated on each car having more people using it on any given day. And these things don't scale infinitely -- each car can only serve a particular locality.
The number of variations that you propose is a fast geometric progression. Kids 0-12 months or so need different car seats from kids 12mo to about 3y need different car seats for older kids. The idea that you could support high utilization on "I want a car that has already installed a carseat for a 1-2year old + a carseat for a 5 year old plus a bike rack" is nuts.
(Regardless, also, parents carry a mountain of stuff that's THEIRS. The heck with a car seat, I need board books, toys, snacks, and wipes in the car.)
Er, I wasn't talking about pricing complexity, I was talking about the ability to actually get the vehicle that meets your needs. It doesn't do you any good to know that "a vehicle with three or more constraints costs $1.50/mile" if you get a vehicle whose three constraints are "4WD, Bike rack, Dog area" when what you wanted was "carseat1, carseat2, large storage area."
And if the fleet can successfully meet arbitrary needs around constraints, then the fleet is going to be heavily overprovisioned, and thus expensive.
But the fleet is heavily overprovisioned now. We're just used to absorbing the costs of the overprovisioning. But the vast majority of Uber rides (or of passenger vehicle travel generally, really) don't require five seats, don't require trunk space, etc. We just equip cars standard with these configurations for the small fraction of trips that actually require them. In some sense self-driving cars should actually reduce overprovisioning by letting single-occupant trips use smaller vehicles (and hey, in a pinch, cars with bike racks on them work just as well carrying people without bikes, just like five-person cars can carry a single person alright).
Car rental companies already manage this quite handily. Roughly speaking: economy, midsize, full size, SUV, 4WD SUV.
"4WD, Bike rack, Dog area" -> 4WD SUV.
You'll probably have to snap on your own accessories (kayak/ski/bike rack) when the vehicle arrives. Or if you order in advance, you can pay extra to have a specific accessory configuration prepared and set up for you (vehicle drives itself to local servicing point to get configured).
I've wondered whether some of this would lead to a personal "pod" like thing. I am happy buying a box with comfy seats and a TV and kids seats, etc.
But I don't need or want to buy a wheel, batteries, cameras, lasers, etc. Those also should be upgraded as technology gets better but my box doesn't really need to be anything but a box.
So what if I could sit in my box and have a transporter thing come over and clip onto me and take me away? You could have child seats and I could have a lazyboy chair, or I might need 8 seats and you need storage space.
This would then let the actual transporters take other types of cargo too, rather than having a people vs items split. It would allow a pony-express style swapover rather than needing huge range in a single one. It would even allow completely different types of transport, what if my pod could just be clipped onto a big train and whip me to the other side of europe?
This would all likely have other downsides, but is a fun thing to think about.
How about making your existing car self driving? Your stuff is always there in the trunk, but when you're not using it, it joins the fleet and earns you some money.
I really want to see a day, where I drive my car to work. It then drives other people around and picks me when I am finished.
1. is never happening. If you want a proper safe car seat for a 1-3 year old, it's going to weigh 30-40 lbs no matter what sintered pixie dust you build it from. And it has to be pretty huge since a small person has to fit almost inside it.
Even the ones for babies, that people for some reason I can't fathom like to carry around in the supermarket, are actually also 20+ lbs, you've just left 15 lbs remaining in the car occupying a seat while you're shopping. And they're also bulky.
2. could happen for children aged 4+ in a minivan-sized car. I've already seen it on some long distance buses. For younger kids however, they really need to sit facing rearward, but also must be well protect against side impacts, so it's very tricky to make something retractable that doesn't take a huge amount of space.
> Even the ones for babies, that people for some reason I can't fathom like to carry around in the supermarket, [...]
These are awesome because you can take your sleeping baby out of the car, carry them with you without waking them up, and put them back in the car without waking them up. Shopping is 10x faster this way than when you have to spend most of your time comforting/distracting an awake and fussy baby.
Maybe I'm spoiled with sleepy kids, but we would just go get one of the shopping carts with a baby seat on, lift baby from car seat to shopping cart seat (95% success rate for not waking baby), go shopping while actually having space for stuff in the cart.
> I've already seen it on some long distance buses
Can we apply some if this science for making car seats for taller people. I am so unhappy when seat ends midway of my thigh or shoulders ram into headrest (mostly airplane problem).
You are over estimating car seat weights. A large, one piece, high end seat for kids up to 70 pounds like a Britax Marathon only weighs 19.5 pounds. No pixie dust needed, just plastic and styrofoam.
The one you link to is a) not ISOFIX, thus not high end, and b) discontinued from the manufacturer.
We have this one, which is what I'd consider the gold standard. It's ISOFIX, and only rearward-facing since kids under 4 should not be facing front, and it's won basically every test there is:
It weighs 40.6 lbs. Quite a bit of structural steel in there, including a full integral roll bar and a brace against being rear-ended, both of which your linked Britax seat lacks.
Also bloody expensive, but if there's one thing I won't skimp on, it's the safety of my children.
ISOFIX doesn't define high end it is just the name of the European standard. The US decided to go with a different standard called LATCH. Britax is high end, and the Marathon is their largest safest seat, and it weighs 19.5 pounds.
So maybe in Europe you are going to have a harder time with self driving cars because you need heavier car seats, but you'll probably find a workable compromise. Yu can just build your base into the center seat or something.
Like I said, assuming your friends have a proper high-safety Isofix seat, those removable car seats are only 1/3 of the whole seat, 2/3 are left permanently in the car.
I think this claim is making the assumption that self driving cars will be much more accessible and available than taxis/Ubers today. Where I live (in the suburbs) I can't rely on an Uber being available at any time I would want to go somewhere, and taxis only exist at the airport. If I could count on a car showing up at any time, anywhere I might go after a short but predictable wait time, I would definitely ditch my car. But the current options available to me don't allow for that. I don't want to own a car - I feel like I have to.
Also, cost is another big factor. I would also assume that taking a self driving car would be a smaller expense than owning for the average user.
Ridesharing currently requires human drivers which limits them to where the drivers want to be and limits them to how many drivers are in your area.
When human drivers are not required you have neither of these problems. It will be significantly easier for ridesharing companies to provide availability in non-dense areas.
Hell, they'll be able to control where these vehicles go by algorithm, so it will simply be a matter of adding more vehicles to their fleet and the algorithm will take care of minimizing response time for everyone.
It takes time to sign up human drivers and onboard them. Buying new cars can be done in bulk at scale.
Also, you can easily send a car from one city to another. A human driver needs to get home to their family. You can't just say "drive to Miami and work there for three weeks".
Also, simply decreasing the cost per mile by cutting labor costs makes further flung areas more accessible. You can't sell a $10 ride if it costs $10 just to pay someone to drive to the area and back. But if it costs $4 to send a car, that ride becomes financially viable.
Self-driving taxis are in principle way cheaper to operate than human-driven taxis. They never have to stop, they can operate at all hours, there's no cost to pay drivers, and if they do stop they can do it somewhere very cheap.
And when self-driving cars are a viable option, insurance will strongly encourage it. And by strongly encourage it I probably mean "absurdly high fees for driving your own car", because people are much worse drivers. Which should drive up the price of car ownership significantly. The increased cost to car storage in populated areas will probably lag a long while behind, so the second cliff will be much slower to hit.
This is a frequent argument I keep seeing about the cost of insurance rising as autonomous vehicles are introduced. There's no reason they'll rise past what they're currently charging; insurance takes into account current risk. In the future, the risk if anything should decrease.
The risk may decrease but liability may increase, as this is a risk you are taking deliberately if you decide to drive your car instead of getting a safe self driving car like everybody else.
Insurance doesn't work that way. Liability isn't an ephemeral concept, it's the chance of a claim X the amount of a claim, pooled in a group of buyers with differing risks. The relative cost of insurance compared to an autonomous vehicle should increase, but not the dollar amount.
I can buy cars with better safety systems being cheaper to insure, whether you're talking assistive driving systems or something more. But why would the existence of cars with better systems make driving vehicles like most people own today more expensive than currently?
> But why would the existence of cars with better systems make driving vehicles like most people own today more expensive than currently?
Safer options will, as they become available, be disproportionately chosen by the safety conscious. This can, potentially, make the less-safe legacy option more dangerous (per use) than before the safer option became available, because the safest users of the less-safe option before the new one are users of the new option afterward, and no longer bringing the average risk of the old option down.
Conversely, if a significant proportion of the cars on the road are autonomous and safer than human-driven cars, then the accident rate for everyone could decline, making insurance cheaper for both autonomous and human-driven cars.
Wouldn't that make insurance cheaper for autonomous cars and more expensive for human-driven cars since humans would be more of a liability behind the wheel?
The comparison isn't between cars with assistive/autonomous systems and those without. It's between cars without those systems some decades hence and those without such systems today. Why should the existence of such systems make driving by humans more expensive than it is today?
I own a car in India, because there are no other means to get to where I want to, when I want to. I now live in Singapore and I absolutely don't feel the need for a car. That said, I am not incredibly happy with the public transportation here. It's still slow - takes about 35-40 mins to cover 10 km door to door. It's not possible to haul kids and luggage in public transit.
Self driving cars address these shortcomings. They are door-to-door. I rent a seat when I am travelling alone, or rent the entire car when I"m traveling with family.
I'm dead certain that traditional car companies are screwed. Just look at the driving license stats in the US. The number of millenials getting a driving license is dramatically decreasing.
> Yet somehow self-driving vehicles are supposed to completely change the mix of transit use when it's not really adding any capability changes.
Self-driving vehicles could be at your door within a 1 minute timeframe (say). That's a huge change, and will nullify the benefits of owning a car.
Of course, this will be possible because driverless taxi companies will have the logistics in place to make this happen (they know where every car is, and they can compute where to send empty cars to increase the capacity).
I can imagine that a cheaper service will offer the same but, say, in 5 minutes.
Also, you will never have to find a parking spot, which may be an even better argument.
Except I like owning my car. I can take it where I want. When I need it, I have it. I can leave my stuff in it if I want to (e.g. if I decide to go to the park for the day I can leave my coat in the car if it is warm). If it's pouring rain I don't have to wait, I can just get in it and go.
I can't wait to have a self driving car it would be even more convenient than a regular car. I gues some people just don't see the utility in cars other than getting from point A to point B. I also love to own a car for same reasons as you do and also for a bunch other personal preferences which are not possible when i have to take a ride in somebody else's car. Just as i prefer sleeping in my own bed, watching my TV in my own living room on my own couch, using my own toilet rather than using a public toilet, i prefer driving/being driven in my own car where i have my things available.
Taxis suck and busses suck. Self driving cars have the potential to give you the benefits of the taxis and busses without as many downsides. Not sure what more you'd need to understand why folks are excited about this.
Not having to own, insure, maintain, or park a car. End to end transportation with no transfers that allows you to internet while you travel. Cheaper than human driven cabs (presumably).
Different usecases drive different needs. If you live in the suburbs, it's easier to own a car than to get an Uber and wait for 20 minutes every time you are going out. If you live in the city, it's cheaper to set aside money for Uber than to pay for a car and a parking spot.
Also, getting a driver has never been this easy. Taxis have existed for hundreds of years, but smartphones are a recent invention. There's a huge difference between calling a phone number and trying to setup a rendezvous than the current workflow on Uber or Lyft, that literally takes 10 seconds from the moment I unlock my phone.
Before Uber, the taxi company I used had an automated system which send a taxi to my place if I pressed 1, and connected me to human that took a new address when I pressed 2. Even with a human in the loop, it rarely took more than 40 seconds. I agree that Uber app is more convenient than that, but it really wasn't an order of magnitude improvement -- what was a deal maker for me was the fact that Uber automatically charged me, and I didn't have to put a card in and input PIN, or scramble for cash.
I own one now because there is no decent public transport in Dallas. I would sell it in a heartbeat if we have better public transport or if Ride Sharing services (with human drivers or autonomous) get cheaper.
> So I'm genuinely curious as to what the magic sauce is that makes people so giddy for this future.
I think the magic sauce is utilization. A fully autonomous car's utilization rate can be close to 100%.
Higher utilization --> same demand for mobility hours/miles can be satisfied with fewer vehicles --> ceteris paribus, fewer vehicles sold every year.
Viewed through this lens, it seems inevitable that, in equilibrium (ie once the non-autonomous to autonomous fleet transition is complete) annual per-capita demand for vehicles will be lower with the advent of fully autonomous cars.
I have to agree with you skepticism. Let's e.g. replace "car" with "apartment". Imagine an apartment where you pay a premium to have all the cooking, cleaning and maintenance done for you. This obviously exists today, yet >99% of people don't use it.
I believe people like owning things, and using them to do stuff. This is a basic fact of being human. Probably goes right back to the stone age. I doubt autonomous cars will change that pattern.
> I believe people like owning things... This is a basic fact of being human
There are some objects we enjoy forming a relationship with. Your favorite kitchen knife. Your bed. Your sports car you take out on the weekend.
There are other objects we just want to work. Your running shoes. Your daily commuting vehicle. Your dish rack.
Of course it varies from person to person which items fall in which category.
There will always be a market for people who want to own and fetishize a car. The same way there is a market for people who want to own and fetishize their record collection.
But it is usually a smaller market than the market for people who just want convenient access to the thing. I.e. Spotify.
Which will be bigger, Spotify for mobility, or your neighborhood hipster automobile dealer?
The connection to ownership works for things you operate. If I do my own laundry I want a nice washing machine, and I'll shop for one, upgrade on features, etc. Same for kitchen knives.
I have a maid who does my laundry so I don't care about my washing machine anymore. I still cook so I want good knives, but if my maid were to start cooking I'd stop caring.
I don't own a car, but sometimes I rent one. And when I do I almost always pay 2-3x for a sports car because it's fun.
When I use Uber, I never pay for the upgrade to a nicer car. I'm just going to read or fiddle with my phone, so I don't care about the car at all.
I predict the exact same behavior for autonomous cars. And why would I pay $25k for a car when I can get the same thing for $1 a ride (and not have to worry about parking or maintenance or upgrades to new models).
> Do you own a personal car right now? If so, why?
Yes.
There isn't a reliable taxi service or public transportation in most suburbs. Even if you have Uber in your town, it's much more expensive than owning/leasing a personal car.
I can't use Uber when I need to go to another town 50 miles away or 500 miles away.
Honestly, I would love to have my own self-driving car that integrates with home robotics.
Imagine you leave your laptop at home... Just get your car to drive home, receive it robotically, and bring it to your location before your next meeting!
It would work just like Uber (even better), only you would own it. Every delivery company would be out of a ride the moment your car can go make restaurant pickups.
> Private car ownership is going to fall off a cliff shortly after autonomous ride services arrive
But how is this going to work? In cities, everyone uses their car at exactly the same time: roughly, between 8-9 in the morning and 5-7 in the evening. If you don't own a car, chances are you won't get one when you need it to get to work on time.
Or, if there are enough cars for peak hours, then demand for new cars will not fall that much (although the structure of ownership may in fact change).
Other things that could change is the need for infrastructure (at least in Europe); you don't need highways and high speed trains that much if speed of travel matters much less. If your car drives itself during the night and you can sleep comfortably in it (or read, or watch movies, whatever), then the length of the trip is not paramount.
> Or, if there are enough cars for peak hours, then demand for new cars will not fall that much (although the structure of ownership may in fact change).
Initially, no, but long-term you can move much parking to cheaper places around the periphery of the urban area rather than needing a parking spot where people live, one where they work, and fractions everywhere else they might go. So sustainable density of everything that isn't parking spaces in the city core goes up, walkability and transit access improve, so the cars/people ratio drops.
In the shorter term, there are still huge efficiencies to be gained with smarter ride sharing. Imagine if something like uberpool served a majority of car commuters instead of a tiny fraction... you could get a lot of cars off the road but still provide a very high level of service.
Or, wait, what if we made the car really long and put in even more passengers per driver? It might even have a pre-determined schedule and be pretty cheap?
No, but busses can be pretty close, and shared vans actually will. I suspect those are closer than trains to what the grandparent post was referencing.
So all the cars also drive back out of the city around 9am and back in again around 3pm? That sounds like more cars on the road (over the course of a day) rather than less.
Upthread I see arguments for why there won't be many fewer cars, because everyone still needs to get to/from work and they do that at roughly the same time. Ride-sharing is what helps. Parking outside of the city has nothing to with demand for cars.
Peak pricing can spread out demand a bit, and also, people can use much smaller vehicles most of the time. We buy cars for worst case size or conditions. Look around the parking lot of a white collar work place that is filled up with four wheel drive all terrain vehicles and pickup trucks.
Ride sharing for rush hour would use a ton of compact vehicles. Even if the number of vehicles is the same to account for peak usage, the majority of the vehicles will be much smaller and much less capable, and therefore much less profitable for manufacturers.
The idea that everyone needs to arrive at and leave the office at the very same minute has always been idiotic. It's a holdover from assembly-line thinking. Presumably in a predominantly ride-sharing, autonomous-car world this would be reflected through pricing rather than through ridiculous commuter traffic jams.
Often you need a car to get to the subway which does not take that long. Slug lanes also become more useful with self driving cars.
Combine the two and you get a self driving car to central location > car or public transit to city center > self driving car in the city. Remember parking is often 10+$ a day due to land value, so two ~4$ trips can be a net savings.
Finally, even a 20-30% drop in demand would be huge in such a capital intense industry.
When I lived in downtown San Diego, I felt sorry for the car people. Some streets are one way only and some parking is two hours only, etc. If you aren't extremely familiar, it can be a nightmare to try to figure out where to park and then find your way back to the building you are going to.
And then one day I realized some of the buildings were multi story parking structures and it kind of made me mad. So much of the city infrastructure is whored out to the needs of arrogant, obtuse car people, very much at the expense of everyone. It brings quality of life down and cost of living up for absolutely everyone because of people wanting the right to drive their own damn car everywhere, no exceptions.
And I stopped looking on them with pity or sympathy and started looking on them as parasites who totally deserved whatever small misery they were inflicting on themselves with driving around, looking for parking. But I did not deserve the misery they were inflicting on me with things like worse air pollution.
I could see apartment complexes and neighborhood associations partnering with these autonomous vehicle services to provide transportation for their residents, then eventually municipalities demanding it of new developments, with a reduction of parking space requirements in combination (similar to what Austin, TX does with ZipCar and businesses: if you have 2 Zip car spaces, you can reduce your parking requirement by 4).
It's much easier to coordinate ride sharing when there's an interested third party in place. ad-hoc ride-sharing for a reduces cost could result in the average car traveling at much closer to full capacity much of the time. If you set it up ahead of time, it could even optimize a route that picks up and drops off along an optimal route.
Don't think of it as replacing what historically was walking outside and hailing a cab, think of it as setting up a carpool with office mates except you don't have to know them, and you don't have to work out all the details, it will be handled for you.
I think you overestimate how willing someone is to sit in a small car with 3 other strangers during their morning and evening commutes. I don't see a huge impact on ownership at least in the first decade or two of autonomous cars. Lots of people still have relatively specific needs, e.g. requiring car seats of various sizes for their young children, needing to pick-up or drop-off kids at school, sports practice, etc.
Mass transit works well in dense areas because the vehicles are large and only need to go short distances or go long distance over a specific route. When everyone has their own custom route, person A's route may make person B's route unacceptably longer. Also, sitting on a bus/train with 50 strangers is less of an odd social situation than sitting in a car with one stranger.
Once autonomous cars get to the point where they don't require any driver attention at all, I think it will make ownership even more appealing since you could customize the entire to match whatever you want to do while you're hauled around.
This is an N of one but, I use either Uber or Lyft to and from work every day. People get used to it. I find it far more pleasant than a crowded and dirty bus or train. Even when the car is small and full, the car usually smells better, is cleaner and oddly feels like there are more defined boundaries than a bus or train seat when crowded.
Everyone has their own door as well so there isn't as much jostling past and touching and being touched by strangers even in a full car.
The only time I've ever taken an Uber or Lyft and I not have my own door and/or I've had to scoot past people is in large party situation and those other people were my friends.
I've talked to women and they seem to prefer an Uber or Lyft over taking the train and/or the bus. Also, being dropped off at your destination rather than walking a few blocks from a metro stop to where you are going also makes a massive difference.
The benefits really outweigh the drawbacks. Especially if this is cheaper than owning one's own car. In theory, a private automated car service (calling a car and being the only rider) would still be cheaper than owning a private car and still have the same conveniences. It's just that eventually it will basically just be 4x cheaper to share a ride.
If everyone took a Lyft or Uber or Grab to work, then the congestion would be unbearable. San Francisco's BART takes about 400,000 would-be passengers off the road daily.
Mass transit remains the most viable method of transporting people over dense, multi-mile terrain.
The last mile and "normal" metros like Memphis, Baltimore, or Cleveland, where it is not dense enough to justify a comprehensive subway/light rail system, are probably the best use cases for self-driving cars and universal car sharing systems.
There is at least one lane on either side of the road at full capacity. Always bumper to bumper. But, the cars do not move save 2% of the time when their operators remove them from the traffic jam, I mean parking spot. There are 275,450 streetside parking spots in San Francisco city limits. Assuming a pooling of some drivers, and an increased utility of roads due to size & traffic flow then we can look forward to a net decrease of cars on roads and a net increase of available land for expanded sidewalks and bike paths.
I actually wonder about this. Is it a matter of most cars being mostly empty, or is it inherently impossible to match the mass transit capacity. If we look at it as passengers carried over space required and time spent, here's how I see it:
Rail mass transit does not as efficiently use the land it's on a occupancy basis (there's not always a train on a specific square foot of track). In peak times, cars are more efficient on a vehicle basis. According to BART system facts[1], there are 107 miles of track. There are 669 cars, seating for 72 in 448 of them with each being 70 feet (with 59 of them having an additional 5 feet for a cab), and seating for 64 in 230 cars[2] (with an indeterminate car length, so I'll use the smaller listed), for a total capacity of 46,976 seated people. BART states that all cars can hold over 200 people in a "crush" load, so we'll assume 200 as the theoretical maximum, and say BART can carry 133,000 people when at peak (crush) capacity, and over the 107 miles of track, that gives us a density of 1,250 people per linear track mile, but with only 8.3% track utilization at any one time.
Cars do not as efficiently pack people per vehicle usually, but can more efficiently use the roads on a per-vehicle basis. Assuming very heavy traffic which is not stop-and-go, so perhaps 35 miles an hour average (the same as BART), and a 4-lane highway (two each direction), if each car is allowing two car lengths between itself and the car in front (slow traffic), we have approximately 33% road utilization (or 25%, or 20% depending on what you think the average space between vehicles is). Since carpooling seems to be at about 10% currently carpooling[4] (ignoring that it may be different in certain arterial routes, as we are discussing), we have around 1.066 people per car[5] as a lower bound. With an average car length of 177.2 inches[6], or 14.77 feet, we can estimate the people per mile on the highway during this time as being 604 people per mile of 4-lane highway.
Interesting take-aways for me:
While 4-lane highways may take more room than rail (not sure the actual sizes here), they are also more versatile.
If the highway bogs down below 35 mph, it's then less than the average rate of BART, and we need to start computing people over time instead of just people over distance.
BART has much more room to increase track utilization, but there is likely unaccounted for overhead here on each train. Optimal usage at current speed is one train arriving immediately after the prior one leaving, at 35 mph exact speed and 20 second stops, for a train of six cars (?) and 425 feet, that would be cars 2.42 train lengths apart, and a utilization of 29%, or roughly a 4x increase over current rates, 5,000 people per track-mile.
Cars have much more room to increase vehicle utilization. If we replaced 50% of vehicles with full size vans transporting on average 7 people each and didn't touch road utilization, we would be at an overall average of 4 people per car, and 2600 people per highway mile. Interestingly, if we somehow moved towards a system where smaller vehicles picked up and shuttled people with small amounts of sharing to bus-stations where they were sorted into smaller buses (40 people) going specific area depots, and from those depots dispersed again to final destinations using individual cars with small amounts of sharing, we might easily surpass rail transit systems. averaging 20.5 people per vehicle, but with somewhat more area used should put us close to 10,000 people per highway mile.
Of course, there's a lot of assumptions in all the numbers, and some speculation in the possibilities, but I thought it was interesting to figure out. ;)
5: 105,476 drove alone, 13,917 carpooled, if we assume all carpooling was just two people per car, we get person to car density by (105,476+13,917/2)/105,476 = 1.066 people per car
7: 5280 feet/mile / 14.77 feet/car * 0.33 highway mile utilization = 117 cars at highway lane mile utilization. 117 cars * 1.066 people/car = 126 people per mile of highway lane. 4 lanes fives us 604 people per highway.
Of course, you could have a separate pull-out track for train stations, and interleave the trains. Then you would be able to double your utilization :).
I can't really think of any reason why one technology is inherently higher-capacity than the other, it's mostly that private cars are very inefficient, not rubber tyres.
Sure. I just wanted to run the numbers to see how it actually broke down currently, and what we could theoretically aim for. Light rail systems are useful in that they are efficient, and can readily move large amounts of people currently. Unfortunately their delivery characteristics are very rigid, and they have a very large up-front cost.
An autonomous car system that supports easy ride sharing, and semi-public transportation through something like uber for busses can fulfill much of the same need, but without quite the expected density, but with the ability to deliver to just about any location.
Both would benefit from a last-mile service . Sort of like arteries and capillaries in the circulatory system.
Actually, you know what would be really cool with this? Autonomous bikes[1]. Not that you need them to drive you, but that you can rent a bike and have it show up for you, or rent it at the bus depot/train station, and have it return after you've reached your destination.
> I think you overestimate how willing someone is to sit in a small car with 3 other strangers during their morning and evening commutes.
People do it already to save money. Making it easier, and adding the ability to set up a standing appointment with the same people over multiple days would alleviate a lot of people's concerns, IMO.
> Lots of people still have relatively specific needs, e.g. requiring car seats of various sizes for their young children, needing to pick-up or drop-off kids at school, sports practice, etc.
Many families have two cars, purely because of the few times a day when both are needed at the same time. For example, I have a Honda Odyssey for family use, since I have three kids. I also have a cheap commuter that I lease, purely to get to and from work a few times a week, since I work from home a couple days. That's a car I could happily and easily remove from service with little impact, as long as I had a way to get to and from work.
> When everyone has their own custom route, person A's route may make person B's route unacceptably longer.
The idea is, with thousands of routes, you might find quite a few that overlap a significant percentage of the time. Compute what percentage each person is responsible for of the whole, subtract some portion of the route for the distance that's spent going to the other pickups/dropoffs that aren't on your path, and provide a a discount depending on how much you share. Use minivans and/or full-size vans if there are enough rides to maximize people, comfort and space. The carrier maximizes profit over people and vehicle wear over a ride, and the people minimize price.
> Once autonomous cars get to the point where they don't require any driver attention at all, I think it will make ownership even more appealing since you could customize the entire to match whatever you want to do while you're hauled around.
Sure, but that's a luxury use. People that drive luxury cars that cost a lot will undoubtedly do that. People that are trying to minimize cost so they can spend that money on essentials or even just other luxuries they value more will opt for the cheaper option. It's why people carpool now.
> I think you overestimate how willing someone is to sit in a small car with 3 other strangers
Why not just keep your partition up between you and the other passengers? Because no human driver needs visibility out all of the windows, you can split the car into private spaces. When no driver is needed, it opens up a lot of different configurations.
Every morning I sit with many strangers in a train car for about an hour commuting to work. Sure a car is not quite the same, but I don't think it's that much different either. Society will adapt.
It's way different to be on a train car with a bunch of strangers than it is to be in close quarters with one or two strangers. There's a reason that a cliche way to meet someone like an online date is in a public location and not, say, a car.
I'm also looking at it from my point of a view, as someone in their mid-30's. I'd assume the generations before mine will be much more likely to accept this than my generation.
Fully autonomous fleets of cars will push societal changes we can't yet predict. Right now everyone shows up to work at about the same time. Is that a fixed position society will always uphold? I don't know, there's a lot of things I don't know. The only thing I can predict is that in the same way the personal automobile upended the social structures of the day, autonomous vehicles will likely do something similar.
There's plenty of reasons to think that. For instance, it's well studied that people value leisure time at the same time as everyone else does. Having your time off disconnected from that of your social circle causes a lot of isolation. Since social circles interconnect with each other, we end up with a single major leisure window, and a set of people in relatively low paying jobs that are stuck working at different times. Those jobs often have out of band vacation days precisely so that they can still participate in regular social life some of the time.
We can make many technological changes in the world. Some will be huge. However changes that rely on humanity not being social are just very unlikely.
Always? No. Will autonomous cars change it? Doubtful. In fact, it could make it worse, because who cares if your commute is 3 hours instead of 2 as long as you can watch American Idol while you're doing it.
I'd love to see what the average household pays for vehicles. If you don't think that money is high on the motivational list for people, you need only look at what happened the last time gas prices were $1/gallon higher than today. In any case, I doubt car ownership will "fall off a cliff". It will be a gradual change. Many people won't trust the software, there will be a few incidents that get tons of media coverage to scare people off (no doubt funded by entrenched interests). Some people won't trust that the amount of time it takes to get a ride will be low enough, especially in more rural areas. Many people like the culture of car ownership. But at the end of the day, if it saves hundreds of dollars per month and is roughly as safe and convenient as owning a car, the value proposition will be too strong to ignore. But that's a change that will take (IMO) 15-25 years to see widespread adoption.
The thing is that even in rural areas the logistics of autonomous cars allow you to get away with just one for an entire family. The fact that the car can come back home and pick someone else up later means that even if there are still people who own cars (and I suspect lots of people still will) sharing amongst small groups will likely decrease overall count of owned vehicles.
Well I certainly care if my commute gets longer regardless of how I'm commuting because there's a lot of things I can't do with my time while I'm commuting.
While people still drive themselves, there's a huge incentive to come to work early or late, to avoid traffic. I don't see that incentive increasing with autonomous cars. If anything, I predict that more people will join rush hour traffic if they don't need to do any actual driving. However, this might be mitigated by the better driving of autonomous cars.
We can predict some of the changes. I'm looking forward to the day where those with poor sight and poor reflexes (eg many old people) can get around cheaply and easily, without much risk to themselves or others. Drunk driving injuries and deaths should plummet as well. Self driving cars will be enormous for blind people that don't live somewhere with robust taxi facilities.
I think you overestimate the percentage of people that a. work b. work 9 to 5 jobs. there's plenty of cars on the road at other times. would be surprised if load was more than 2-3x as high at those peak times (which could easily be accommodated by ride sharing)
It could. It could also go from one-point-something to zero-point-something. Until now, the floor for car occupancy was 1. It's now zero. There is the possibility of empty trips (repositioning for next shuttle run, driving to distant parking lot, etc) adding rather than subtracting traffic.
I agree that a certain amount of repositioning trips is likely, but I also expect that they will be calculated for when they are fastest and cheapest. Even in an all-electric-car-all-the-time scenario, fuel isn't free and sitting with the engine off is easier on the car than idling on a highway.
If the cars no longer need a driver, then self-driving cars combined with route planning may open the door toshuttle-based door to door ride-shared commutes...
Potentially cheap enough to be a good compromise between current uncomfortable public transport, and owning your own vehicle (but unable to put the commute time to good use).
With that being said... I am also pretty doubtful. People like to own cars.
> In cities, everyone uses their car at exactly the same time: roughly, between 8-9 in the morning and 5-7 in the evening.
Good question. How the system deals with congestion. Two things:
1) Taxis make variable pricing possible, which gives markets an incentive to spread out demand. "Would you like a $15, 45 minute ride to San Francisco by 9:15, or $4, 25 minute ride arriving at 10:30?"
2) In big cities we'll start to see quick transfers to- and from- mass transit.
You'll push a button on your phone. The phone will say "Ride begins in 8 minutes to minimize transfer time". You finish your email, put on your coat and walk outside. A car picks you up, drives you to the transit center. You get out, walk across the platform, get onto a train. Train stops, you walk back across the platform into another car. It takes you to your destination.
Why? It will save you time in traffic if you care about, and it'll be cheaper if you care about that. Of course during off peak times, you can just take the cab the whole way for the shortest trip.
I've gotten plenty of sleep in cars, but never has the word "comfortable" crossed my mind when I woke up afterwards. I reckon it's something about that whole annoying safety thing requiring you to sit still upright strapped to your seat that ruins the comfort factor for me. YMMV though.
Self-driving cars of the future don't need to look / be shaped like a conventional car. Not all of them at least.
The ones designed for slow night travel don't need to be air efficient, they could be shaped like a big square with almost real beds (like on a train).
That would be hilariously unsafe in case of a crash. Look, we already have big square cars with beds in them, called RVs, and it's illegal and highly unsafe to sleep in those beds while driving.
It doesn't matter how slow and safe the autonomous car is, as long as it's traveling on the same roads as other cars, some of which are fast and driven by humans, crash safety is very much an issue.
* All the big auto manufacturers are shipping self-driving cars.
Volvo is the first to ship a Level 3 system to real customers.
* Tesla is doing OK, but the big electric car brand is Chevy.
* Waymo is a Tier I supplier to Fiat/Chrysler.
* GM and Ford are using their own self-driving technology.
* Uber is still around, but is out of investor cash, and has to be profitable to pay back its loans. The service is much more expensive to use. Their self-driving thing never works out for them because they can't spend big money to buy cars, build garages, and staff up for maintenance.
* Lyft is still around, about the same as it is now.
* The mini autonomous bus people have some installations, but they're rare outside airports and campuses.
Yes, I believe they have an L2 but are working towards and L4. Google is still going straight for L5, which is the only case where you can fully remove the steering wheel. It's a huge bed, and a bit silly, honestly. I feel like at this point, they've definitely got L4 down.
L4 basically says the car can drive itself without attention in most cases. Driver only needs to take over in extreme cases where the computer can't figure it out (something on the road blocking, sever weather, super abnormal situation).
Uber is literally all over the world, though. Google can barely launch a keyboard app outside the US. Or a payment system. Lyft is still constrained entirely to the US and a handful of large Asian cities.
I think people severely underestimate Uber's worldwide push in preparing themselves to be a force to be reckoned with when the autonomous vehicle takes off.
User is ubiquitous only because of its ability to burn an astonishing amount of money. Their bet is that they am do this long enough to make themselves the market leader. I am not convinced that they can continue to lose money at the rate they currently are for the future potential as a monopoly provider. At some point investors are going to ask for a return or bail on the company.
Amazon's biggest quarterly loss ever was about $437 million. That's roughly Uber's average quarterly loss, and Amazon did it on quarterly revenues of $20B, while Uber's revenues are optimistically $5B.
Amazon can collapse really fast. My bet is that if it does it will be almost overnight and everybody will claim that nobody saw it coming.
The reason I believe that it may collapse is because now when I want to buy something online I rather go to Walmart,Target or to the specific brand store. When I think of Amazon the picture that comes to mind is of shady people trying to scam me out of my money. My perception, even if erroneous, right now is that Walmart is less shady than Amazon when it comes to buying merchandize.
If enough people start to feel this way about Amazon there may come a tipping point and the entire company will collapse. Once the tipping point happens I don't think it will be possible for Bezos to reverse it, though he will spend mountains of cash trying to.
Do you actually go to the brand stores though? The other day I needed something that wasn't on Amazon, so I bought it from the retailer directly. And then I remembered why I love Amazon and Prime. The ordering was terrible, and I had to wait 10 days for it to arrive.
With Amazon the purchasing is seamless. That's where they really win. And they just keep making it easier. Now I can order things with my voice (using Alexa).
A few times yes. So far I've had good experiences. For example, when I decided I wanted to get a new leather notebook from my favorite brand [1] They had it in B&N, Amazon and the brand store themselves. I finally decided to buy it from the brand store and had no problems.
Of course, take everything I'm saying with a big grain of sand since this is all anecdotal. I've also purchased from Columbia Sportwear Store and some other brands that I do not remember right now. I guess so far I've been lucky.
Amazon is 1000% easier to use than to compare prices from different places. I also view it as more secure than entering my information on 10 different websites. I don't know that many people share your perception.
I'm the opposite, Amazon is a bajillion times more convenient for me to use than other online stores. And judging from how their business keeps growing, I'm guessing more people are like me than you.
I really doubt they'll let it reach a tipping point. Major companies keep tabs on user sentiment, etc, and tweak their operations accordingly. Small losses like some people finding them shady are expected in a profit maximizing model, much like lawsuits by burnt people are an accepted cost by McDonalds to avoid having the rest of their clients complain about cold coffee.
Do you have any reason to believe that many other people share your perception?
Also, doesn't Amazon have unique defenses against this possibility? If sales suddenly fall off a cliff, Amazon has AWS and lots of B2B products that could keep them afloat for a time while they tried to recover. If people en masse decided to stop shopping at Walmart, they would have nothing to fall back on.
Amazon's infrastructure is not easily replicated. As far as I can see Ubers infrastructure is essentially their app and whatever routing/mapping/job matching algorithm they have. Both of which could be replaced by a better implementation relatively easily.
The same economy of scale that keeps people using Facebook despite constant privacy issues and bad press: the network effect.
That said, Facebook has a much larger network effect than Uber. The friction to create a profile and attend an driver orientation on a competing ride broker network is much smaller than telling your everyone on your Facebook friends list you're moving to Diaspora.
You know it's funny though because I stopped using Uber with the most recent update that completely screwed up the UX. There's a billion variables that will go into which companies succeed and fail.
That may have changed with a recent update. I'm pretty sure I set it to "only when in use", but the app updated soon after. Now I went back to check, and it got set to "never".
Why does Apple even allow apps to remove the "only when in use" option?
How does that work? Is it an iOS specific thing? I haven't launched the Uber app in a while and use Lyft instead - but I just turn location services on when I want to use it and off again afterward. How does the app get any say in the matter?
They changed the UX, you now drag the pin after setting your general location and choosing your destination. I think this is because they want to suggest more efficient pickup locations instead of just wherever you happen to be standing. Felt a little weird at first but I'm already used to it.
Which is fair enough, but nearly every driver I've spoken with hates the pin and a reasonable percentage get lost unless you put a landmark in as the address.
Weird, I've literally never used a landmark and drop pins 100% of the time. I've been taking 2-6 Uber rides a week for about three years. I've never had a problem.
It won't be worth anything because autonomous ride-sharing will be a pricing game. Anyone who undercuts Uber, and there will be plenty of willing and able companies, will beat them.
There are a few other things they could compete on:
- Availability. Who has the largest network and highest availability?
- Car Design. Who has the most comfortable or fully featured cars?
- Safety. Who has the safest car? Who has the safest track record?
It might eventually become a race to the bottom, but I think there's still plenty of opportunity to compete on other fronts for at least the next five years.
Availability barely matters though. When I need a ride I open Lyft, see if they have a car nearby, and if the answer is no, I switch to Uber. The switching cost is so low it's negligible.
We tried this with our fleet of landrovers. Each one has a waterproof interior and is plumbed with sprinkler pipe with hoze-lock connectors on the back of the vehicle so it can be plugged in to the water supply.
It doesn't work, the water spray is not strong enough, or directed enough to remove the mud and sand. We still have to have people going round with water jets to spot clean. A have my doubts a fully automatic system could ever work effectively.
If you are competing on price for commuters the obvious thing is a waterproof interior and some high pressure nozzles in the ceiling. Car drives up to machine with hose, hose attaches, water sprays.
An app that books the cheapest ride for you out of all the ride-sharing companies. Like Kayak but for ride-sharing. This comment is the same as a patent, right?
You know how a person nowadays can register for uber and drive people around?
In the future, maybe people will register their CAR (which can drive itself). So they sit at home, getting work done, while their car is off autonomously driving itself and picking people up. When you want it to come home, just call it home.
Good ideas are a dime a dozen. Whatever you're thinking, a thousand other people on the planet have DEFINITELY thought of as well. The difference between people with ideas, and entrepreneurs, is that entrepreneurs make an attempt to bring that idea to life.
> In the future, maybe people will register their CAR (which can drive itself).
Until an optimized business supporting lots of cars supporting fewer people undercuts the individual owner on price per ride, sure. Uber with individual driver/owners makes some sense, because you need the individual driver no matter what the ownership structure is. But with autonomous cars, fleet ownership is going to win over individual.
Why wouldn't Toyota just launch Toyotas On Demand? If the cost of the car and fuel are the only significant costs, Toyota could probably do it more cheaply than Uber.
Uber's network effect is its driver network. It will find it a lot harder to compete if you no longer need drivers, when it has to compete with real car companies that could be able to undercut them on cost.
> Toyota could probably do it more cheaply than Uber.
I agree. The fact that autonomous cars will likely reduce ownership will put a squeeze on car companies, and it might make sense for them to branch into service areas, as they can get the cars at cost. I guarantee that for 10,000 Camrys purchased by Uber and 10,000 Camrys built and shuffled to a ride service division by Toyota, Toyota's ride-sharing division will get the sweeter deal on price.
Put another way, it's much easier for Toyota to start a ride service company than it is for Uber to start manufacturing cars to get them at cost.
I think that is suboptimal vertical integration. Hertz is a fleet buyer for a good reason; it doesn't want to be reliant on its Hertz Automotive subsidiary to meet a production quota. It doesn't want that capital expenditure. It wants to buy cars at volume and sell them as they depreciate.
Ah, but the car companies have already expended that capital. The problem here is they will have a lot of building capacity from all the factories they own, and possibly less to build with that capacity. There's nothing stopping them from buying extra needed capacity from another company, but in the meantime it may help them fill the lack of utilization they may see. At some point, some factories will need modernization as normal, and they can choose at that point to reduce factory capacity.
The point is that it's a possible strategy to get them from point A to point B when they've already banked heavily on certain future production trends if they've overestimated.
Hertz lives in an equilibrium where lots of people buy and operator their own cars. Autonomous cars may well move us to an equilibrium where most people do not own cars and instead rely on services. That totally changes the economics of the situation.
How does it change the economics of the situation? More volume or shorter rentals? Would having more frequent rentals mean that rental companies would want to produce their own cars?
It isn't efficient for Toyota to fully stock every city in the world for peak demand, and yet that's what they'd have to do to compete with the surging multi-manufacturer fleet of Uber.
They don't need to stock for peak demand. They don't need a monopoly (nor will they be permitted to get one).
To succeed worldwide, they simply need to be as good as Uber, but cheaper. And if the capital cost of the car becomes half of the cost base of taxis, a 10% saving matters.
To get as big as Uber means they need to be as convenient which means that there's always a car available. Uber handles that by being multi-brand, and by incentivizing their fleet with surge pricing.
Toyota would have to prepare for worst-case load and let those cars sit idle to keep them available. And they'd have to do that in almost every city in the world or their app would just be another local annoyance.
Also, if you order a Toyota and a Ford shows up you're going to laugh and tweet it.
On the other hand, it does give them a stable production target which would level out manufacturing spikes...
> On the other hand, it does give them a stable production target which would level out manufacturing spikes...
That's the underlying assumption I was working with. If they've invested heavily in production yet they see a downturn in individual ownership, it's a way to capitalize on that resource (factories) to stabilize production until some factories reach end-of-life and have theoretically paid for themselves. Factories which require large capital expenditure are not something you want to see sitting idle. Every year is another year where the tooling is becoming more obsolete.
I take it that Uber is betting on being thee brand for ride sharing. Maintaining market leadership is critically important for when autonomous cars take off. Even if any car manufacturer were able to undermine Uber on cost, Uber will already have a significant upper hand with people already seeing it as the company you call to get somewhere. The only question I ask myself when looking to go somewhere is, Uber or Lyft?
If car manufacturers tried to clone what Uber has built they'll probably flop on making the technology usable, and the experience may just be overall frustrating.
People will still buy Rolls-Royce or Mercedes S-class for the comfort, but most Toyota customers don't care if they're car pooling with 3 other people or sitting in a car that's not their own.
Furthermore, if there are a bunch of self-driving car services, that service will be commoditized the same way cabs are. These ride-giving services will [and kind of already do] only compete on price.
Uber's ride sharing model is fine for the major markets around the world (that have accepted it), but I think they're going to have to stretch quite a bit to successfully operate elsewhere.
Uber's autonomous trucking model is likely to aid in the complete change of that industry, though.
Not only is Uber no longer burning $1B to be in China, they now own a 20% stake in what is now a monopoly in China to (ostensibly) make all of it back and then some.
Yes, Uber may have a global precense, but that's from burning VC cash and they're not exactly successful in major markets like China. And for what? Uber is essentially a messaging app between users and drivers, and their users/drivers are not loyal.
Uber as a self-driving company would need to sell hardware to drivers to convert their vehicles into self-driving cars, and I don't think they'll have the bank to get that far, especially against larger companies with deeper pockets and existing hardware experience/partnerships.
They're already a force to be reckoned with. How do you compete against a company that might as well burn cash in the streets?
The rich are buying the future on the backs of the working class. When uber goes under, the jobs won't come back, but transit can't be that cheap. It's not like the cost of a ride was actually based on the cost of the car OR the labor.
Why would anyone own a house when they can rent an apartment? Why don't we share bathrooms with the neighbors? Why have a small garden that needs maintenance when there is a huge park nearby? Why did I just buy a printer when I need it so rarely and there is a copy shop downstairs? Why do people have washers when laundromats exist?
I think ownership means more to most people than just convenience or being the cheapest option.
It does mean more, but it's not immune to other constraints, and it may be enough to make demand fall off a cliff. Cars are expensive (relative to the average income) and for many, it'll be hard to justify ownership if the alternative is cheaper.
Is there a particular reason to think Uber is going to be last to autonomous vehicles? I can see a good argument that they're head of the pack today, between Tesla's not-really-autonomous vehicles and Google's perpetual inability to execute.
Head of the pack of dreams maybe. What does Uber have, in a tangible manner, right now? At least both Tesla and Google have real systems that are on the verge of full autonomy.
This is a fair point. Uber has yet to ship any hardware, and really any software aside from an app that is a couple steps above "Uber for X" that are whipped out in a weekend.
I know they gobbled up a bunch of researchers, but moonshot engineering feats like autonomous driving is not (yet) in their DNA. And their current business model is not profitable and will die the moment they go up against an autonomous service.
They don't need to be last to market to lose, they need to be first in order to survive.
I don't particularly care for Uber, but they do have prototype self driving cars on the street in Pittsburgh. So it's not quite fair to say they're failing at this.
I saw an uber car covered in a crazy amount of sensors a couple of days ago on Van Ness avenue... There was a human driver, so I assumed it was mapping the environment, but it might have been a test vehicle; it looked similar to the pictures of the autonomous cars in Pittsburg.
Uber will continue to hold a lot of the chips in this race. Consider the Tesla or GM examples - a regression in revenue will not be acceptable. So if they're currently selling x cars at y $ a year they will need to continue that revenue without actually selling cars. This means they'll have to get really good at understanding how many rides over the span of a vehicle will continue that revenue and provide just the right amount of vehicles to meet a regions needs. Uber has that information in spades.
Netflix makes a good example it is possible to make a major business model adjustment and return shareholders very well. I think investors have a very realistic view of how much the automotive market is about to switch. The companies who provide no evidence of preparing for the switch are just as likely, to me, to see their market cap plummet as ones trying a risky pivot.
Certainly some auto makers may choose to go down the route of executive assisted suicide and maximize shareholder (and executive) profit distributions in the process, as IBM decided to do. To pretend than the next 50 years will be like the last 50, I'm not sure how long that charade can hold up.
Personally I think Google and Apple are the ones who hold all of the chips. This definitely can be turned in to a commodity business with a few OS adjustments. There are some cards Uber can play to avoid this, like exclusive infrastructure agreements, although not really sure where that will be permitted geographically.
A more interesting question is will this look like smartphones where a single company gets 90% of the profits and the rest of the companies do huge volume but capture small to zero margins.
This means they'll have to get really good at understanding how many rides over the span of a vehicle will continue that revenue and provide just the right amount of vehicles to meet a regions needs.
And a pox on all of us if we hand firm control over regions to single players.
I dont see how that is the case. User is currently burning through more eyes to maintain their market position and that is without a fleet of vehicles to operate and maintain. Moving to autonomous vehicles may cut their staff costs\remove drivers from their costs column, but this cost is moved to vehicles and in many jurisdictions at least for the medium term a human driver will be required even if it is just paid to sit there.
My point was that regardless of the dollars and cents for Uber to transition from staff to cars, or even if it works out at all, Uber still has the information to decide it will take X cars in Y region to make Z revenue. The only other player with that kind of information (on a smaller scale) is Lyft.
> Private car ownership is going to fall off a cliff shortly after autonomous ride services arrive. Which probably means general demand for vehicles will fall off a cliff.
I really dont agree with this. Cars a so cheap now and I'd much rather use my own than some dirty shared car.
Already renting a car is annoying as you never know if you'll get charged for some damage that you didn't cause, I can't imagine having the same issue every day.
Having the car drop me off then go away and autopark is something that I'd value though. :)
Presumably, in a shared autonomous vehicle, the only thing you'd be responsible would be the interior. I would think this would be fairly easy to monitor on a user to user basis.
It's not either or. Autonomous service for commuting is still highly valuable even if you retain your own car.
Over the medium term, this will only encourage ever more massive and distant suburb creep. Also, with longer commutes less tiresome, the negative cost of having a longer commute will be mitigated.
As for the stranger effect, there are many innovations on that front as well. Google is already has a working app that tries to arrange commutes for true ride sharing where passengers only compensate the driver for the pure cost of the ride. It's cheap and effective if you are a bit flexible in the morning.
I can see autonomous services being so cheap that they become the new corporate perk and companies find value by having employees ride together, share ideas, socialize, etc.
Also, the economies of scale alone would make a private automated car trip still cheaper than owning privately an autonomous vehicle.
The technology will move fast enough that the vast majority people will be switching from a human operated car to a self-driving car service. I can see the ad campaigns now. Basically showing people how much cheaper and safer and more convenient it is.
Cars can be seen like horses were before them, an expensive hassle that requires land for storage and usage. If you think about it, people have whole buildings just for storing cars, garages, like we did with horses, barns.
15 years from now suburbs will be built without garages and driveways. There will be a cottage industry for people to convert garages and carparks into more usable space.
20 years from now, multistory car parking lots will be the new conversion buildings like the old meat packing and textile plans that used to be in downtown areas that were turned into lofts. Polished concrete floors will feature the original parking space numbers. The open sides will feature panoramic views.
Lastly, having your own car will be a throwback element. People who like to 4x4 or work on their own cars (a huge market that might actually grow in the medium term) will fill this but, even then they'll use automated services to commute. The number of people who have hobby cars (off-roading, auto racing, vintage collectibles, restoration) will skyrocket because a car no longer has to be a daily driver.
Or people might care less about cars, I don't presume to know the complex of a future of human desire. But, in a world full of cheap cars that you driver yourself, automated car services will still create so much value that people will still switch.
a whole industry of workers will be out of the job. How will society react? Will we have a socialist revolt? Want to stay clear from politics... but seriously why isn't anyone talking about the socioeconomic impact this will have once drivers are put out of the job?
IMHO this is the reason Trump won the election, because there is no plan for these people.
There are plenty of conversations around that idea here. UBI for example. Some people think that the population in general will retool (some will get left behind). Others point to the horse population in the US that went from some 5 million to a few hundred thousand following mechanized transportation as an example of what will happen to the middle class.
"Retooling" sounds like such a great idea when you're a 20-something urban male single software engineer, you've got nothing to lose, and the world is yours to take. It also sounds like a great idea when you're independently wealthy and you don't have to work for a living, so you can afford to bloviate from your ivory tower.
Most other demographics have very a different perspective on that concept, and for good reason.
There will be cataclysmic changes ahead if we don't seriously reconsider, well, everything about society. Under the current way of thinking, millions of people will get sent to the glue factory - to persist in your horse analogy here.
I have no idea how to solve this problem. My cousin's husband is a truck driver. He makes a six figure income running between Indy and Chicago (bang, bang). He's going to retire before they can get rid of him, but this is a huge issue.
I tried to explain this problem to my father, who's a CPA. I've pointed out that his firm can get by with a dramatically smaller office than it use to due to tech. Same's true for his lawyer friends. I pointed out that the law firms are not training the next generation like they use to since they use fewer low-level lawyers for research thanks to Lexus Nexus. Account is next.
I'm not a huge fan of UBI since I think that makes people slaves of their State. We've seen such enslavement in the US under welfare programs. We've set the program up to lock people in. If they try to get out, they will lose benefits (donut hole issue). I think UBI is like that, but worse.
> I'm not a huge fan of UBI since I think that makes people slaves of their State.
When America will be able to shake off its own mythology about this notion of the "State" as an inherently bad actor - then, perhaps, America will be able to get out of the current rut it's stuck in.
Yes, the "State" can be bad (I should know, I grew up in an actual bad "State" in the Eastern Bloc before the revolutions). But at the other extreme, societies can sabotage themselves from within, trigger the social equivalent of an autoimmune disease, and collapse from endemic distrust in the system - which to me is pretty clearly the direction America is going now.
Isn't the whole point of UBI to be Universal? People don't lose the benefit if they earn additional income and everyone gets it so it shouldn't have an inherent stigma to receiving it.
I'm not actually sure its a practical option, but I don't see 'enslavement' as being one of it's downsides - where's the 'lock in'?
My concern is what happens if a block of people start to become disenchanted with UBI. Will the State cut their funding? Can the State, as the sole employer, lock the group out of the political discourse?
In general I'm not a fan of going to the United States government hat-in-hand. Look at Scotland. They do that with the Brits. They're kicked around all the time.
Here's an example. South Dakota v Dole, the Supreme Court said the government could tie general funding to states to enacting state laws. In this case the states had to raise the minimum drinking age to 21 if they wanted road funding. What's to stop the government in a UBI world, from putting different life-rules on its people.
The government goes from working for the people to providing for the people. That's a massive power imbalance. I don't know of any government I want to give that power. I do know it's not the United State's.
A UBI should be universal, so you can't lose benefits - it's only as bad as how you become slightly less rich when your income goes up under progressive taxation.
The closest real problem I can think of is not being able to move to a $15/hr minimum wage city from a poor area because you don't have the starting capital to buy the basics there.
> I'm not a huge fan of UBI since I think that makes people slaves of their State. We've seen such enslavement in the US under welfare programs. We've set the program up to lock people in. If they try to get out, they will lose benefits (donut hole issue). I think UBI is like that, but worse.
Bad policy structuring, because welfare programs are legacy programs and were designed when Economics was still in its infancy.
Also keep in mind that the welfare population is a minority. It's similar to the bank anecdote - if you owe the bank $1000, you have a problem, if you owe the bank $10,000,000, the bank has a problem.
When tens of millions of people in a country with gun ownership depend on UBI - any government structure that tries to reduce benefits or eliminate it will become a target of the masses.
Won't happen in the USA. Due to our puritanical obsession with "earning" a living through labor, we'll never see Basic Income. It would be more politically do-able to pay 1/2 the people to dig holes and the other 1/2 to fill them in, than to implement a sensible UBI system.
What I don't get, even now, is the lack of public works. Digging a ditch and then filling it is dumb. Why not have individuals on welfare work cleaning out the forests in California? Playing a role in sensible land management implements? There are many places that the unused labor could benefit the people. At least then there is a work/trade for the money. In the end, those people get life lessons that would benefit them in the workplace.
Right. Instead of just giving people money our governments at federal / state / local levels should become employers of last resort for everyone who has exhausted their welfare or unemployment benefits. Guarantee 30 hours / week at minimum wage to anyone who wants to work. Even if it's just trail maintenance or graffiti cleanup they will at least be maintaining basic employment skills, and have enough spare time to retrain for something better.
What? Everybody is constantly talking about it. I have heared almost as much about that, compared to the actual technology.
The answer is of course the same as with any other reason people lose jobs and have done so in the past. It really is nothing new. Neither in scale, nor in severity.
Also, every state already has system to deal with that stuff. The hole point of a generalised systems is that you don't have a major political crisis everytime insert <big company or industy> suffers job loses.
That Trump won because of them is an assertion that is made often, but hard to proove or even show good evidence for.
What happened when cars were first invented? What happened when interchangeable parts were invented? Every time some new tech gets close to being revolutionary, the same sky-will-fall arguments get repeated. New markets are created, new industries develop; the economy continues to grow and worldwide poverty continues to drop.
Why would it be different than previous instances of whole industries being out of a job? There was no revolt when the US shipped manufacturing overseas.
You've mentioned $1 a ride a couple of times. Where is this number coming from? It doesn't pass the smell test to me. Even bus fare is several times this amount.
If I have to bet against anybody it will be Uber. They will be burning cash and given that Google just gave up on building hardware why would one think Uber will succeed?
Uber hasn't demonstrated a capacity to succeed, but unlike their rivals, they need self-driving cars to succeed for their company to succeed. The CEO has called self-driving cars "existential" to Uber.
Contrast this with Google. Self-driving cars are an expensive hobby to Google, and completely inconsequential to the success of their core business.
To be fair I don't think Google is the gold standard for hardware. Outside of a few successful hardware releases (Chromecast and arguably the pixel line) they seem to struggle in this area.
But that doesn;t mean companies should be afraid to try new things. History is evident that more companies died resisting the change than anything else. Car ownership going down should not mean companies like Uber should not venture into self-driving cars although by burning huge VC money. Even the VC's who funded would want Uber to the fore runner in the self-driving cars. The other factors why companies are still investing although they might loose money are:
1) It will take atleast 5-7 years for self-driving cars to become ubiquitous. Till then there is good uprising market for every player to make money. secondly, if you have a product ready, it might open up new markets in future which currently no one realizes.
2) Car ownership will still be there. People who can afford it will still prefer private self-driving car than waiting for Uber/Lyft. I would say self-driving private car will be seen as a new luxury status.
3) car manufacturing will definitely go down with less demand but car manufactures will venture into different pricing model with ride companies. Right now, a car is a one time sell for a manufacturer. It might be subscription which ride company uses.
That's what people said five years ago. I am sure something will happen eventually, but it will take longer than people expect and the results will never be as good as people are currently fantasizing that they might be. We'll get to "good enough" but never to "great".
I don't know of anyone who was saying self driving is only five years away in 2011. It was still a fantasy then.
Now we actually have working software and hardware, and real people are experiencing self driving. I live in the bay area, but not a month goes by that I don't see a self driving vehicle at some point. And that's not counting the Teslas that are autopiloting their way down the highway next to me that I don't even know about.
How is it that "real people are experiencing self driving"? I was under the impression that Google's autonomous vehicle and other projects like it were still all technology research prototypes. There is a very big difference between the sort of next-generation cruise control driver assistance features Tesla and others are providing, and the sort of fully autonomous wheeled robot people are fantasizing about when they hyperventilate over the city of the future and its wondrous transportation technologies.
Maybe my negative feelings are excessive; I'm just really, really tired of having people jump in with the whole "but self-driving cars will change everything!!!" routine every time people try to have a serious conversation about public transit investment.
They're more than just prototypes. Their employees are using self-driving cars for their daily lives. If you have a friend who works at Google (well Waymo now) who is using the car, you can get a ride. Also, as a driver on Mt View roads, I have experience self-driving as someone behind and next to the vehicle. They aren't the "robotic" drivers you'd expect, but fit in pretty well. They still show signs of being a robot (like the way they approach a turn with a biker is very mechanical) so you can tell it is self driving, but it clearly works.
And also there is Uber's self driving fleet. They are actually taking people from place to place who use the Uber app in a self driving vehicle.
In all cases someone is still at the wheel, but they aren't doing anything other than monitoring and logging the ride and cases where the car didn't behave like a "normal" driver.
Will they stop when they see a ball fly across the street, because a child may be chasing it. Fringe cases will determine their success. Until then, it will be glorified cruise control.
Then the question arises who is responsible for that imperfection. Remember the Toyota auto acceleration issue and how that played out socially.
Safe driving is one of the most complex things done by people. If tech companies are good enough to solve that, there many things more profitable and easier to solve.
I beleive it will be more realistic for them to develop seprate and exclusive infa for these vechiles. We can then plan rest of the things around that infra.
Integrating self driving cars into current infra will require a true AI or something closer.
I'll be a believer the first time they actually put a fully unmanned vehicle on a public road with no "safety" driver on board. That'll be big news indeed and indicate that the technology is far enough along to be taken seriously on a massive scale.
For some very limited use cases five years may be realistic. For a car that can do everything and go everywhere a human driver can it's more like 50 years.
> Also: Private car ownership is going to fall off a cliff shortly after autonomous ride services arrive. Which probably means general demand for vehicles will fall off a cliff.
Or, if everyone has a self-driving car of their own (or one that can be converted to self-driving), people will rent out their cars AirBnB-style when they aren't using them. I mean, why not? You've already got a garage, a driveway, etc. Might as well make a few extra bucks with your vehicle...
Not going to happen, or if it does it won't last. The so-called sharing economy has mostly just devolved to low-barriers-to-entry entrepreneurship. AirBnB listings start off with ordinary joes renting out their spare bedrooms, but they got crowded out in a few years by businessmen providing real services, and consumers follow.
Uber drivers at first were mostly just driving in their spare time, but over the years the professionals came in and crowded them out.
The lesson here is that it takes work to provide a level of service that consumers are willing to pay for. That work is best done through organization and professionalization. Some guy renting his car out during the work day isn't going to find it profitable enough to continue doing when he has to compete against dedicated on-demand rental services.
But AirBnB and Uber require the owner to invest effort: cleaning the house (or arranging the cleaner) and driving the car. If you could hire your car out fully autonomously that changes the equation.
It's like you're saying I shouldn't bother with my zero effort $100-a-month passive income just because somebody else is making $200-a-month with higher margins.
As an AirBnB user, that's completely opposite to my experience - 9/10 have been ordinary people doing it on the side, not pros. I suspect the picture is different if you raise the "price" slider (I go for the cheapest third, usually), but I still see plenty of supply.
Two sides of the same coin. If you can rent your neighbour's car, why own your own?
Heck, why have driveways? In cities, at least, I could fully imagine new townhouse developments operating more like apartment buildings: one central parking lot for the "pool" of (mostly) driverless cars. Your car drops you off at your door and then parks in the pool; then comes back out to get you in the morning.
While driverless cars obviously make short-term car rental more attractive on a cost basis, I think that people here tend to overestimate the impact of this.
* A large percentage of people will want to have their own car because they don't want to put up with the occasional hassle of unavailability of rental cars during a peak time, and/or occasional long waits for the car to reach them (particularly when they are being picked up in un-dense areas). And before you hand-wave this away, note that the more of a cost advantage you give to rental cars (ie, the higher the utilization of those cars), the more of a disadvantage cost/wait time will be.
* People -- and very especially families with children -- also value cars as a mobile depot for their stuff. If you have kids of certain ages, this is practically a necessity. I definitely am not going to screw around with not just a carseat but like a dozen books and snacks and wipes and cups and so forth every time I get into/out of a rental car.
* Then cleaning and status and all the other things that people regularly bring up here.
* Availability can be addressed with a dedicated option, where you pay extra (if a fleet) or get paid a bit less (if renting out your self driving car) to ensure there's always a car whose destination's time distance from you plus remaining trip length is less than some amount you specify such as 15min.
* To address the mobile depot use case, I've heard of proposals to have storage areas (possibly trunk, but not necessarily) that are only accessible by the owner, not by renters. Even that may not be perfect though, I do see your point.
Sure, availability can be addressed by paying more, but the entire value proposition here is that you pay less.
Is there a sweet spot where you still experience significant cost reductions but the inconvenience is small enough that you don't mind it? Well, certainly, for some people! But I think the number of people is going to be a lot smaller than many commentators think.
(The only cases I see where there's no increase in the rides-for-hire business after full automation is if the cost of cars drops to the point where there's very little incentive to try to reduce the costs for most people. Given that cars are quite expensive now, that seems unlikely. I can construct a scenario, but you basically have to believe that, first, the additional cost for driverless sensors+software is very low, second, that driverless software is so safe that you can drastically decrease physical safety features on cars, and third, probably that automation drives significant GDP growth).
> Is there a sweet spot where you still experience significant cost reductions but the inconvenience is small enough
It's not that hard. For one thing, you don't need to increase the price much, you just need to increase it a little so that you push cost-over-everything riders out of the queue. There are enough of those that whatever capacity is left should be plenty.
And you are worried those riders who were pushed out of the queue will go back to car ownership, but they are by definition cost-over-everything people, so of course they won't do that. If they could afford to own a car they could afford peak pricing.
My point is that millions of them already exist, with a distributed network of roads leading to them, as well as storage (garages). It would be costly to suddenly rip all those up.
They could be used for other purposes. People filling their garages with other junk and then having to put their car somewhere else is so common already. The driveway can become a front yard.
This is an interesting thought experiment. What do people feel about recreational driving or weekend leisure driving ? For commuting, I am totally in favor of autonomous vehicles. But for a weekend or leisure drive I may not want to bother with shared ownership.
I am not from U.S. but the experience I have had with using local air travel (with Southwest) in the U.S. I am prepared to caffine up and drive up to 400mi rather than bother with the hassles of flying. (I have done that a few times and found it quite enjoyable, maybe there was a novelty factor to it)
You would call a car on your phone (with maybe a ski rack or whatever you need). Car will drive itself to you and you can drive it manually or in automated way over the weekend. When you return back home, car drives itself away.
No hassle with maintenance, repair, parking.
As someone who once had a long-distance relationship with a 400mi round trip, I'd say it's a little of both. There's a novelty factor, but there's also a part of it that some people find enjoyable.
Having the right car helps: trips made in my almost-busted Honda Civic were much less enjoyable than the ones in my brand-new-custom-ordered Mini Cooper. :) The right soundtrack, the right weather, and the right caffeine vehicle are major factors as well.
I tried renting out my car for a month in Oakland; it routinely smelt like weed and McDonald's fries, which got annoying enough that I stopped renting it out.
A fleet would be able to handle these sorts of things better.
So far sharing economy companies don't want to vertically integrate their whole service stack down to physical goods. It requires capital raising, and is low margin compared to the value that they provide by scaling their main services. Maybe they will once they've achieved full market penetration and are struggling to increase their profit, but that's decades away IMO.
But why have a garage and a driveway in the future? If you can "hail" a self driving car within 5 minutes then having one in your house becomes a waste of valuable space.
I wonder what the worst case scenario is here. I'll put Muni-level syringes and human waste as the worst. But (1) fare evasion wouldn't be a problem and (2) the 1% of worst offenders who account for 99% of the disgustingness can be readily blacklisted.
I suspect the typical case will be a drunk person throwing up in it on the way home. Hopefully they self-report; if they don't, it comes to the next rider to report it, who does and gets a generous reimbursement, charged to the credit card of the last rider.
And the fallout for the economy will be interesting. What does it do to home prices when a long commute is less of an issue? What happens to all the cottage industries built around the assumption of someone owning and driving their car? Etc.
I would disagree: fierce competition is what will get us there the fastest. Every company is already a team of people working together. Let the games begin, and may the best team win. And may we, the consumer, forever reap the benefits :)
In a cooperative environment, bad players bring everyone down. In competition, they drown, and only the strongest survive.
It feels there's a mountain of hype around autonomous vehicles. I think the core challenges to a fully autonomous vehicle (with no human backup), are still far from solved. They need to reliably deal with an almost infinite number of edge and corner cases, each quite different from the last. For example:
- communication with other human drivers. In London, this is required all the time, like when parked cars block the road, allowing just one car through. Or traffic light out of action, so you negotiate with other cars using hand gestures, light flashes etc
- endless roadworks, that change what lane you're allowed to go on, turn a two way road to one way road.
- random debris on road. Plastic bag - safe to drive through, wooden plank - safe, plank with nail - not safe.
- loss of GPS, mobile data, or both (again, surprisingly frequent)
- making way for emergency vehicles (sometimes need to drive into lane you're not normally allowed to go, I.e. Bus lane, pavement)
- policeman coordinating traffic
So far, I haven't found any evidence of autonomous cars dealing with the above. If anyone has, please post.
> By early 2015, Fairfield thought they were getting close: The cars were clocking full days at Google’s testing grounds without needing human takeovers. They had driven 1.2 million miles on public roads and could pull over for emergency vehicles. They could detect and brake for squirrels, and read hand signals from cops and construction workers. They knew when to honk politely (Just making sure you see me here) and when to blare it (You’re about to slam into me!). Then the team spent the next year putting another 800,000 miles on the fleet’s collective odometer, to fine tune everything. And it gave 10,000 rides to employees and “guests.”
The car-human interaction bit isn't from a quote, is there any evidence of this actually happening in the field? Would be great if there was a video or something.
You can see one example at the end of the video on their technology page where the cyclist can't make up his mind. It recognizes the left turn sign from the rider (you can see the sybmol change when he extends his arm).
This kind of predictable reactions to specific human gestures.. I fear it will have great uses for certain type of crimes. What will you do if someone stops you with the purpose of assaulting _for eg, a robbery_ you and the car refuses to run the dude over because he made a gesture and is now in front of your car while his buddies are waiting aside?
There are good reasons why things like these happened:
> The law of some states, such as Louisiana, explicitly lists a killing in the course of defending oneself against forcible entry of an occupied motor vehicle as a justifiable homicide.
While I'm interested in self driving cars that still feature manual controls you can override, like Tesla, I have no interest in a future controlled by Google's AI without my input.
Do you really think the companies building autonomous cars haven't thought about all these things? They may not have solved them yet (If they had, we've probably see their products on the road) but I'm sure it's in progress.
I agree that it's safe to assume folks at these companies are aware of these issues. But if I'm going to risk my life by using one of these systems, I'd want to have good answers to how they have solved them. Until then, I think it's fair to keep asking to avoid unrealistic hype.
I think safety is fine, especially in cities with not too crazy weather.
It's more that if an autonomous car with no human stalls, due to a traffic light that's stopped working, or an oddly shaped plastic bag in the road, it's stuck, and would cause a traffic jam. Not sure how to resolve such a scenario, can't rely on someone remotely taking control, as mobile data doesn't have 100% coverage, so a human who is allowed access to the car would somehow need to get thorough the traffic jam to move the vehicle on. Can't think of an efficient way to resolve this situation.
First, some baselines: humans suck at driving too. Next, coverage isn't an all or nothing thing, it'd be quite easy for an autonomous uber like service to restrict itself to areas that permit remote control when cars need assistance. Last I'd encourage reading some of the literature behind what folks are doing: it's a very different sort of system, one that blends a lot of complex probability estimates. It's not like everyday coding where a human developer needs to anticipate every combination of corner cases.
Machine learning tools also cannot cover corner cases, if they're not present in it's training data. Communication with human agents (other cars, police officers) is also extremely tough.
It's not as if human driven cars never get stuck in an intersection. I imagine the solution will be much the same: it will be moved manually if possible or towed if not. If a person is deliberately fucking with traffic by deliberately messing with an autonomous car, they'll also probably be fined.
I mean the other day on my daily commute there were two cars right in the middle of a freeway, one attempting to jump start the other and they were at it for a while. This is not shockingly new ground.
But nobody's asking you to use any of these systems yet. they're in development, not products for sale. There's no benefit to anybody in questioning whether google knows that snow exists, or if uber understands that traffic stops unexpectedly sometimes. Obviously everybody involved in developing self-driving cars understands the most basic definition of the problem.
If google actually starts selling a self-driving car and doesn't give any information about whether it can deal with a policeman-controlled intersection, then you can be concerned. Until then, implying their self-driving car programs haven't considered the basic problems that any uninformed commenter on the internet can think up is not incredibly helpful.
That may well be the case, but it does not follow that a 98%-complete self-driving car would actually be a viable product. It's not good enough that the robot can beat a human in a head-to-head skills test; the robot also has to be as versatile as a human, across an unpredictably wide range of circumstances. That's hard.
Depending on where you live, "can pass a standard driving test" isn't necessarily saying very much.
In any case, I'm not talking about whether it "should be allowed to drive", I'm talking about the level of capability an autonomous car would have to reach in order to become a viable product.
No its not, that is kind of my point. It only needs to be better at driving than the average human to be technically viable. And yes, it may very well be allowed on Indian or Chinese roads before its allowed in America, but that's human fiction at work not demonstrable ability.
I'm not talking about technical viability! I'm talking about product viability. Can this hypothetical device be manufactured and sold for a profit? Can it be used, profitably, for some purpose, for which it will be cheaper or in some way preferable to existing solutions? It is not enough that this autonomous car can drive around the block without crashing; it has to be good enough, reliable enough, efficient enough, that you can build a profitable business around the unique service it can provide. That is much harder than simply passing a standard driving test.
Yes of course it is enough, if that is your definition of 'viable product', that should be self-evident. It only needs to be better than a human at driving for it to be a viable product. Are you prepared to pay the insurance premium on a non-autonomous transport appliance in 10 years time ? I'm betting you won't.
"Driving" is not one single monolithic problem. "Driving" is a whole array of complicated problems. The "standard driving test" exercises a very limited subset. "Better than a human at driving" is a very broad and poorly-specified claim.
Hell yes I will still be using a "non-autonomous transport appliance" in ten years - it's called a "motorcycle" :) - and I don't want your high-tech self-driving future if it has no room for that.
Clearly 'driving' to you is an expression of personal freedom - not body transport. And that's good and well, but finding unrelated justifications to defend what is essentially a hobby is silly.
People will continue to spend stupid amounts of money on hobbies. I'm not going not hand in the keys to my Landcruiser or my CRF250. But I don't leave those parked at the train station as a rule either.
You asked me a personal question, so I gave you a personal answer, somewhat flippantly as I disagree with your premise.
That remark was separate from and unrelated to the actual point I've been trying to make: there is a very significant gulf between the notion of a "self-driving car" as a form of automation built into a vehicle which is still under the control and responsibility of a human driver, and the notion of a "self-driving car" which is a fully autonomous robot built in the shape of a car, that must be capable of doing something reasonable in all situations without depending on a human. The latter is a much harder problem than the former, and we don't yet know whether it can be solved in a way that yields a practical, profitable transportation system of the sort many people like to fantasize about when they discuss (for example) Google's autonomous vehicle program.
Seeing as there are already people trying to use their Tesla autopilot as autonomous capabilities, I'm pretty sure they'll be viable as soon as they are road legal.
It is very clear that autopilot will be a successful feature and we can reasonably expect continued improvement in human-driver-assist technologies for years to come.
But people keep on handwaving past the very important gap between "automation which assists a human operator" and "autonomous vehicle which gets around on its own", as though doing well at the first problem necessarily makes the achievement of the second problem inevitable.
Driver assist technologies make all kinds of sense but they do not fundamentally change the nature of our transportation systems. They just make it easier to do a better job as a human driver.
Getting to the point where most wheeled vehicles are self-piloted autonomous robots with no human driver interaction...? I still think that's a really hard problem.
Sales. That's what makes a viable product. One that will sell. Despite elsewhere in this comment chain you saying "I'm talking about product viability not technical" I can't help but think that your response to me was about technical viability.
People are already willing to buy and trust their cars to drive for them right now, when they're specifically being told "don't trust it to drive for you". The second it's legal to advertise "let it drive for you!" autonomous cars will sell regardless of their ability to perform well in unregular circumstances.
If "self-driving car" just means "I'll drive for you, most of the time, while you sit behind the wheel", then we are talking about nothing more than a better automobile. Of course that will sell, but in that case the world of the future looks pretty much the same as the world of today, with fewer crashes.
But people keep talking about "self-driving car" as though it is going to be a robot capable of running errands on its own, with no human driver and not even necessarily any human passenger. The question is still open: can this be done well enough, enough of the time, solving a valuable enough class of problems, that producing such car-shaped robots will actually become a significant industry? Furthermore, will the problems solved by such hypothetical car-shaped robots be significant and widespread enough to justify the hype currently being lavished on them?
Meeting the hype expectations of people who practically worship technological change is a pretty high and imo unreasonable bar for viable product.
Personally I'm of the opinion that because we've been making tools for 2.5 million years and don't seem any happier than the average animal, the very concept of a new tool making all that much of a difference is absurd.
I agree that autonomous cars are happening, but there are a ton of assumptions about human capabilities embedded in a standard driving test. It's not nearly rigorous enough greenlighting autonomous vehicles.
Sure, but put a tumbleweed in front of the driverless car and it's stuck. Or wood with a nail sticking out/generic movable obstacle, a human can remove it, the auto-car cannot.
Yea so ? Harping on the 2% issues does not negate the benefit you have the rest of the time: It can swerve in microseconds, you cannot. How many times will your nail removal skills save you or your kids live compared to software that knows that rolling the car to the left have a 87.34 better chance of occupants survival that the head-on collision with the truck that just swerved in 0.4 seconds ago ?
Forget it, driver-less s cars will be the norm in the future. You won't be able to afford the insurance associated with a meat-driven car.
My nail removal skill will prevent gridlock on the road. If an auto-car gets confused due to a 2% issue, it becomes a brick, clogging the road. If multiple auto-cars do this in a city, the result is standstill.
You make it sound like people are gonna be strapped into their autonomous vehicle like Hannibal Lector. Seriously, a couple pieces of debris are not some thing where we go "oh crap, we're done, guess the whole city will just sit here all day."
Can it not drive around the tumbleweed? Or, if it's completely walled in by them, make an assessment of their density and likely harm caused, and decide to just run over them?
If we're crediting the car with being able to handle merging into a crowded multi-lane highway and dealing with idiot human drivers who don't indicate or hold lane position very well, it seems a bit odd to arbitrarily presume that it won't be able to identify non-hazards like said tumbleweed.
Yes, a computer drives better than me. But that's not it's competition.
Does it drive better than me driving while being supervised continuously by a computer? I still make the important decisions, and my mistakes are pointed out by the computer.
The competition for self driving cars are not humans, but humans being adviced by driver assistance programmes
No sure why you make that distinction, or what the benefit of that will be. I want to read HN on the back seat on the way to work, not "supervise" the driving appliance.
Op has a good point. You might want to read the web in the back seat, but if you supervising the computer makes it safer, will the public let you doze in the back seat putting your life and the lives of others at increased risk?
If I have to supervise, there is no point. Having a human intervene in what should be 100% automated is putting my life and others at risk. Look at road death statistics: the vast bulk of incidents are are caused by humans sucking at driving - why should anyone want to continue doing that ?
I would say that humans should drive and computers should supervise and correct obvious mistakes, that however is status quo right now. It doesn't sell as futuristic fiction
The number of times that "you can't stop / brake / slow because someone behind you will run into you" comes up in these threads is always alarmingly high.
> Miller and Valasek’s full arsenal includes functions that at lower speeds fully kill the engine, abruptly engage the brakes, or disable them altogether.
Are there seriously cars on the road with complete fly-by-wire brakes which have no pure mechanical backup? That's horrifying. How did we go from dual redundant hydraulic systems that work (if in a reduced capacity) even when the ignition is off, to something at the mercy of some code that can be bamboozled over 3G?
I imagine in an emergency the car would show a set of options on a screen and you click one and the car will do it.
For example getting out of the way of emergency vehicles on a blocked road. It will ask you what to do and if you tell it to drive up the pavement or into a bus lane or cross to the wrong side of the road -- the car will just do it.
I do wonder about all the non-verbal human-to-human communication. But a friendly baxter type screen with eyes that move (and messages that scroll across) would solve that.
We have to remember that these cars are all feeding unusual situations back to the google BRAIN that is learning. It'll watch us react to an emergency and learn. It'll see that 20 or 30 times a day and generalise a response.
Bringing human decision making into the loop makes things messier. If any of these human decisions need to be made quickly in the interest of safety, then the driver needs to be paying attention at all times. That won't happen. This is very dangerous. This is also complex in terms of liability. If the car hits something while driving partially autonomously, where does the responsibilty lie? The software, the driver, or somewhere else? We're far from the widespread use of autonomous cars. I'm not even going to mention the dangers of buggy software written by large companies that can't get iCloud notes syncing to work.
Agree 100%. This back and forth between fully-autonomous and partially-autonomous highlights the largest obstacles for "self-driving" car adoption.
The fully-autonomous route runs head-first into the inevitable, messy problems with software: security, maintenance, feature creep, complexity, unreliable networks etc. In theory, humans could nail all of these in car software. In practice, it's highly unlikely. The software industry can't even get basic security for IoT devices right.
The human-in-the-loop route runs directly into the only problem more daunting than reliable software: human nature. Human overrides for self-driving cars that need to be activated in a timely manner will not work. The drivers won't be paying enough attention to react in sufficient time. Note that overrides add even more complexity to an already fantastically complex system.
I still don't see what's wrong with current systems (driver assistance systems)
Say an always on self driving car, but it doesn't drive. Human has to drive, and the computer can correct any obvious mistakes the human does. Keep a really high liability for the car manufacturer though
I think it's fine if there's a human in the car, simple, they just take control to manage the incident. But to truly disrupt our lives (24 hour robot trucks, autonomous taxi fleets ), which is what all these companies are aiming for, the car needs to be able to drive without a human.
Having a human on call to remote into a Car is still a viable option in most of the US. This is not a good fit for sub second response times or tunnels, but it vastly reduces the endless .0001% edge cases.
ED: Not hitting kids is manditory, not blocking traffic is mandatory, parking in every random stop is optional as long as their are safe places it can drive to.
Yes, it may provide a good impetus to upgrade our mobile data infrastructure. It would have to be ultra fast, ultra reliable, ultra wide coverage. Because if you can't remote in, and the cars caused a huge traffic jam, it'll be very difficult to get the owner companies human in to help out the vehicle. I guess the recovery man can just use a bike :/
It still needs to work 99.99% of the time. I am more thinking extreme edge cases like: "Drove off the highway to avoid collision and stuck in a ditch. Bridge out now what? etc."
Basically, you don't need a perfect solution in every case, just a safe solution in every case. A self driving car that get's to a gas station is a good enough step 1 even if you have some human attendants there to fill up. Sure, it might add $2 bucks per fill up that's still cheaper than highway driving.
PS: Self driving car, only works in NYC is also useful.
And if you can remote in, I'm worried about hacks. Imagine the chaos if a terrorist group got in and turned the wheel of an entire fleet in a random direction. It seems to me anything can get hacked/any individual can be pressed, and it's alsways just a matter of if it's worth the trouble. Seems a hack like this is would be worth some, to some groups. Also: they only need to succeed Once. The rest of us need to have perfect security among all fleets of all corporations and brands for all time.
If you can call for a driverless Taxii then there is some level of remote access. So, it's going to be a concern in either case. However, you can set the car to dial out when it's confused rather than waiting for remote connections which reduces the attack surface.
The page about technology says, "Our sensors observe that the cyclist has extended their left arm. Detecting the cyclist's hand signal, our software predicts that the cyclist will move to the left side of the lane." This suggests to me that they're trying to deal with situations like the ones you've brought up.
So far the focus was primarily on what it would take to produce an autonomous car, which is primarily a technological challenge. But what I see as a harder problem to solve is the one imposed on all humans participating in traffic: we will have to strictly adhere and be very self-disciplined in traffic, in order to make sure that the non-humans get the correct message.
Imagine I am a cyclist and I don't signal but I swiftly swerve to the right. Unless the self-driving car is already at a slow speed, it would impose an undue risk.
Or, it is hard to expect that humans will rigorously adhere to this traffic discipline.
This is why these cars will drive slow for the foreseeable future and will frustrate a good chunk of the drivers.
Unless it gets dedicated lanes, the way bicycles do.
Not really, they're correctly hyped. The concept is revolutionary.
> They've been just around the corner for the last 10 year
B-S. Holy revisionism BS!!
We were NOT talking about self-driving cars in 2006! There was no company seriously designing fully autonomous cars in 2006! It was fully in the realm of science fiction, no investment, no major players, nothing.
And you make it sound like it was supposed to be released 10 years ago! Vaporware! They told us we'd have AI cars in 2006! It never came! <---- False reality that never occurred.
Why do you feel the need to lie about recent history? Shouldn't you observe that if you have to make a nasty lie to support your argument, maybe your argument is wrong?
You're definitely parroting the contrarian point here, but you're being very low-effort about it. Contrarian can be cool but please try harder.
We absolutely were talking about self-driving cars in 2006, when no less a "major player" than the US Government had already been investing in this "science fiction" idea for two years, in the form of the DARPA Grand Challenge races. The 2007 race even required compliance with traffic laws in order to win. Google's autonomous vehicle program is a direct result of this project - that's why they hired Sebastian Thrun, whose team won the 2005 race.
I don't know how serious it was, but part of the initial PR about the Grand Challenge included the goal that it would enable the US military to convert some significant fraction of their fleet of ground vehicles to autonomous operation within ten years.
I do think the previous poster was overstating their case a little, but not so badly as you seem to believe.
Wasn't there a breakthrough year with DARPA challenge? It seemed like for a while the results seemed to indicate that autonomous vehicles were basically impossible, then suddenly a few vehicles just killed it one year. I could be wrong, but seems like that threshold may have been less than ten years ago.
Hey I just wanted to remind you about the Principle of Charity[0] as well as HN's guidelines for commenting[1]. It's easy to forget when you feel strongly about something but that's a real person you're replying to and they are probably acting in good faith. We should be civil on HN even when we have strong opinions.
Caltrans and UC Berkeley were very seriously working on technology to have cars fully autonomously run on freeways. In fact my father was involved in the project some odd 20 years ago. The difference is that today you have more advanced technology, more private interests, less government interests and more media involved.
> In 2006, Volkswagen showed off a self-driving version of its Gold car, named "53 plus 1" in honor of the original #53, Herbie. Autoblog.com posted a video of the car in action.
Should I go further back in history? Because we can.
Yes, the concept is revolutionary. But so is matter teleportation, and that's not going on sale next month either.
What do we call technologies that have been hyped for 10+ years? This is not just a knee-jerk contrarian point of view. Vaporware is actually the correct term we use for these products, not a slur.
It's not bullshit revisionism. People really have been talking about these things forever but not delivering them.
> Why do you feel the need to lie about recent history? Shouldn't you observe that if you have to make a nasty lie to support your argument, maybe your argument is wrong?
Settle down, nobody's doing any nasty lying, see the link above, this discussion has been going on longer than 10 years.
If you use autonomous miles driven as a metric, there is actual progress happening. That metric is increasing exponentially btw. This tells you nothing about time to market, but it does tell you that the time to market is a definite (but maybe unknowable number).
In contrast, fission reactors have been around the corner for a long time, however, there if you use the metric of net power generated from fission as a metric, no real progress is being made there.
I see a real difference between these two "always around the corner" technologies.
I would encourage humility when considering whether to make a claim about a field that you are not well informed of. Real progress is being made, by a number of different groups, in a number of different directions.
Absolutely false.
A bit more than 5 years ago I thought quite seriously to invest on self driving cars because I thought it would be viable in the not so far future and it seemed that no one else was pushing for it.
10 years ago there were only research projects with abysmal performance compared to what today seems not too far from the market.
Any verifiable source for your claims?
I get that we have to be open with tech. After all most tech seemed "impossible" "unrealistic" and "unfeasible" even a couple of years before they go mainstream.
But you have touched some use cases that are more frequent than most people think. London is a great example. My hometown Athens is much more closer to London as far as driving is concerned as opposed to LA or NY. I guess many other metropolis also.
How will AI brand A communicate with AI brand B about life critical decisions ? Hey we can either both crash to each other or kill that kid on the pavement. ( just an example , I don't want to go to the morality of the situation)
I assume there is going to be a "standard " . But have a look at your everyday tech. Standards are not easy. There are still WiFi routers/repeaters that cannot talk to each other nicely, just to give an example.
Usually in life critical systems you don't have interoperability issues because there is no interoperability. Huge fail-safe systems from a single vendor.
I can see communication with other autonomous cars happening, but with humans is another question altogether. We enter the realms of 'general ai' here, which is way out of scope of our ML tools.
You don't use your general AI to see someone extending their hand in traffic ex. General AI is the ability to both drive car, score a girl, cook food, clean your room etc.
I think you would be surprised how little general AI have to do with this.
Sure, not Turing test level AI, but the car needs not to just correctly interpret what the other human driver is communicating, it too needs to communicate back, sometimes assertively. Sometimes you need to actually talk to the other driver.
Anecdote: I was driving down a long, narrow country road, lots of turns, space only for 1 car, land rover coming in opposite direction, and another car behind me. We had to verbally negotiate who reverses back 50m to let the others go past. I cannot think how an ML algo could resolve this situation, even if it could, how would it communicate it's intentions to the humans, when flashing your lights isn't enough?
I think a good way to think about the last 5% automation problem is to look at other examples where automation has replaced human work. In every case that I can think of, the work of the human was never 100% replaced. Automated check outs at the supermarket still require a number of attendants. An automated assembly line still needs factory workers. Automated rail systems are still controlled by humans in a control room. You can also look at how the environment has been adapted to better suit the new technology. For example to allow big heavy horseless carriages roll along roads we replaced dirt with tarmac (asphalt).
Driverless cars will never (until we have general AI) be able to fully replace a human driver. However we don't need to get to 100% for the technology to become useful. It might be that driverless cars will be restricted to only driving on certain mapped roads. Or every car's movements are tracked in a central control room and if a problem is detected either by the car or the passenger it could be controlled remotely. Or let's say a GPS sensor breaks an actual human driver might be issued to come and rescue you. Lastly we might even accept a certain failure rate. Society is willing to tolerate failure of technology to a certain degree in cars and planes today. All we need is is for the technology to make us all better off on average for it to be successful.
My scuba instructor shared this piece of wisdom: "You don't have to swim faster than the shark. You just need to swim faster than your buddy."
People get this stuff wrong all the time. Especially when they're drunk, tired or texting. These cars won't be perfect but I think pretty good and conservative under uncertainty will be better than human, drunk and overconfident.
Not to get personal, but I'm getting tired of seeing this argument.
Why are people looking at the extremes? Only human drivers or a completely self driven car? What about a driver assistance system, that corrects a Human's stupid mistakes and still lets the human drive. I'm sure this hybrid (which is the current industry) is safer than both your extremes
It has be better than the best human driver or they'll get sued out of existance. You can't sell a product that you know will hurt or kill a certain number of people.
What happened to worst case analysis? Don't we all know the flaws of best case?
I think (and I could be wrong) that there are many more cases where conservative driving is safer than aggressive driving.
In a lot of the cases mentioned, doing nothing or stopping are the safest choices. What's safe to drive over? Just assume everything is unsafe and only drive over it if it puts others in danger to avoid it.
If the road is blocked and you need to communicate to navigate, don't. Wait.
If the road conditions are changed due to construction in ways the car can't understand, stop. Find a new route.
Most humans wouldn't do these things. The nice thing about the autonomous car is that it can drive extremely cautiously.
Sorry no. Cars kill thousands of people every year, and will continue to do so. Car companies are doing just fine. Power companies literally kill millions every year, a-ok. McDonalds, Coca Cola ? ditto.
Drivers kill, not the cars. That's totally different from an autonomous car killing someone. Humans make mistakes. That's different from a defective product.
> Traffic lights will send out the same messages they send to humans by red/green lights to the cars using radio.
I live in the capital city of an EU country, population 2 million, and there's a traffic light about two blocks away from where I live which has not been functional for about 2 weeks now. And the bad thing is that the other traffic lights from that intersection DO work, so you cannot rely on the road-signs, like one usually does when ALL traffic light from an intersection malfunction (from loss of power, mostly). So I'm a little skeptic that traffic lights will be able to "reliably" guide no-driver vehicles using radio signals anytime soon.
> Traffic lights will send out the same messages they send to humans by red/green lights to the cars using radio.
Most of the traffic lights in Las Vegas already support vehicle-to-infrastructure (V2I) communication, although you need an Audi A4, Q7 or Allroad manufactured after 1st June 2016 and an Audi Connect Prime subscription ($33.33/month) to make use of it [1].
Yes, I think you're right. We'd need to update a lot of our infrastructure to make it friendlier for these cars.
I should also add:
- very fast mobile data, with no blackspots, so a remote driver can take over a car that's confused by a situation
- a replacement for GPS, it's not reliable enough, the signal is far too weak
- digital signal beacons embedded in lane markings, and edges of the road, and on traffic cones for roadworks
Why should local municipalities foot the bill for adding beacons to stop lights, etc? I get trying to make infrastructure friendly towards self driving cars, but the general public should not be footing the bill for someone's life choice to buy such a vehicle.
Because if we do resolve these issues, our roads will be much, much safer with autonomous cars. 2k people die on the roads each year in the UK, multiples more sustain life long injuries. I was in an accident, which could have been prevented were it not for the 1 second it takes for a human to react to a situation.
Right. Just like we paint roads to tell humans (and camera driven cars) where the lanes are, we should have digital markings to tell self driving cars where the road is directly.
The remote takeover seems entirely separate, and both promising and terrifying.
A lot of these problems can be solved by simply making them fly.
... that's right, now you know what the driving force behind all those flying cars in sci-fi was. It's simply a way to separate the autonomous and old-style vehicles.
This seems to have been said in jest but the main thing keeping flying cars in sci-fi and out of the real world is the human element. Most people have a hard enough time driving in two dimensions. Adding a third would be chaos without the aid of our robot overlords.
> This seems to have been said in jest but the main thing keeping flying cars in sci-fi and out of the real world is the human element.
I thought it was that winged flight makes for crappy cars and wingless, vectored thrust flight is horribly fuel hungry. (Sci-fi versions are usually wingless and either have magic drive systems or just ignore the fuel issue.)
I would also add; that I will stay consider these vehicles to be unsafe until the manufacturers can prove to be beyond a shadow of doubt that I or the local computer are the ones piloting the vehicle.
When a random person can remotely take control of a vehicle; whether it be the windshield wipers, or full control of the vehicle, then it is not safe to contain a power source in the vehicle near a public road. I don't want 4chan driving me into another vehicle.
I'd consider autonomous cars to be more safe than a car I drive when statistically I'm less likely to get hurt or die.
If there was some chance my car could get hacked, or other weird vulnerabilities that robocars had that traditional cars did not, I would still want to be in a robocar if my chance of dying was cut in half.
Sure. But you are assuming a good way to estimate that chance. Say you base it on previous performance: in the last 5 years, twice as many people have died driving non-self-driving cars compared to those riding self-driving cars. Then, the very next day, a massive hack / cyberattack happens that causes 10 million fatal accidents...
Getting targeted for a hack isn't a purely random thing though.
It may be true that the chance for an average person to get targeted multiplied by the chance for a targeted person to get their car hacked is lower than the chance of dying while driving yourself, but that's not reassuring at all if you think you have a higher than normal chance of being targeted or are already being targeted. From experience of being on the receiving end, it's really not that hard to piss off 4chan or get unwanted attention from other crazed parts of the internet. (For an extreme case, consider a certain pizza place in the news lately...)
This is going to be one of those things where a lot of people have trouble quantifying the actual risk because the fear when things are out of your control is so powerful. Like flying.
Not to mention, the live hacks of moving vehicles have already occurred multiple times. Thankfully, there are not yet enough skiddie tools and the bar to entry is just high enough that mass casualties have not yet occurred. Give it a few more months or maybe even another year before your car is just another IoT that can be controlled remotely. I do not believe it would be irrational to consider the risk of 100k cars driving into oncoming traffic at the same time. That is a lot more damage and chaos than a few little 747's could do. Am I incorrect?
My only questions are
1) will the skiddie tools be on github first, or will someone have to buy/leak them and repost on github?
2) If a minor executes the scripts and lives are taken; and assuming you can determine who ran them, will they be tried as adults and will the verdict or punitive actions make up for the damages?
His point is that the remaining 5% of a problem can take considerably longer to solve than might initially be anticipated by looking at the quick progress made on the first 95% of it. A textbook example of that is voice recognition.
It's also not particularly clear that the costs will drop rapidly to zero or even to a level that leads to mass adoption (see air travel at Mach 1+, Segways...).
Your examples aren't very good, sorry. We're talking about adding AI, autopilot, autonomous capabilities, whatever you want to call it to already existing hardware. It's mostly a matter of software and you're a living example that it can be done, in 20W, with nothing more than stereo cameras and mics.
If I had to compare, I expect the change to be more akin to what Excel has done to paper spreadsheets.
Current voice processing tech is still very much worse than an average human, no matter how fancy the microphones and how much power it's allowed to consume from the grid.
Good sensors aren't cheap, the optical and mechanical elements aren't likely to drop in price soon. Only the computing power is getting cheaper, but harnessing it isn't particularly easy (we're not getting better at programming as quickly as we're making faster chips).
If you need another example, consider automated cleaning robots. It's a considerably simpler problem to solve, and they're still only a marginal product. At the moment, autonomous tech is only really used in well-controlled industrial environments. It will surely break out of there progressively, but chances are it will take a while.
we don't disagree in anything. i'm just saying that it will happen, it's a matter of time - 5 years, 10 years, maybe 20 years, but it's going to happen. there's thousands of smart people working on it, millions if not billions already invested and tens of billions at stake.
You're oversimplifying it and being a little pessimistic. Why? Because there's an order of magnitude or more difference in the complexity of AI of cars and AI that a human uses to navigate
That people are talking as if it's just round the corner, that taxi drivers and truckers are about to become obsolete. Whereas I understand the technical challenges are much greater than is stated, it may even be they won't be all overcome.
It's not that taxi drivers and truckers are going to become obsolete tomorrow. It's that that happening is such a huge cultural and economic change that we need to start planning for it today if we're going to not destroy the livelihood of millions of people when those industries do become obsolete in five or ten years.
Tesla is launching Autopilot 2.0 (in stages) starting in January.
It's still NHTSA level 3, not 4 or 5, but it's a major step up from pretty much anything currently available and will be in consumer's hands. Even more so when the Model 3 comes.
I think you'll be surprised how fast this technology will exit the labs. Just that Google has decided to brand around it is a big indicator that they're growing confident.
But so what? Full autonomy needn't even be a business goal of Waymo. They just want full deployment. Having a system in place which reduces or eliminates many/most of the accident root causes would be worthwhile to install on any vehicle on the road even if it technically had a human driver.
The "goal", such as it is, is to have a technology worth using. It's not to transcribe whichever comic book utopia someone has in mind.
The point is that we are far far further away than commonly accepted. The day a computer can distinguish between a plastic bag in the road and a child is the day I'd support them on the road.
Perfection is of course, impossible, the best we can hope for is progress asymptotically close to ideal. The minimum bar should be performance surpassing human performance, but the details do matter. For example, a 99.99% safety rated car which almost never crashes but randomly explodes once every 10,000,000 uses would probably not be accepted.
My mother once almost killed a child who ran into the road. The kid was on the side of the road and suddenly ran into the street. She did not see him, and he ran into some sharp piece of metal on the car (this was in the early 60s). He bled profusely, and the neighborhood people came out and almost became violent with her, blaming her for the incident.
If she had been in an automated car, it would have almost certainly have stopped before the collision.
And what's wrong with slowing/stopping for a plastic bag? How often do plastic bags really sit in the street shaped like babies?
I would shorten that to "other humans", to specifically include pedestrians. That 1/2-second eye contact between driver and human is key. Would anyone here dare walk in front of a car without that little nod from the driver?
The other edge case, one that is never discussed, is how these things expect to deal with off-road driving. I regularly use car ferries. Getting on and off those, and negotiating the terminals, stumps many human drives.
These issues could mostly be solved by the car having a "remote control" where an expert in a remote immersive environment could take manual control of the car for a few seconds to resolve whatever the problem is.
Now you've reduced all your problems down to simply "identify problem and alert a remote human".
If each of those problems only happen once per day, and they are resolved in 30 seconds, it's going to cost only about $30 per car per year to have that remote expert on hand.
That's an option, but London has loads of signal black holes, or the speed is just not good enough. Infinitely worse in the country, if we want autonomous lorries.
I think we'd need a huge infrastructure upgrade to make life easier for autonomous cars. It'd have to be >100 Mb/s mobile data, with extremely good coverage (for the recovery mechanism you mention). Think GPS would need to be replaced too, the signal is way too weak and unreliable. We'd have to embed digital markers/emitters alongside lane markings, traffic lights, signs etc, so they're easier to deal with.
No need for high speed - all these issues can be resolved by inching forward or back to get the car back onto the open road. (although obviously high speed is a bonus)
Connectivity you see is from a single mobile phone. A car could have a better antenna, service from all the mobile networks rather than just one, wifi fallback and a backup satellite uplink. If all those fail, you dispatch an engineer to find the car wherever it's stuck and add a solar powered LTE repeater.
If you have a map of the whole service area and know where you started and have a camera looking out the window the whole time, there is no need for GPS really.
It might even be the sort of thing where you tap the video of the street ahead to tag the bag as passable and the car then slowly drives through it. No remote guidance per-se.
Considering these are car companies doing this, true remote control is not a good idea. The industry response to the car hacks was pathetic. Only Tesla reacted properly. (And probably Google...)
I'd imagine all the GPS data could be loaded into the system beforehand and updated when there is a consistent signal. The same way cars now deal with GPS.
Humans generally look at it, make an informed guess and drive accordingly. The recent Tesla accident is a great example of how even the best optics can be blown out by environmental factors like sun, whereas a human would likely readjust their vantage point, lower the sunshade, put on sunglasses or even pull over if they are unable to cope.
Comparatively, the car did not attempt to do any of this and actually didn't even realize its cameras were in a state of failure.
Which says a lot about relying specifically on optics. Readjusting vantage points from an autonomous vehicle is as simple as having multiple sensors and deciding to rely more on a particular sensor in particular conditions. In particular, Google has some of the most expensive arrays of sensors.
What you're missing is "how do humans cope with these scenarios?" -- by either cautiously slowing down or by following those humans ahead or both. Machines can do the same
All these are old "problems", either solved or in progress.
Something I've never seen adressed is how SDV fleets behave when change occurs. Are there chaotic oscillation patterns that emerge from AI trying similar avoiding behavior.
There's research in massive multi agent behavior, but I've never seen it tested on the field with SDV makers actual models.
> They need to reliably deal with an almost infinite number of edge and corner cases, each quite different from the last.
It doesn't need to be perfect. It just needs to be better than average human drivers. Sure there will be cases it can't handle. But the same is much more true for human.
> “We’re now an independent company within the Alphabet umbrella,”
> Google is currently equipping a fleet of 100 hybrid Chrysler minivans with its sensors and computing gear that will soon join its nearly 60 prototype autonomous vehicles. The company hasn't yet disclosed when and how it will begin generating revenue from its efforts and Krafcik declined to discuss specific business plans today.
I attended one of Google's talks on this. The rationale is that humans can't be trusted to take control in a timely manner for that 0.01%. Their internal testing with an L4 system on trained Googlers (who were told to remain alert in the event that action needed to be taken) had instances of testers sleeping, eating, putting on makeup, etc. They concluded that if Googlers with explicit training couldn't behave themselves, then L4 definitely wasn't sufficient for the general public; autonomous driving had to be all or nothing.
I still don't understand why every company doing self driving cars is focusing on consumer cars while no major player is doing interstate trucking. Interstate trucking could pretty much be done now, and has several advantages. Driving on interstate freeways is orders of magnitude easier than driving in a city with pedestrians, bikes, cars parked in the road, etc. And there is a great monetization scheme--no driver means you can get it to its destination more quickly and more cheaply.
(The thought is you'd hand off to a real driver once you get in to a city)
We'll see commercial self driving trucks before we see consumer self driving cars. Self driving trucks will first connect depots outside of cities, with humans solving the local driving problem, but eventually, they'll just completely drive themselves.
Because it's all marketing bullshit. In fact that's a good indication that it's marketing bullshit - the fiction that the tip of the self-driving car spear will be consumer vehicles.
Admittedly pre-orders are essentially marketing, but marketing of an actual product with intent to deliver, I'm not convinced it's "marketing bullshit".
Trains don't even need AI. It's the lack of good incentive (the required capital investment vs. potential for future savings) and angry drivers' unions that are stopping innovation there, I'm sure.
I am not so sure. It could also be the Christmas present Larry got for himself. At least more affordable than a rocket company.
I guess it is a mix of both. The SV elite pretending they can do another trick and keeping themselves busy and entertained and Google/ABC pretending they are a growth company (beyond search ads). Why should they bother with worldly billion dollar markets such as office communication equipment (which Microsoft is successfully attacking as we learned earlier today).
The industry will change/disappear overnight. Trucking will be 4x cheaper with 2x the transportation capacity when you remove the trucker. Doesn't matter how much unionized, etc. the industry is now. No group survives that.
I assumed all trucks stop at a truck stop and the driver sleeps while another driver continues the journey? That way the valuable truck and cargo never stop moving.
In the US, sometimes truckers drive in teams and switch off, but I've never heard of them not sticking with their trucks. Trucks frequently sit idle while the driver sleeps.
True, the controls challenges to a self driving truck may be easier than a personal car, but the stakes in the event of an accident are an order of magnitude greater in an autonomous truck. As proof, let's napkin-math some kinetic energies using 0.5mv^2.
If we are riding along in the current heaviest production car, the Rolls Royce Phantom Extended-wheelbase[1], which weighs in at 6000 pounds, and we are moving at 70 miles per hour, we have a kinetic energy of 1.33 Megajoules. [2]
Now, let's hop in the heaviest federally allowed tractor-trailer, which weighs 80,000 pounds, and get ourselves to 60 mph, a more reasonable speed for such a massive object. This comes to 13 Megajoules. [3]
The upshot is, one fully loaded truck moving at 60mph could cause the same amount of damage as 10 passenger cars all moving at 70mph. This means that as a firm, your liability is 10 times greater per vehicle. Insurance is 10 times more expensive. Regulation is 10 times as strict. At that level of risk, it'd be stupid to not put a human in the cab of your autonomous truck right? Just in case, and to shield yourselves from legal issues. But wait, why not just get him to drive and save yourself a lot of R&D money?
I work in the commercial trucking industry and real-world challenges mean this self-driving equation is not as lucrative as people believe. And we haven't even touched the Jobs Problem. Hence the focus on passenger vehicles. I do think autonomous trucking will become popular eventually, but this industry moves at a glacial pace for a reason.
Amazon will be willing to take the risk to avoid the trucking unions. FedEx is pretty forward thinking and might also be willing to take the risk since they self insure. Same with UPS.
Once those folks prove that it works and it's safer than a human, the other companies will quickly fall as the insurance rates become lower than a human driven truck.
A self driving truck doesn't have to go 60mph to be as cost efficient as a human driven truck. They could be governed at 45 mph and still come out way ahead and cause far less damage if something does go wrong.
So now you have a greater disparity of vehicle speeds on a typical highway. Having just returned from driving on highways cutting through mountain ranges, I'm acutely aware of the "interesting" driving that results from having vehicles traveling at very different speeds on long uphill grades.
I think that we'll see a ton of companies going after this space, but they'll be as quiet as they can. Why? Jobs. You are directly competing with humans in this space. Truck driver is one of the most common jobs in many states[1].
"Driving on interstate freeways is orders of magnitude easier than driving in a city with pedestrians, bikes, cars parked in the road, etc."
Yes, it's easy and safe if competing traffic is only fellow truckers or other level-headed, observant drivers.
It's when you mix in the distracted, newbie, or skittish drivers that things get difficult. Long-haul truckers (the majority, the good ones) develop a sixth sense about such risks; I can't imagine thoroughly coding a software replacement.
About the sixth sense - does this not also apply to good doctors? And yet, they are easily outdiagnosed by many ML implementations. I can imagine deep learning to also gain this sixth sense. E.g. a cyclist in front of you looks to the left - a proficient driver predicts that cyclist might turn left very soon. Why should a self driving car/truck not learn the same?
In terms of execution, I agree. Decision making is where human experience wins. I think an experienced human driver will better predict the irrational things another human driver might do.
If you solve the consumer car problem, you own personal transportation. That's why. Besides, your premise is untrue anyway. Otto is the autonomous truck thing. Rolls Royce is the drone ship.
While the benefits you listed are correct, you're missing the biggest one: relative cost of the sensor array. High-fidelity LIDAR tech is cost prohibitive to include in general consumer vehicles (hence taxi as a service), but it's much more practical for trucks which typically run in the hundreds of thousands of dollars.
Ie. you're gonna have to get new laws in place allowing empty vehicles in not just one, but many states.
Current driverless carpool plans might well only allow use when someone is in the vehicle, and they will likley only operate in a single state. The people-transport industry is also worth more than the goods transport industry.
It’s going to be interesting watching how local police departments and municipalities handle the widespread adoption of autonomous vehicles.
In my town (a suburban town with a low crime rate), the police spend the majority of their time enforcing traffic laws. The municipal courthouse is always filled with people, and if I had to guess, I’d say that 80% of the people are there as a result of a traffic violation, while the other 20% are there for drug offenses/other.
It seems like autonomous cars would lead to a drop in traffic offenses and thus revenue. Even if only a tiny portion of the town’s revenue actually comes from traffic violations (when compared to local property tax), there would probably be a lot more idle time for police.
I suspect that they will first start to raise our property taxes to compensate for the lost revenue, but I would think that the long term effect would be a reduction of municipal workers/police.
Of course, this doesn’t apply to other areas like Philadelphia, where police spend only a small percentage of their time enforcing traffic laws.
Of course, the flip side to this is the notion that autonomous vehicles will free up the existing police force to do more important things. In most cities there is a distinct lack of officers on the street, usually due to budget constraints. Now they can take all of those officers used for monitoring roads and highways and have them do other, more important things; all without increasing the number of officers or budget.
Anecdotally, one of Google's autonomous cars was pulled over in Mountain View, CA a while back[1]. The car was going 24 MPH in a 35 and the officer didn't realize the car was driving itself.
At least they could have given a more up to date report on what they are doing. 7 years passed since 2009 and all we've seen is a couple of promo videos with scant actual information. We have no idea how they compare to Tesla or other self driving car startups.
Tesla has lane-assist, adaptive cruise control and a few other goodies. Many auto manufacturers have those things. You still have to manually follow GPS directions.
Google/Waymo is actually piloting a car from origin to destination without a steering wheel or pedals. That's a much more difficult challenge.
Put another way, no blind person is going to sit behind the wheel of a Tesla and tell the car to take him/her to the supermarket. Google's car actually does this, on real roads, today... and that's amazing.
> Put another way, no blind person is going to sit behind the wheel of a Tesla and tell the car to take him/her to the supermarket.
That's exactly what Tesla is marketing and working on with their new self-driving features, and you can buy cars that will allegedly be capable of this today.
"All you will need to do is get in and tell your car where to go. If you don’t say anything, the car will look at your calendar and take you there as the assumed destination or just home if nothing is on the calendar. Your Tesla will figure out the optimal route, navigate urban streets (even without lane markings), manage complex intersections with traffic lights, stop signs and roundabouts, and handle densely packed freeways with cars moving at high speed. When you arrive at your destination, simply step out at the entrance and your car will enter park seek mode, automatically search for a spot and park itself. A tap on your phone summons it back to you."
>Google's car actually does this, on real roads, today.
It should be obvious that Tesla is going to win this race. Tesla already does fully autonomous driving, they released that video in November. And they build real, production cars. "Real artists ship", as Steve Jobs said.
If you a manufacturing ace, would you join the company with a long list of failed projects and no actual production cars, or would you join the company that is building 500,000 of the most advanced autos right now?
If I was a hardware expert, would I join Google who shies at giving a fraction of it's cash to hard tech, or Tesla who buys german automation companies?
Or plain old software. Tesla has magnitudes more data and can update neural networks to hundreds of thousands of cars at a whim. Soon millions.
I'm not super convinced by that video. That drive had almost no challenges/obstacles. No bicycles, no construction, basically no pedestrians, no other vehicles behaving erratically.
Google has cars that can respond to everything from a cyclist making hand signals, to a school bus (which must not be passed), to a police car pulling it over, to a woman chasing a duck across the road: https://youtu.be/tiwVMrTLUWg?t=8m49s
I'm also not sure this is a simple case of "throw data and hardware at the problem." If it were that easy, it wouldn't have taken Google so long to get to where they are today.
Again, not to take away by the amazing feat Tesla has achieved, but Google has had similar videos going back 2-3 years, and they are still not ready to release.
So either the problem in practice is much harder than a simple video can show, or something else is up.
Though, I think the real issue is that google wants to go straight for L5 (meaning 100% automated, and you can remove the steering wheel). As compared to an L4, which is more like 99.9% automated, but you still need steering wheel for those rare edge cases.
If Tesla executes perfectly they might stand a chance. They don't have much margin of error though. One lost lawsuit because of an accident caused by a bug in their software, and they are out of business...
I find that notion very frustrating. 30,000 people die on the road in the US every year and no one seems to bat an eye, a self driving vehicle has one and it's the end of the world. The expectations seems rather unrealistic and the media seems to love creating a controversy.
On their webpage, waymo breaks down that 30k into smaller pieces. 94% of those deaths were due to human choice or human error. At least 71% were due to speeding, alcohol, distraction, and drowsiness. It's probably safe to say that driving like those 94% during your DMV driving test would mean a human wouldn't even get licensed. Why lower the bar to just being better than them? I want my autonomous car to be better than a good human driver.
The trouble with just throwing the statistics around and saying "it's fewer deaths!" is that it has to be fewer deaths in comparable situations. A tesla having fewer deaths per mile in perfect driving conditions shouldn't be compared to a possibly-drunk person in possibly-awful weather. This isn't the media creating controversy, it's people expressing skepticism when a corporation's incentives are to let a couple people die, and trying to maintain a high bar.
But they should be compared. As you indicated and the statistics show the death primarily happen when people are impaired (drugs and alcohol, fatigue, distractions etc), the machines substitute these human errors by fewer and further apart potential engineering faults. With proper Root Cause Analysis of the investigations (helped with all the data collected of all the accidents), overtime fewer and fewer would be expected.
If you can go from 30000 to 1000 just by changing to good enough autonomous cars that's a change worth doing. Then you can look at improving it to 0 and encourage that process if not through insurance liability processes through regulations that require ongoing improvement in autonomous safety.
As rtx said, the difference is choice. We could save a ton of lives where people choose to put themselves in danger, but we'd save them by killing a smaller number of people who choose not to put themselves in danger. I don't think that's ok.
Sure, but if the average person really would be safer with a car that self-drives 90% of the time, wouldn't you recommend that everyone get a self driving car? You can't say "Only get a self-driving car if you're a below average driver" because then nobody would get one (and they'd be worse off for it)
I don't think there's a clear answer to that question, but my gut says that would be immoral. What we'd be doing is saving a ton of lives where people would have otherwise gotten themselves killed through their own choices, but at the expense of a few deaths where the fault is entirely our own (the car maker). It's laudable to protect people from themselves when it doesn't otherwise affect them, but when the cost is killing totally innocent responsible people, I think it crosses a line.
On the other hand, if you think fault doesn't matter and it's just one life for another, then it essentially becomes the trolley problem, which doesn't have a clear answer either.
Well, in the former case, that is the fault of the driver (or another driver). In the latter case, it would be the fault of the manufacturer. It's not even remotely the same thing. Remember those out-of-control Priuses, where the manufacturers were at fault. Toyota got plenty of backlash for it.
I find this an interesting question. With the exception of OTC and prescription drugs (for which there are still lawsuits) I have trouble coming up with other consumer products when used as directed and properly maintained are still directly involved in killing people because... well, stuff happens. This isn't an argument against using autopilots but it is an issue that hasn't been really addressed so far.
The issue is Google is doing it with (prohibitively) expensive hardware, so we don't know yet if Google's self-driving system works with cheaper Autopilot 2.0-level hardware that the other car makers are going to use as well.
"The person in the driver's seat is just there for legal reasons" -- but note just how closely that person's hands are to the wheel, in an obvious state of readiness. That person is 100% ready to take over the car if it messes up.
Tesla had the luxury of running test drives until they had a success. This is a different thing than putting a car on the road with no steering wheel and saying bon voyage. The first is a demonstration of a single successful instance (which is probably repeatable on average); the second is a much stronger demonstration of a > 99% success rate.
Googles car only does 25 mph. They might have made that video when they only have a 95% success rate, because in the 5% chance of collision it would be at low speed, with a very light vehicle, and very unlikely to cause a fatality.
I have friends that live in one of the neighborhoods they're testing in, here in Austin (Mueller, if you're curious). And they've said the vehicles have been performing well, with no issues at intersections or dealing with pedestrians or cyclists. It's a new-urbanism style neighborhood, with the houses positioned close to the streets, with lots of on-street parking. I imagine it's a challenging environment for the car.
Yep! Plenty of pedestrians with strollers/dogs/bikes/etc, vehicles stopping in moving lanes for pickups or dropoffs, a mixture of stop signs and traffic lights, one-ways, two lanes, four lanes, divided roads, and the cherry on top: a two lane roundabout, which endlessly baffles drivers here.
I'm not sure there's a better spot in Austin to test autonomous cars.
That said, they aren't perfect.
1. They wait just long enough at stop signs when it is their turn, which confuses other drivers.
2. They drive well under the limit, causing small jams at peak congestion.
I live almost in the backyard of their facility in Silicon Valley. No issues here either, the vehicles are very well behaved. Of course, how much is that an indicator of their progress, I don't know.
Tesla will have 100,000 prototypes being trained in shadow mode for a billion miles per year across the globe within a year. Tesla will compare quite favorably
One can scale up the first 99% of an autonomous driving os pretty fast just using pattern recognition, but the devil is in the edge cases.
We haven't heard much from Google because they've been programming and validating all the unglamorous little problems that training data and a neural net can't solve alone. 4 way stops, passing motorcyclists and stuff like that, where old fashioned hand coded features are necessary to adress situations in which specific kinds of reasoning is necessary.
Tesla has to solve all this boring stuff as well before they can offer comparable capabilities in an AV.
I don't believe your hand coded features will outperform ML trained over a billion miles, even in the long tail rare events.
Sure ML is bad at rare events, but hand coded features also map poorly to such rare events, because it's unlikely you'll be able to imagine them all, or even manage to handle categories of rare events well.
Furthermore, I believe it would be possible to train an ML based rare-event detector, which simply detects the difference between "regular driving I'm familiar with" and "something funny is going on, I haven't seen this much in the training dataset", and then refer to a remote human to resolve the issue.
> and then refer to a remote human to resolve the issue
Successful remote control is a whole different project, very different from self-driving. In order to give the remote pilot enough situational awareness, you need plenty of streamed imagery (recreated or recorded). And you need a feedback loop with a very small latency. Not saying it couldn't be done, but it won't happen "organically" by cars becoming self-driving.
That's the benefit of having 100,000 cars with all the sensors but humans performing all of the edge cases. Tesla can take data from all the cases where there was a deviation between what the car calculated it should do and what the driver did.
I don't think hand coded features are necessary. No one hand codes humans to drive.
And how many of those miles are in conditions like:
* that road with the one-foot deep pothole (it's more like a sinkhole at that point, though) in the middle of a lane
* the perpetual construction zone where there are five layers of lane markers, only one of which is the actual layer of lane markers
* the 10% grade two-lane winding road in the dark with 2 inches of snow on the road and more coming
* open highway with a constant 30mph crosswind
* dense fog with less than 100m visibility
* torrential downpour with so little visibility you can't even see the lane markings on the road right in front of you
Those are all conditions I've driven in, and I can think of several other horrible driving conditions that I haven't yet had the (mis)fortune to drive in. Just because you can drive well in good conditions doesn't mean you can drive well when things get tougher--as the driving records of many human beings can well attest.
There are some conditions a human shouldn't be driving in. For example a hurricane or while intoxicated.
We can likewise say there are some situations a machine shouldn't be driving.
As long as the machine can identify when it is incapable as a driver, it's only the same as the human pulling over for a nap because they can't stay alert.
Sure, if you ran a taxi service with such a system you'd have to send a regular taxi to pick up the customer, but it's only the same as any other kind of breakdown like a puncture.
This is why Tesla's strategy is better than Google's. Tesla will get the data on those situations, because their data comes from every driver of their cars. Google gets data on Mountain View and downtown Austin.
That's a super conservative prediction which has zero growth from current sales numbers in it. Tesla grows has historically grown at about 50% per year.
Not talking about the number of cars sold, talking about that translating into self-driving capabilities. And in that sense, it is a very aggressive prediction, given that Google has had fully-autonomous, no-backup-driver, public-road-testing for over a year.
So you can say that Tesla is at least 1-years behind. More like 4-5.
Tesla is definitely behind Google now, but I'm saying their trajectory is orders of magnitude better, because they are collecting real world training data much faster. Humans show the minimal set of sensors needed for driving, and Tesla have a superset of that being delivered to customers today, which is being used to train their model. They get a lifetime of real world driving experience on those sensors every day.
Tesla models equipped with Autopilot enhancements do in fact have some form of emergency braking (automatic emergency breaking) to avoid rear-end collisions and primitive auto-steer.
Looks like that page got killed in the Waymo switch. Archive.org is missing the PDFs, but it looks like they still work if you follow the archived links straight to the URL.
Google's commitment to SDCs has always seemed half hearted so I wonder if this indicates that they are getting more serious about it. Moonshots at X have been getting the axe, the SDC stuff seems to be the only obviously viable one right now.
To the surprise of many folks (myself included) it has turned out that SDC tech is probably for existing manufacturers to develop versus software guys learning how to build cars without the massive supply chains needed to assemble 4000lb widgets. Before anyone points to Tesla, try getting inside a $140k Tesla and then a $140k Mercedes and it will be obvious what advantages there are to having manufacturing experience stretching decades.
Tesla recovers a lot more of the individual vehicle cost as future R&D than Mercedes. That's probably a good thing as they are a start-up and Mercedes is a company delivering a few orders of magnitude more cars. Don't think it's a valid comparison though to say that Tesla is putting the money into tech and mercedes isn't.
Rather, Mercedes has access to better credit terms and availability than Tesla because they produce 100x more cars per year. And because of that, they don't need to recoup future R&D money from current cars. Ergo when you buy an expensive mercedes, more money goes into the car you bought than future cars.
I think mostly it is the fit and finish. I've been in a 2016 BMW 750i and a 2013 Model S, for about a total of maybe 30 minutes in the BMW and 10 in the Tesla.
The interior of the 7 series seemed special to me. All of the buttons felt great, seats looked beautiful and were super comfortable to sit in. The control panel had many small OLED displays and they looked great. And small details like when I went to adjust the seat, by placing my fingers on the buttons, information about my seat configuration popped up on the center console, so I could see what I was changing. Also the rear passenger seats and rear windshield had power sunscreens.
Interesting that this very same day, the US Government is looking into proposals to require vehicle-to-vehicle communication in passenger cars to enhance safety.
I think we do know it would help. If your car can't physically see the car 1/4 mile ahead, but gets info that it has slammed on its brakes, thats very useful.
If you've ever driven in bumper to bumper traffic on a highway you know how invaluable this info would be, even to a human driver.
Plus info about road conditions (icy spots ahead!), etc..
Agreed. This is exactly why I hate driving behind SUVs. I like to be able to see the car in front of the car in front of me. It gives some additional reaction time.
I also feel like if you know where every car on the road is, you can also optimize the traffic light timing.
There is no reason to sit at a red light at 2:00 in the morning when there are no other vehicles, pedestrians or bicycles on the road. But the law currently requires it.
Chris Urmson, back when he headed Google's self driving car effort, said car to car communication was unnecessary. The car has to have good enough sensors and decision-making to deal with all the things that aren't communicating, from pedestrians to deer.
It could still be helpful, however. If your car gets a signal that there are an unusual amount of cars stopped half a mile ahead of it, it might be able to deduce that there has been an accident or something else stopping traffic and navigate away before it becomes an issue. Sort of like an automatic version of Waze
I agree that you shouldn't assume any information will necessarily improve performance, but I feel that, naively, inter-car communication would have tons of useful applications.
If anything else, during the transition to fully-automated, it would be wonderful if human drivers were able to visualize the intent of the automated vehicles around them. To know that a car is going to merge into my lane in 50m, or is attempting a U-turn, is so much more useful than simply seeing a turn signal.
Although this site doesn't really offer any substance, I'm glad that Google is making this project more public. Not only it nice to have updates on the project (I do hope they continue to update this site with their progress) but I think PR campaigns like this will go a long way to swaying popular opinion on self driving technology. The faster they can get the public on their side, the fewer regulatory hurdles they will face.
On days like today, when there are several inches of snow on the ground and more to come, I always wonder how autonomous vehicles will handle such situations. What happens when snow (or mud, etc) accumulates on the camera or sensors? I don't doubt that a computer can react quicker and with more precision in an "event", but what about when it simply can't "see"? I don't see a steering wheel in these videos, so I guess there is no manual failsafe?
Ford has been quietly testing in snow in Ann Arbor, MI.[1] Volvo tests in snow a lot. Volvo discovered that ice on the radars could be a problem. (So, de-icing heaters on the front of the sensors, like aircraft.) The big auto companies have tested their cars in bad conditions for many decades. They all have test tracks in awful locations and performance standards for bad weather. Even Google tests in Tahoe occasionally.
What's becoming clear is that the self-driving car industry will be the car industry. Autonomous driving equipment may come from suppliers, but they'll just be suppliers to the automakers. That's not a great place to be; the margins are low and suppliers are totally under the thumb of the automakers.
Volvo, which will deploy 100 self driving cars with actual customers in Gothenberg, Sweden in 2017, is way ahead on the user interface.[2] Volvo takes the firm position that, in autonomous mode, the driver is not required to pay attention at all, and if something goes wrong, it's Volvo's problem. They have redundant sensors, actuators, and computers.
Volvo is also way ahead on ads for self-driving.[3]
These cars will likely have many sensors of different types at overlapping locations, its not likely to keep going if all are completely unusable.
But I wonder about the snow though. How does the car know where to go when there are no identifying lane features? Maybe its not as hard of a problem as I believe it to be?
How does a human do it? It's extremely hard but recognizing landmarks like a guard rail or plow markers would be possible. Or maybe self driving cars will apply a sane approach and stay off the streets in bad weather unlike human piloted vehicles. (There is a whole other rant about letting employees work from on bad snow days)
I really don't get these questions. "How will the car drive in whiteout conditions where the road is frozen over and not at all visible?" Same way a (smart) human drives: it will pull over.
For substantially less crazy conditions, you can infer where the lanes would be by (a) looking at the spots of the road you can see, (b) prior knowledge of the road from experience, and (c) looking at oncoming cars and the flow of traffic, and driving in a way that doesn't surprise them. Though it's likely the case that you shouldn't be driving in these conditions anyway.
You must not live somewhere that gets a lot of snow :) For those of us that do, sometimes driving conditions are much less than ideal, even unsafe. Several times per winter I have to drive home from work on unplowed roads. As you put it, I have prior knowledge of the road (maybe precise GPS will help here?), and I can figure out where it is safe to drive based on tire tracks. I guess these will just be places that you probably won't have one of these vehicles (just as you wouldn't ride your motorcycle in such weather).
> Same way a (smart) human drives: it will pull over.
You mean carefully get to the nearest settlement as soon as possible? Some of these storms can last for days at a time. Stopping to wait it out in the middle of nowhere, especially if you don't have a full tank of fuel to keep the car warm for extended periods, is a pretty scary situation.
Hmmm. If the car decided, before your trip started, that it's too dangerous for anyone to drive where you want to go because of possible weather and environmental conditions mid-route, would that be useful or enraging?
I imagine a bit of both. Getting caught in a snow storm is a pretty horrible experience, and I would gladly stay home knowing I would have otherwise ended up in one. But at the same time the forecasting isn't all that great, so often the storms don't amount to anything, and you would have been perfectly fine to be out there. Sometimes the storms even come up without any warning (most of the snow we get – especially that which causes driving issues – is from the lake, not the clouds). Having to essentially hibernate in the winter because the car is always erring on the side of caution could become pretty enraging.
As haywire said, if you live where it snows, you drive regardless of whether or not you can see the road. I got a job to do, computers don't fix themselves, and my job is necessary even if the weather sucks outside.
While I think it's plausible to teach a self-driving car to drive in snow (like I do, generally following in the tracks of the cars ahead of me), I honestly think this is one of the best justifications that cars should continue to have steering wheels and pedals: Because automated snow driving is going to take a lot longer to become a solved problem than fair weather driving.
Presumably the cars won't continue driving if their systems are inoperable or unsure how to proceed. One could also argue that riding around in several inches of snow is a bad idea regardless of who's driving, human or computer.
That's a little concerning. Especially to the lee of the lake, snow storms can pop up out of nowhere and can last for days at a time. Usually, with care, a human driver can traverse to the nearest shelter to weather the storm. But if the car just stops out of the middle of nowhere, with limited energy to keep the cabin warm in the harsh cold, that seems like a recipe for disaster.
We're talking about mobility-as-a-service, so presumably the service would use other means to get you to your destination, like having the car towed, or sending out an SUV with a professional driver.
These kinds of extreme situations are scary, but that's what insurance is for.
Everyone is worried about how we get autonomous driving that's perfect in every situation, but that's not how automation works. You just have to automate the common case so a small number of humans can focus on the corner cases.
We are? I went up the comment chain again and still did not get that impression the second time around. What statement in particular gave you that idea?
> These kinds of extreme situations are scary, but that's what insurance is for.
Like... life insurance?
> You just have to automate the common case so a small number of humans can focus on the corner cases.
If I'm going to be sitting in the car anyway, I can drive it in those conditions, assuming the vehicle allows me to. However, Google has stated before that it is unlikely that they will release a car that needs to be taken over by human drivers. It's all or nothing for them.
It's a fair question, but it doesn't seem unsolvable. And humans have huge issues driving in snow themselves, so by comparison even sightly impaired self driving will be preferable.
One is the basic "environmental" interference with competing radar signals from other devices. Imagine how much more complicated this gets when a majority of competing traffic is likewise equipped.
But the other risk, which seems to be mostly ignored thus far, is sabotage. Can you design a system that is hardened not only against stray competing EMR but also against attempts to hack or vandalize your system into misbehavior or malfunction? For example, consider a terrorist plot. Or even a smart neighborhood crazy who wants the neighborhood kids off his lawn and neighborhood traffic to stop during his midday nap. Or, a British organized crime gang using stopped traffic to cover an escape[0].
When all the cars on the road are self driving, I don't see why they would need to have all their radars on at once. I imagine they would be able to communicate with each other and form a complete picture of the entire roadway without all of them having their radars on (or perhaps running at a lower power?)
This would be very difficult to train with ML, however. I'm sure there are other downsides I'm not thinking of
From the "How it works" Google self driving car page: "Sensors
Lasers, radars and cameras detect objects in all directions"
It uses many kinds of sensors. Yes I know about clutter, I've spent quite a lot of time in the radar industry. But by combining data from radars of multiple wavelengths, it becomes pretty feasible. Though yes, difficult.
It strikes me as curious that someone who spent a lot of time in the radar industry would make the original statement you made, at least in that fashion. Also, radar is subject to interference as well. In addition, there are many things that act as "natural" corner reflectors that would pose more challenges, I can only think of some mountainous snowy / icey nothern US/Canadian routes and just thinking No way Light sensors and Radar sensors will conquer this.
We need to fundamentally redesign our road systems to accommodate self driving cars. That might happen... in 50-100 years...after we address the already crumbling infrastructure we have. For perfect, sunny conditions, like the roads in Nevada all these self driving startups are doing their testing, with straight flat landscapes, I am sure the tech can work fine. Rest of the country, maybe not so much.
The original comment was hasty, I admit. But there are a lot of ways to tackle these issues.
Radar clutter, attenuation, penetration, etc are all affected by wavelengths. By using multiple frequencies, you can get a better idea of what is and isn't really there. As far as interference, there are ways to filter out noise. The car presumably knows how fast it is going. Therefor, it can do doppler filtering on the received waves. IE... car knows it sent waves at 50 GHz and is traveling at 60 mph. It knows to expect a response at ~54 GHz from the front and 50 GHz from the side.
For tracking objects, you can use a pulse-doppler radar [0] to get both range and rate information.
There is an announcement today too that Google will not be operating that technology, but focus on selling self-driving to car manufacturers.
Does this mean that Alphabet will keep on working on building their own cars but under a different company? The association between the two is a little confusing at the moment.
A lot of comments here and I haven't seen one that addresses the problem of impulsivity. Many / most of the times I need a car are not schedulable events but impulses. For example when I'm working on a project around the house, it's typical to need to run up to the hardware store 3 times in a day.
I think autonomous transport will be very popular but shared vehicles will be hardly more popular than Uber / taxis are now. If you think that shared autonomous vehicles will be significantly more popular than shared conventional vehicles are today, I'd be interested to hear why. Is it primarily cost savings?
How do you get to work every day? How often do you go to the grocery store? What about going out to dinner?
For myself, living in the Bay Area, anything related to driving or owning a car is a nightmare -- not that that has stopped me from owning a car. Driving from Mountain View to San Mateo takes about 45 minutes in the morning at 75 minutes in the evening. I went to SF to have breakfast with a friend this past weekend and finding parking took half an hour. I'd love a self-driving car that could just drop me at my door. Autonomous cars should also lead to drastic traffic improvements, assuming the cars will eventually be able to communicate with each other in a reliable way.
Maybe you live in a less densely-populated area? I have an aunt that lives in a rural area and I couldn't imagine a need or desire for autonomous cars there.
There is definitely a long way to go, but think of all the advantages! Parking can be automated and condensed, which means more land is available for buildings or parks or anything else. If you suffer from traffic problems now, your daily commute could be cut by 50% or more. Shared vehicles should cost less than what you pay for gas and insurance now. High availability is a real possibility if someone can figure out how to make the economics work. Ride sharing can be almost completely automated, leading to less pollution.
I don't think you read my comment clearly. I think autonomous cars are going to be quite popular, but I don't understand how that necessarily leads to shared instead of privately owned autonomous cars.
Ah. We might be talking past each other. Looking back, a number of the advantages I mentioned are inherent in having self-driving cars. But the benefits compound when we have shared self-driving cars.
Example:
Traffic is improved with self-driving cars because most traffic jams are caused by humans. However, traffic is improved even more when carpooling is automated with a shared self-driving fleet.
Shared autonomous cars will become popular as these compounded benefits are realized.
I understand the points you're making. But I think you failed to address my question. Let me rephrase.
We have fleets of shared vehicles today, but most people prefer personal transportation.
What is it about autonomous shared cars vs shared cars with a human driver that makes you believe that driverlessness will lead to increased use of shared vehicles vs personal vehicles?
What about on-demand transportation doesn't meet the needs of your impulse to run to the hardware store? Just call for a car 6 (there and back) times and you won't worry about having to park on either side.
I think the concern is more about availability. How long does it take for a car to show up each time OP needs to go to the store?
This is a problem with Uber now. I need to be at the airport at 5:00 AM tomorrow morning and I'm just hoping like hell that there will be an Uber around. Most drivers will be home sleeping.
Low-volume times like my example above are where autonomous vehicles can shine. Making the economics work during times of high volume is more difficult. You can always buy more cars, but that's just that many more cars you need storage space for outside of peak hours.
Yes. most people when confronted with the choice of a cheap shared ride, or the convenience of their own vehicle will choose a cheap shared ride. The cost on a per mile basis, will be close to 10x cheaper. I predict even you will change-over when confronted with that kind of choice. It seems inevitable.
> "The lack of control is still kind of terrifying so I think they're trying to make it as non-threatening as possible."
This is true of all Google products, and has been the basis of Google's branding strategy from the very beginning. The dorky, colorful logo isn't an accident. They are a scary, dangerous and harmful company with an extremely friendly face.
This is the point that I see everyone discounting. Whether autonomous cars turn out to be feasible or not, we know for a fact that humans are bad drivers. Don't we owe it to ourselves to try something different?
There's a nice analogy with fighter jets taking off from aircraft carriers. They have handlebars in the top of the cabin that the pilots are required to grab during take off. Experience had shown that, when they took off with their hands in the steering stick, the acceleration was so high that they'd react by pushing the stick, plunging the plane into the ocean.
How long until someone puts a personal gym inside a self driving SUV, so you can workout during your commute? I can imagine people running on a treadmill inside a car while it is stuck in traffic. Or at least sitting on a trainer.
And feed the energy back into the motor. Or even directly power the wheels. It could work if you make the car more lightweight and use a pedal system on gears instead of treadmill. Let's call it a... "Bicycle". :)
Given that these cars are supposedly going to strand us on the side of the road (or in our workplace parking lot) on many snowy winter nights, they had better have beds.
What is it about car hardware+software that gives people confidence that the operating system / some core process won't crash or reboot in the middle of a turn or brake? I'm not familiar with the architecture of "mission critical" systems which I assume would be similar to what they use in these cars.
I'm not sure how I would feel about the absence of steering wheels and pedals within their cars.
Tesla's auto-pilot can be over-ridden due to the presence of the steering wheels and the pedals. But in a car that has none, you're not in control. And that, is scary no matter how you look at it.
Obviously this vision is compelling. I'm confused by the decision of their prototype car to not have manual control overrides (e.g steering wheel or something similar). Air travel has been revolutionized with autopilot, but there are clear overrides for safety in case systems crash. I don't think we need to be wed to the pedal + wheel paradigm - but having a manual override option seems critical to safety.
a big red emergency switch telling it to pull over and shut down may be useful.
Anything more than that would mean your system is not fully autonomous, because the driver would have to pay attention all the time. What good is manual override if there's nobody paying attention? It would also mean you need a qualified driver, you couldn't use it to ferry around children, disabled people or simply people without a license.
They seem pretty clearly, however, to have changed their minds on the topic. It's enlightening to compare the information on this new site with information on the old site (https://www.google.com/selfdrivingcar/). You can see what language they've either removed, or toned down.
The argument is that if you aren't paying attention 99% of the time, that you wouldn't be capable of dropping whatever you are doing in order to take control. Also, these are meant to enable transportation for people incapable of driving.
How is this news? So what, they slapped a new name on the project. What about actual progress on the tech?
Honestly, with Google's track record, I'm starting to doubt that this will ever ship. It just seems like a huge marketing tactic at this point. Plus. all of the talent has moved to companies like Otto and comma.ai; who are making tangible progress in this space.
Its great that waymo has its own brand now. I really wish udacity's open self driving car takes off as well. I predict there will be a linux of self driving car software/hardware kit that will be battle tested, reliable, secure and solid with contributions and test cases from all around the world.
I also really think someone should invent a solution to convert existing modern cars with brake, gas and steering from CANbus to self driving cars by installing a kit.
The current car manufacturers can simply make really good cars with cameras and sensors integrated. One can then install open software adapted for the region that is always getting updates. The test cases and lidar data is put in a giant shared repo whose ownership is the community. E.g OpenStreetMap and wikipedia.
Same software can be used for smaller robots to dispatch packages to the front door, or robotic lawn mowers, rubbish trucks, cleaning trucks e.t.c
I can see the potential of self-driving cars for long distance trips (e.g. SF-LA) or for big trucks, but I can't see it for everyday use especially if someone has kids. How many times an average person in US uses her car? 2-3 without kids and maybe more than 4-5 with kids(I think my numbers are probably too low)?
For this to be viable economically, the pricing has to be very low and in order to be very low all of the people has to use this "self-driving service". This creates another problem though, you will have to plan your "short ride" ahead of time, e.g. what happens if you want to leave at the last moment and no car is available around you? And then there is another problem, if the self-driving service is cheap the ownership of the car is going to be even cheaper. A lot of variables and difficult to answer questions. I don't know if we ever going to be a society without car ownership and to be honest as long as there is no traffic I like driving, and it makes me relax. I see it more as a fancy option for cars, like what Tesla does, but to completely remove the car ownership is going to take many decades. Before cars there were horses and stuff so the transition was easier. This is going to be very tough.
My real question is whether they will be able to convince american customers to give up one of their most prized status symbols. It feels from this webpage their main argument will be safety, so I guess we will find out how much human life weighs against the thrill of owning and controlling fast, expensive and shiny machines.
More power to anybody who's willing to put their children in a car and not pay attention to the road or have control the outcomes of relentless traffic threats all around them.
In my view, all of this technology is much better suited to aid the human driver as a safety enhancement system rather than a full replacement for vehicle navigation.
The problem is that many people learning web development nowadays don't realise there is anything other than impressively complex client-side MVC frameworks. They are taught that traditional HTML pages are a dinosaur and doing progressive enhancement with some simple jQuery or vanilla JS is something for amateurs and old-timers.
I'm really pleased I was forced to learn the web in layers. I learnt basic HTML, then CSS came along. I realised I needed some dynamic behaviour so looked into CGI/Perl and later PHP. The DHTML became all the rage and I picked up a bit of javascript. PHP began to show it's weaknesses and luckily Django and Rails popped up. Ajax became a thing somewhere in all that and returning JSON instead of HTML snippets was sometimes useful.
If you start now and dive into learning Angular/React/Ember/Whatever - you are so insulated from the lower layers it must get a bit baffling trying to understand how all the parts fit together and why some things are done the way they are.
Probably not. My point is that what (I'm assuming) is a brand new site is not built using the latest version. I would have thought that an organization such as Alphabet would want to dog food as much as possible and in doing so tell the world that 'Hey, look at this! It's super easy and pretty! You can have a site like this too if you adopt our technology'. Instead, it suggests to me that the barrier to learning the new shiny is big enough that whoever made it didn't want to learn it. That raises the question, should I learn it? Why, if even people at Alphabet don't want to? Do they doubt the long term prospects for Angular 2 so much that they hesitate in adopting it? Do they not consider it mature enough?
TL;DR I read too much into things.
Framework Ecosystem and current developer experience/knowledge/comfort level have a lot to do with it in addition to uncertainty.
Dogfooding is nice, but not at the expense of risking not having the site showcasing your new autonomous vehicle company ready by the time the press release hits.
When your train is carrying hundreds of passengers, the cost of somebody driving it is pretty small per passenger. In a car alone, that cost isn't split, so is more expensive. That makes it more worth pursuing.
Really shallow of me, but the name kind of turns me off. "Uber" or "Lyft" are just whatever, but Waymo sounds like... something an infant would sputter out. Or "Lame-o".
On their website they have pictures of both the prototype and other manufacturers' cars. it makes a lot of sense to just sell the driving system, like the android model. But I assume they designed the prototype to capture the public's attention. I just don't know why they can't spend a bit more effort to make it look desirable.
... meanwhile Microsoft successfully attacks the billion dollar office communication market.
Seriously, Google has had so many growth businesses in their hands and just fails to execute on them. This might be for one part because of its internal structure but I get the impression that many of these things look not exciting enough for its leadership.
If you look at the cumulative miles driven, the slope has a tendency to increase, as you'd expect as you added more cars to your fleet.
In 2013, though, there's a regression in number of new miles logged, and it took until 2015 to catch up to 2012 numbers. Can anyone give insight into the cause?
this comment section is hater news at its finest. autonomous will change all of our lives for the better sooner than you think. edge cases will be solved. the competition will be interesting, rooting for everyone.
This is really exciting! I see these cars all over the place in the Bay Area, so it's nice to see that they're at least one step closer in making it into a product and thus bringing it to reality.
I wonder about the internationalization of this product, especially if (or when?) it's brought to other countries and regions, such as the Middle East. In some of those countries, women aren't allowed to drive and men drive with extreme speed (and park in the most horrible of ways). While this will cut down on things like speeding / drinking / etc, it may also potentially impact social norms as well.
I'm hoping to keep a level-head about this project, but any step forward in this endeavor is worth being excited about.
Women aren't allowed to drive in, e.g. Saudi Arabia, because they aren't allowed to be out of their home without a blood relative or husband. That could lead to mingling with the opposite sex or something like that. “What would happen if a woman got in a car accident? Then she would be forced to deal with the male driver of the other car, a stranger, with no oversight." [0]
So I don't think self driving cars will solve that problem at all, sadly.
The self-driving car is just a component of bigger system, as new generation of servers on cloud centers. Riding service is cloud computing. I bet on Uber or Didi in China.
What's the story in the video with 'pulling at heartstrings' by featuring a blind person? The problem of a blind person getting a ride is obviously solved. But more importantly it's an edge case of a need for a vehicle that is self driving. And actually I wonder whether someone who is blind actually would rather have a human driver in the car and feel safer generally that way. Seem to me almost like a strawman in a way. In other words 'if you feel that you have to help the handicapped you need to be onboard with self driving cars'.
The website is pretty explicit about it being just self-driving "technology". I wonder if they just plan to work w/ one manufacturer for now or if they're in talks w/ many. With so much competition, if they're all executed to the same standards of safety, my wallet is going to go w/ better design & aesthetics.
> Our testing fleet includes modified Lexus SUVs and custom-built prototype vehicles. We also plan to add modified Chrysler Pacifica minivans to our fleet.
Probably safe to say they're in talks with other auto manufacturers that are looking to add self-driving software to their portfolios.
Well, they only tested in US cities, so it's unclear what has occurred to them. How will it work in small crowded streets in India, China, or just Italy?
They are a US company, after all, who generally aims to release their innovative products and services in the US first, and the US has comparatively predictable and routine driving compared to other major countries.
Why not walk before running? Especially when this is a technology that is heavily driven by machine learning?
I agree it is unlikely to work on non-US roads for a while but I can't really blame them for starting in the easiest conditions. It doesn't have to work everywhere at the outset to be useful.
Truth be told, two thirds of America is worthless in a consumer sense. You can be successful targeting literally only the Bay Area. Why bother solving every problem at once?
Everyone working on autonomous vehicles is facing a similar issue, and it's also one that human drivers are pretty bad at. Most people solve it by not going outside as frequently with snow on the roads.
Tesla's goal is high fidelity GPS mapping of roads and their landmarks, so that the car can navigate down a lane in the absence of normal visual signals (i.e. clear road markings). It will be interesting to see how everyone else deals with weather conditions, but it's also not a problem that needs to be solved instantly.
The weather problem needs to be solved if you want fully autonomous vehicles with no human on standby, otherwise when it snows you'd have a bunch of autonomous vehicles stood still, blocking the road not knowing what to do.
There are tens of millions of US customers who live in areas that never receive snow and rarely receive rain. It's clear there are a few different technical solutions, and probably the engineers working on these problems will think of a few more in the next decade, if it's a major challenge SDCs will just be targeted at minimal weather markets to start out.
I have my doubts that any of the current systems out there - even Google's - work well in extreme inclement weather. I doubt many work well in even "normal" inclement weather (ie - light rain or snow). The biggest problem is sensors.
I am not a self-driving car expert (software or hardware). Nor am I an expert in ML, computer vision, etc. But I do have minor experience in each of those. Currently, I am enrolled and taking Udacity's new "Self-Driving Car Engineer" nanodegree course. For my cohort (November), we just completed the first basic project - Lane Finding.
This project utilized Python and OpenCV, and was far from a robust enough system for it to be used on a real vehicle (for one thing, it wasn't fast enough - it does well with non-real time video and still photos, but it would likely need to be re-written in C/C++ for it to stand a chance in a real-time situation). It couldn't deal with curves, and it has other corner cases.
We had several still pictures to work with, then two required videos, and one bonus "challenge" (ie - extra credit) video. If we had time, we could try our hand with our own videos, etc. It was very easy to get things working with the still images, and the two regular videos. But the challenge video was something else entirely.
In that video, you had to deal with not only different color lines, but also a section of different color roadway, shadows and varying light conditions. With enough effort and thought, it was easy enough to get things working properly, but it highlighted the fact that humans do some amazing things when they drive.
For one thing, our eyes are better capable to handle subtle differences in color, shadow, and lighting (particularly brightness and contrast) than most traditional image sensors. Furthermore, we can "fill in gaps" and "infer" where and how things should be, based on other information in the environment. In cases of rain or snow, we can - for instance - watch where other cars go, follow the "tracks" other cars make in front of us, follow road "edges", and use other subtle cues to help us guide our vehicles as we drive (another instance - recently they repaved the road near my work - but haven't painted new lane markings - I and probably others instead used the "seams" between the asphault runs as "lane markers" instead).
These and many other issues likely are "edge cases" that haven't been fully explored in this relatively new cycle in self-driving car engineering. I am confident that the problems will be solved, likely with deep learning systems of varying kinds, as well as better sensors. Sensor data fusion and priority could also help (ie - use a LIDAR to "follow" a car in front of you, perhaps? Of course, filter out the noise of the falling rain/snow, first).
So far, this course has enlightened me more on just how hard of a problem overall self-driving vehicle systems are to solve. Furthermore, there aren't many players (companies nor individuals) in the market (which is why I am taking the course - to expand on my current skills). Solving these kinds of hard problems will be paramount for a successful self-driving vehicle to last in the market.
Google and Nvidia and maybe Tesla (probably just using Nvidia's system) as well as perhaps a few others _have_ solved those prpblems using things like deep learning, LIDAR and sensor fusion.
This is such a minor nitpick, but I'm bothered by it because I see so much performance-shaming coming from some Googlers:
The site loads large pictures for all viewport widths (ideally they'd load downscaled images for smaller viewports — it's wasteful to load a large image for small devices) and the image files are PNGs when they should be jpeg or webp (example: http://waymo.com/static/images/journey/streets.png)
I've noticed this quite a bit with Google stuff. There's blogger, for example, which is one of the rare sites that doesn't show me anything until I whitelist the entire site in uMatrix.
Hmmm. I'm more usually aware of helpful, insightful, useful articles on how to improve performance, and a constant stream of optimizations to web technology to the same end.
Not that the site in question isn't bloated brochure-ware. Just that I wouldn't characterize Google that way.
I'm optimistic. I agree with Musks estimates. At most I see first fully autonomous vehicles being on the roads within 5 years.
Yes, they may only work in places with clearly marked streets but if I could use my autonomous car for daily commute or roughly 50% of the time it's still an incredible achievement and well worth the cost.
Those two ideas are contradictory. It's not fully autonomous if it only works in ideal conditions. There's already a bunch of prototypes that work perfectly in ideal conditions.
I live in a small enough city where I need to drive for some things but I can walk to work. If an autonomous car could help with little things like shopping for groceries or driving to get a haircut, then my last mile problem would be effectively solved.
I think that there are a lot of little steps towards 100% autonomy that will still provide tremendous value to people, even if the product is far from perfect.
- Instead of talking about their car's capabilities on the front page, they include a pathos about drunk driving. I feel embarrassed how Google doesn't have more things to say about the car.
- It looks really unattractive, I could hardly call it cute.
- It's not in production, it's in testing.
These companies are sitting on mountains of cash! And they still fail to do things effectively! Apple and Google are just sitting on their fat cashflows while the world is getting very scary very fast. I'm tired of these "technology" companies doing complete fails of R&D projects, too damn shy to leave their advertising revenues. We need leaders with real courage.
They've told you exactly what matters to them: not sexy cars, not the kinds of technology used, but the goal they have in mind. Tesla, by comparison, is building sexy fast cars that happen to be self-driving.
• Tesla bootstrapping a ride service on the backs of buyers
• Waymo directly rolling their own fleet
• Uber trying to get a self-driving fleet up, burning mountains of money to maintain their "monopoly" on Uber for rides
• Lyft working with GM to get a fleet up
• All of the other car manufacturers trying to get autonomous vehicles going, presumably hoping for consumers to still want to own a vehicle rather than just pay $1 to get a ride
So: How much is Uber's market share worth? I suspect it'll evaporate overnight in every market where another service has autonomous vehicles and they don't.
Also: Private car ownership is going to fall off a cliff shortly after autonomous ride services arrive. Which probably means general demand for vehicles will fall off a cliff.
I predict blood on the walls.