Actually these Uber Volvo self driving cars have been picking up passengers using UberX in San Francisco for weeks. They're easy to spot because they have a huge lidar contraption on the roof and a lot of camera's mounted on the roof, rear vision mirrors, and the rear tailgate of the vehicle. The depot they use is on Harrison and 3rd so you see a lot of them driving around the area just south of Market Street, which has a lot of traffic obstacles, construction, and pedestrians. They go very slow and stop often out of an abundance of caution, much to the consternation of impatient SF drivers behind them.
Indeed, and I'm almost positive that I've seen models other than the C90s. I didn't realize that they were toting real passengers, though. I thought they were just Alpha Tests As Marketing. Given that they made this announcement the day after Google announced Waymo, I'm now convinced that's what they are, but I guess with real passengers?
The Ford Fusions were used for data collection and always had two engineers in them. The XC90s are used for passengers and they had a free ride program for a little bit here in Pittsburgh.
If you think self driving cars are cool, wait until two years from now when you're stuck behind them in your morning commute every day, while they drive 15 miles under the speed limit.
The depot is the former Otto building, I believe - seen plenty of Otto tractor cabs pulling off of the Bay Bridge, left on to Harrison, then straight down and into their garage/depot there.
I've never understood why people think traffic will move faster with self driving cars. The status quo is drivers driving well beyond the physical capabilities to stop given the conditions at hand. Sure, they might be able to cut down on reaction time, but they still can't do anything about physics.
Given that these vehicles will shift the risk model from driver to manufacturer, and subsequently programmed to obey the laws (which include right of way), we might actually see pedestrians and cyclists using their right of way instead of being bullied away from it.
The idea that cars can communicate with each other so they could drive closer together and faster is complete bullshit...Sure, it's possible, but what profit-driven company would ever take that risk knowing the real-world reliability of wireless communication?
Pretty much all the evidence points to very slow traffic in the future of self driving cars. As someone who mostly walks everywhere, I'm pretty excited.
A lot of traffic jams are so called shockwave/phantom traffic jams and a lot of other ones are where lanes have to merge in each other. Those could be avoided with self driving cars. Maybe in the beginning they will be too cautious for that, but after some time and experience, I'm sure these optimizations will be implemented.
It may not make the traffic in downtown SF faster, but it sure will make the traffic on the 101 from SF to San Jose faster. And good luck walking that.
Those won't go away. They might be mitigated, but once again, you can only mitigate the portion that is due to driver error, and the major problem in those scenarios is physics.
The shock wave phenomenon is one caused by physics (it literally is modeled by fluid dynamics simulation), exacerbated by limited visibility, and exacerbated once again by human reaction times. You can only get rid of the human reaction times, and maybe a little bit of visibility due to communication (although I'm extremely skeptical that there is any incentive for cars to rely on wireless communication to make decisions).
Lanes merging is a physical bottleneck. Sure, humans might make the merge worse, but speed is still limited by the capacity bottleneck, not the friction of the merge. The zipper merge has never proven to be faster, merely safer and more space efficient. You'll get to the bottleneck faster, and cars that try to exit before the bottleneck will get out of your way faster, but that's about it.
Those traffic waves are in large part psychology though. If every car traveled at the average speed you wouldn't have them. If even just 10% are self driving and have data on average speed for that section of road, they can not only avoid the wave but break them up, making traffic better for everyone.
I've even done this myself. I see a standing wave in front of me and slow to what I know is about average for that time of time. People cut in front of me as the gap gets bigger, but by the time I get to the choke point the wave dissolves as I pass through, and everyone behind me who was pissed about my slow speed is suddenly happy.
I'm not convinced, and I don't think there is very much evidence either way. Are the traffic waves caused by erratic driving due to driver error at misjudging average speed? Or is it more caused by small fluctuations due to actual and unavoidable conditions? Would a self driving car see a plastic bag blowing in the wind and slam on its brakes, or drive right through it? What about a kitten? What about a bowling ball, or an unsecured 2x4 falling off a truck? Any one of those situations can cause a shock wave.
We know, due to fluid dynamics simulation, that reduced reaction times will make traffic more resilient to small fluctuations in speed (they can "recover" from the shock wave with lower space requirements), but when the freeway is at capacity there is no recovery room and the shock wave will happen regardless of reaction time.
I try to leave 10-20 car lengths between myself and the next vehicle in this kind of stop-and-go heavy traffic. No waves for me! The anti-traffic essays indicate that doing what I'm doing will actually speed up the breakup of a traffic jam.
That's true, but depending on the local driving culture, this may just lead to many vehicles cutting in front of you and making the gap vanish. In metro NY/NJ, at stop-and-go speeds on freeways, 1 to 2 car lengths seems to be about as much as you can safely leave without it quickly being snatched away. Too bad.
This is an interesting insight. What if we replace 20 sequential cars with a huge "road train" ala Australia - a tractor with multiple trailers?
The road train has no cushion (ignore the few inches in pintle hooks) so it brakes as a unit. And accelerates as a unit. Does that make it's recovery from shocks 20x as good as 20 cars?
The idea of traffic reduction is twofold. First is the fact that there can be a lot more carpools because you don't need the owner of the car to be inconvenienced to make the carpool work, so there will be an overall reduction of cars on the road.
Second, and more importantly, when you have only autonomous vehicles on the road, you can make assumptions that all drivers are perfectly rational, and then your safety margins can be smaller, allowing closer travel at higher speeds.
I suspect in the future, much like how when we transitioned from horse drawn carriage to motor vehicle, at first they will share roads, but then there will be autonomous only roads where the speeds will be higher.
> Second, and more importantly, when you have only autonomous vehicles on the road, you can make assumptions that all drivers are perfectly rational, and then your safety margins can be smaller, allowing closer travel at higher speeds.
Are you saying that a rational driver will never apply maximum braking?
There are non-rational things like animals, kids, and adults with their face in their phones to watch for, so it will never be fully rational...
However, the fact that the computer in the following car can detect instantly that the front car is slowing, it can react instantaneously and (theoretically, with similar braking distances) the following car would stop exactly the same distance (give or take a few inches) from the leading vehicle as when they were driving at speed...
OK, so we agree that the lead auto-car may still apply maximum braking unexpectedly. Therefore the follower needs to leave enough space to compensate for reaction latency, difference in braking power, etc.
So how does data communication (or lack thereof) affect this?
The follower can instantly detect that the leader is slowing via sonar/lidar/radar, right?
Seems like it would be near instantaneous, right? As soon as the leader brakes, the brake lights would be on (speed of light essentially) ... the following car would recognize that, along with radar/lidar type technology. I would imagine the reaction time would be nearly instant...
Then the distance between the cars that they should hold should be calculable by enforcing braking performance on the automated cars...
Assuming its enough, a simple 2 carlength rule would 1) give enough distance to stop and 2) allow for merging easily (second car would slot into place, following cars would slow slightly to make the gap again) ...
Maximum breaking is still a possibility but you can do things like assume you won't get cut off and you can safely ride someone's blind spot and assume they will see you. Things like that.
> I've never understood why people think traffic will move faster with self driving cars.
Because there's a huge difference between individual agents trying to maximize their own benefit to the detriment of everyone else (while also operating based on limited, local sensor data, and with limited processing power and slow reaction times), and a fully integrated traffic flow where fast, powerful agents communicate with each other, share sensor data and telemetry, and are able to make global optimizations.
Scenario #1 often produces pathologic outcomes that are way below global optimum. Scenario #2 can optimize traffic to the extent that human drivers would be unable to keep up with it (think rush hour traffic moving at legal speed limit nearly bumper to bumper with no accidents).
The video you linked to was interesting - thanks for sharing - but I didn't follow how that's related at all to self-driving cars driving around looking for parking. The video seems like more an example of what happens when individual households all own cars and don't carpool when taking their kids to school.
If the kids being dropped off at that school started using self-driving Uber Pools, then that video would be better described as "Welcome to your traffic jam of the past."
The relation is that instead of driving around looking for parking, they queue up at their new bottleneck: loading and unloading. Nobody in the video needed to park, and yet traffic within that area was still moving slower than 3-4mph. Place this phenomenon on city streets where they won't have space for conveniently designed queuing areas, and cars will stop in the road, blocking anybody who is trying to move by. Downtown traffic during rush hour will be moving at a snails pace.
I'm open to the idea that carpooling might increase, but where will it come from? If it is coming from public transit, that would be a net increase in congestion. If it is coming from cars, it would be a net win...but self-driving cars should make ubers cheaper, not more expensive, and the relationship with supply/demand at lower cost would suggest that people would move away from uber pools and towards uberx. I would imagine that the more flexible ride-matching of uber pool (compared to traditional carpooling) would make it easier to use uberpool than it currently is to carpool, and with the right incentives (congestion charges that are waived for carpooling?), we could probably make it work pretty well. I don't know though, the complexity of the dynamics here makes it pretty hard for me to predict with any confidence.
This! Self-driving cars makes car use much cheaper, and I can't imagine any other outcome but substantially more car use as a result, leading to more traffic. At some point we have to deal with this in some way, such as tolling, especially in denser urban areas, where land values are expensive, and the hassle/time cost of driving/parking are the only things metering motorists' use of this valuable resource.
> Sure, it's possible, but what profit-driven company would ever take that risk knowing the real-world reliability of wireless communication?
I think it can be done; the following car may have to brake hard if communication is lost with the leading car.
Let z1 be the safe following distance behind a non-smart car, and z2 be the safe following distance behind a smart car. z2 < z1.
Follower approaches leader and settles at z1. Follower attempts to contact leader. If contact is successfull, follower measures link quality and computes z3=f(z1, z2, quality). If link quality is perfect, z3 can be z1 - in which case the slightest link disturbance causes hard braking.
If link quality is 50%, z3 might be avg(z1, z2), and the follower has more time budget to avoid braking during tiny dropouts.
I think traffic will get worse with self driving cars. If you don't need to be in the car, it can run errands for you during the day. Remember how computers were supposed to lead to a paperless office?
I don't really agree, though as someone who also walks everywhere I'd be glad if you were right. My guess is that as self-driving cars take off to the point of being generally accepted rather than a curiosity, there'll be less and less incentive to operate a car of your own - much as you can still in theory get around by horse but it's so much hassle to own one that hardly anyone bothers to do so.
> The status quo is drivers driving well beyond the physical capabilities to stop given the conditions at hand.
because that's not the most efficient way to move traffic. Yes, a road at it's limit is going to be worse off. Perfect merging and other better habits will help in a lot of places. Just think of all the times you see some idiot merge in at 50 way before the end of the lane.
Uber's product is the company's stock. All of their "AI" and "self-driving" stunts so far have been transparent hype fuel that anyone with the slightest domain knowledge knows have no practical significance.
Their PR department are, as can be expected, top notch though. I especially like how they put a populist spin on their announcements, like the beer delivery (yay, beer!) and now picking up passengers with their proof-of-concept vehicles, to make it look like self-driving cars are already part of their business.
Or... they're trying to learn the nuances of operating with and around real customers for their actual application.
Too many tech companies (and esp. robotics ones) do their development and testing under limited conditions, and then they're blindsided by real-world realities.
Getting contact with customers early is key. It's like the old military adage: No battle plan ever survives first contact with the enemy. In this case, the "enemy" is the real world.
Kudos to Uber for getting out there and working with customers -- it's much better than Google's autonomous cars, which have driven millions of miles and never had a single paying passenger!
Ya, Uber just wants people to think they provide better service at lower prices than their competition; and the tricksters, they're doing it by working to provide better service at lower prices than the competition.
Feeling peckish? Ignore it. It's just the McDonald's corporation trying to raise the value of it's stock.
> Ya, Uber just wants people to think they provide better service at lower prices than their competition; and the tricksters, they're doing it by working to provide better service at lower prices than the competition.
The prices are lower because they're subsidized with invested money and debt.
That subsidization with invested money and debt is not uncommon, atypical, or untoward. In fact, those same instruments are used by most companies (not just startups) in one form or another. Actually, they're used by almost every entity in our economy, even individuals (ie. credit cards and mortgages). So the GP's point was moot.
Nah, Uber and Lyft CEOs both said they envision a future where we don't own cars and just call self driving taxi services via app when we need them. They both envision this happening in less than a decade. This seems to be coming true, millennials have the lowest rate of car ownership and it seems to be the best explanation for why both companies are willing to bleed money to expand market-share as quickly a possible in the mobile taxi app game.
The age old last mile problem. I don't think anyone is suggesting getting rid of mass transit and in fact self driving cars could be the best thing ever happen to mass transit (no need to park at the train station if coming from the suburbs).
When a whole suburb is taking self-driving car every morning to get to the train, I guess someone will innovate and start offering rides in a minibus or even a bus.
Nah, self driving electric vehicles with 6 seats operating on an UberPOOL model will dominate, and it wouldn't be surprising to see the UberPOOL model be extended to a subscription service as well where you simply pay a commit for the service level each month, get an included number of trips, and a further discounted on trips over that number.
Having the commit will make it maore feasible to fund vehicles that are intended for a specific area, and the pool model allows Uber to recover margin during pricing surges and users who don't want to pay the commit (after all, the discount on the fare would only apply to the portion of the pool fare ascribed to the subscribed to the passenger).
One neat thing about this (that probably won't end up making that big a difference for most people, just those few with high-mileage commutes in low traffic areas) is minivans can safely go faster on the freeway than buses.
Adding more smaller trains without a fixed schedule over a broader area is exactly what they're doing. These smaller unscheduled trains are called "cars."
Running empty high capacity vehicles over a really large number of routes is immensely wasteful, much more wasteful than on-demand vehicles.
And even that is ignoring how ridiculously expensive it would be to run trains every 15 minutes for the 6 people a day that have to get from West Bumblefuck to East Nowhereville.
It's all about keeping the same user experience for most people (those with cars).
One thing much of this analysis is missing is how much more in demand commuter vans will be. Transportation will be much more hub based for both long and very short trips. A middle of the range option will fill the void, which is large. Given much of transit is pre planned with times and locations, for jobs, van use in self driving cars will be enormous.
Well, since I put significant work into doing napkin calculations on this and would like to get some mileage on it, here's my comment from a discussion yesterday (https://news.ycombinator.com/item?id=13172888).
Edit: It should be noted this was in the context of a discussion regarding ad-hoc pairing systems for carpooling/ride-sharing possibly making car travel much more efficient, possibly with vehicles that seat more (minivans, full size vans, small busses).
---
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.
Replying to this one as I don't want to reply to yesterday's thread.
(numbers unrelated to yours)
1. BART track utilization at peak times is already pretty much at capacity. The typical westbound rush hour train will be waiting in the transbay tube for trains ahead of it to unload passengers at Embarcadero.
2. Seating capacity on BART is underreporting the ridership. At least 30 more people fit into each car during peak and late hours.
It's great that you dove into the napkin math and worked out your idea. I think that the notion of using many busses is worth looking into. When it comes to capital expenditure and ongoing subsidization, BART is an expensive remedy for traffic congestion.
I think the most vital part of the exercise is to examine the choke points: the Bay Bridge and Market St. Without another bridge and another level for road traffic, I don't see how the city could accommodate the extra usage from the loss of BART, van/bus use notwithstanding.
> BART track utilization at peak times is already pretty much at capacity. The typical westbound rush hour train will be waiting in the transbay tube for trains ahead of it to unload passengers at Embarcadero.
I was hoping people would chime in with actual experiences. I've lived an hour North of SF all my life, but have never ridden BART. There's little reason to if you already have to drive quite a distance just to get to a station, and you only visit the city occasionally.
> Seating capacity on BART is underreporting the ridership. At least 30 more people fit into each car during peak and late hours.
30 more people from the per-car numbers, or from the "crush" load they report, which is "over 200" (and I used 200 for my calculations).
> I don't see how the city could accommodate the extra usage from the loss of BART, van/bus use notwithstanding.
I wasn't really making a case for eliminating BART, but more for where future load capacity might come from. If we already have a mass transit system the numbers show it's a very efficient, if fairly rigid solution than we should keep that and expand it as much as feasible (which likely means, maximizing current tracks, not creating new ones).
That said, a lane packed with buses is much more dense with people than even BART, so dedicated bus lanes (10+ seat vehicles) combined with a shift to ad-hoc ride sharing might do it. Getting from here to there would be hard though.
Except they forget that people like owning their personal space -- especially if it's they need to inhabit for multiple hours in the day, pretty much every day.
And (as with the "smart home" advocates), they forget that, outside of the tech crowd - not everyone likes interacting with (and hence, being dependent on) computers for every conceivable need in their life, 24x7.
The ownerless model might work for some people -- but not too many, I suspect.
I dunno, of my friends under 35 (San Francisco and Oakland) it is rarer to own a car than to be carless. Most couples I know with cars are also single car families. In cities and the surrounding metros (which continue to grow as a proportion of the US pop) I could totally see an ownerless model working. How many Manhattan residents own cars?
An important insight here is that if you and your friends already don't own cars, then moving from drivered-rides-for-hire to driverless-rides-for-hire for you and your friends won't reduce car ownership.
Clearly, in every world, the modern one and the future driverless one, childless adults in dense urban areas are the people most likely to benefit from a rides-for-hire approach. Families with children and people who live in suburban, exurban, and rural areas will be the people least likely to rely on rides-for-hire.
One interesting (though not, I think, probably overall very likely) scenario is where total car stock increases in a driverless car world because:
1. Childless urban adults use rides-for-hire, but they already used rides-for-hire and already didn't have cars. Meanwhile, they start substituting some amount of their former public transit travel for the new-cheaper rides-for-hire (so more vehicles are needed to serve them).
2. Families with children and suburban, exurban, and rural families purchase driverless cars instead of signing up for rides-for-hire, because rides-for-hire still don't work very well for them. In fact, they purchase more cars, because the driverless cars offer them more utility than their old cars did.
3. Meanwhile, a smaller number of people in relatively dense suburb do go to rides-for-hire and abandon ownership, but there are relatively few of them and the cars in their localities are relatively poorly utilized, so they don't offset the increases in vehicle stock driven by people in groups 1 and 2.
As I said, I don't think that's overall a likely scenario, but I don't think it's an insane one. I'd say 10% chance of coming about.
Indeed. Manhattan residents rarely own cars (IIRC around 20% do), and when they do it's always for some weird miscellaneous reason, these being the most common ones I've noticed:
- it's a toy (like a motorcycle or boat would usually be) or
- it's a business tool (like a sedan owned by a cabdriver or a minivan owned by a deli owner) or
- it's owned by someone who recently moved from a suburb to Manhattan and brought their car along by default (which most of the time, lasts no more than a few months before they learn the transit system, get fed up with parking, and sell the car) or
- they're a law enforcement officer or family member (because LEOs, even personal cars, usually get away with parking murder here, so from a parking perspective, manhattan looks as convenient as a sleepy suburb, to someone driving a car with the right protective LEO markings to ward off tickets)
FWIW, I'd love to get rid of our car. I don't like fueling, insuring, maintaining, parking, or driving it. I'd rather use that time and money on something else. Cars are a lot of work if you take care of them.
This works in dense areas. Much of the Midwest is very sparse. I can't imagine people dropping their cars anywhere in Texas (where I live) anytime soon, but I can't wait for my car to be driverless.
Seems potentially unlikely that a consumer could own and (auto-)operate a self-driving vehicle at a rate competitive with Uber & Lyft (or whatever corp.), who could purchase in bulk and have their cars operate for significantly less.
On the other hand, I imagine there will be lots of locations where Uber and Lyft won't serve with their own fleet, either because it's too small, or because it's too sparse. Having a program, or a separate competitor, that allowed people to submit their own vehicles and do their own upkeep might help those areas get served. Then again, those areas might be geographically small enough that you don't need a car (depending on weather). Who know, maybe there's a long tail there which is hard for the big players to capitalize on completely, given they need fleet servicing facilities if they are running their own cars.
If the robocab revolution happens, a lean ride broker where local community pools can offer spare capacity, within well defined availability goals for their members could be the Uber killer. Or Uber itself could be exactly that, given their roots in the claim that their drivers were not taxi operators but just ordinary people who occasionally give someone a lift for money in the cars they own primarily for personal use.
Yes! The far suburbs of Nashville aren't going to see UberTeslaWaymo fleet cars; the utilization would be far too low. They'll see what they see right now: one or two drivers on Uber or Lyft. And the trip price will be higher to reflect the human driving the car.
Well, with respect to the topic of this thread, those are specifically the jobs which people might purchase the car and sub it out to a company for a shared profit. If the cars are autonomous, there's no need for the driver to be there. It could become the equivalent of a paper route, and extra way to make some cash on the side during what are traditionally non-work hours (in this case, the cleaning and mechanical upkeep of the car).
Cars deprecate by mile as much as by year (ignoring the initial drop which is only based on a non-zero preowner count instead of time or distance). So unlike a driver, who is as expensive while driving as while waiting, a robocab spending much of its time on cold standby would not require a terribly high premium on trip prices.
Adoption in the suburbs will still be much slower, but mostly because unused car storage is so much less of a headache out there.
That's only if you think that people are 100% OK with using a random car every time they want to travel. I think they would be willing to pay a little extra to own a car to keep their stuff in at all times (e.g. I always leave my Sunglasses in the car)
Using a random driverless car will be much cheaper than owning (perhaps roughly as expensive as taxis would be if there were no taxi driver who needed to get paid). Random cars can be in use virtually all-the-time every day. The portion of costs a single passenger pays will thus be small fraction of the total costs. For a car you own, at least as it works now, you generally pay 100% of costs. So it's not going to cost just "a little extra" to own your own car; as it works now owning a car results in huge levels of waste, while the expensive item sits doing nothing, occupying a (valuable) parking space.
I suppose you could own a driverless car and have it continuously deployed as an income-producer moving other people when you're not using it. That creates a lot of additional headaches and problems, though. E.g., how to make sure it's always available whenever you decide you want it? Or, will it be safe to leave your items in the car when it's used by others? (I would guess not).
As a city dweller, I would take random car over random parking spot every single time. Certainly does not apply to those who already pay a little extra for private parking though.
A year after the tech is available, I'd expect a thicket of regulations and permits to emerge that make it economically unviable for anyone but a large company fielding thousands of cars to profit.
Sort of like how you couldn't open your own payment processor for the few hundred people in your neighborhood today.
Well... depending on who you source, the majority of people already live in urban environments. In the US, the most recent figure from the census is 62.7%, which is more narrowly defined at % of population living in cities[1]. If you count "urban areas" which includes towns and villages, it is north of 80% [2]
Worldwide, we are at ~54% living in cities. [3]
This trend seems to be continuing globally as well [4], so pretty soon people in non-urban environments may have no choice (assuming, as I do, that cars will get more expensive as they have more and better technology and are much more highly utilized due to sharing)
No choice? In rural environments, where you have to drive an hour to a small city and 3+ hours to a large city, and there are no local businesses, no one is going to be operating self driving cars for hire. Electric cars aren't currently practical in such places either, though over time that will likely change as battery and charging tech improves—but I don't see shared ownership or commodity car usage replacing ownership rurally at any point, and there will always be people living rurally. If new cars stop becoming available at a reasonable price to buy, then people will continue repairing, rebuilding and using old cars.
Seems possible that as automation of delivery services becomes more widespread (e.g. drone delivery), and as both remote work opportunities and entertainment technologies increase, there will be less of an incentive, or at least less of a financial/work-related incentive to for populations to accrete around centralized urban developments. So the growth of urban vs. rural or suburban population rates may potentially slow in a few decades. Maybe, but probably not, idk.
I live in what's considered an urban area due to population density. But I wouldn't personally describe it as urban. I have multiple farms near my house, and on the way to work, I pass by numerous farms for a good part of the way. Be careful confusing urban with city. It's merely population density.
Regardless, Uber et. al. doesn't reach where I live. It's not an option. It's not a choice at the moment. Living in an urban area doesn't mean you are living in the city. Public transportation is a long way off from being viable out here in these urban areas.
Right now, you cannot live and work here without having a car (unless you limit yourself to very specific conditions, but that could be said for all locations). And again, this is considered to be an urban county.
> The U.S. Census Bureau defines an urban area as: "Core census block groups or blocks that have a population density of at least 1,000 people per square mile (386 per square kilometer) and surrounding census blocks that have an overall density of at least 500 people per square mile (193 per square kilometer)."
By that definition, which would seem to completely rule out farming, I doubt you're actually in an urban area.
I have a friend working for a major supplier in self-driving vehicle sensor tech who went to Pittsburg to learn from Uber. I imagine they at least did their homework before flying him out there.
How is this going to combat people who make a mess inside the vehicle after going 100 percent without human drivers/test engineers? If it picks up a drunk person at 3 AM who then throws up inside the car, is it vomit-aware and knows it needs a cleaning before picking up the next passenger?
The same way stores keep unattended restrooms clean: periodic inspections throughout the day and a timely response to customer complaints: If cars report to a central facility for a 5-minute inspection after 45 minutes of service, they've already cut labor costs by 90%, but with some increased capital costs due to decreased fleet availability.
I was thinking it would be kind of up to the next passenger. If they are ready to get in the car/Uber, and notice it's extremely dirty, or god-forbid someone puked in their recently, maybe they have something on the app or in the car that says "Report car for service inspection/cleaning" and "request another ride".
Uber already charges a cleaning fee of $40-150 at the driver's discretion. There's margin there for the cleaning cost, opportunity cost of the vehicle time, and a credit to the reporter for the trouble of seeing vomit and waiting for the next car.
I imagine that the next user would just tap a button on their smartphone saying that the car is dirty, upon which the car drives back to a service center and the user receives a new car (not the same car).
The problem with this is, as a rider, I personally would not want to have to be constantly expecting a dirty car. If one out of every 3 cars has vomit or syringes in it or feces, I would personally stick with a service that has human drivers at the wheel.
This is remedied with a service charge for the responsible person.
Unlike a Tenderloin sidewalk, deterrents are easily enforced under a scheme like this.
Uber and Lyft currently charge $100 for a cleaning fee to the driver when their passengers soil a car's interior with vomit, etc. I don't see why Waymo, Tesla, or Uber would be any different.
Zipcar and Reach Now don't seem to have an issue with this, I would be surprised if it's any different when you don't have to the move the steering wheel yourself. If they can make a car that drives itself they can surely figure out a way to keep it clean.
Sure if one out of every three had vomit, syringes, or feces it would be a huge problem. One out of three seems incredibly unlikely and off by at least a couple of orders of magnitude.
Maybe this would self-regulate. If someone is the kind of person that would make a mess on an autonomous car ride, maybe after calling the service again and having to wait for a replacement due to someone else’s carelessness, their attitude would change.
1 in 3 seems ridiculously high. I'm not sure what the vomit-rate is currently for Ubers and taxis, but I'm pretty confident it's not even remotely close to 33%.
This seems solvable. As a rider, you have the option to select "This car is too dirty to ride" or "This car needs cleaning."
If you indicate it is too dirty to ride, the car reports for cleaning. If the cleaners agree with your assessment then your ride is free, paid for by the person responsible for the mess, who would also pay for the cleaning. If the cleaners don't agree, then you pay for the cleaning and your own ride. If you indicate that the car is dirty but want to ride it regardless, then the car could just check in for cleaning after your ride at no expense to you.
If you don't want to send the car away, perhaps you have an urgent appointment, you could just ride the messy car and have it cleaned after your ride. If adding a few extra minutes to your trip isn't a big deal, then the reward of having it be free may compensate you for the inconvenience.
A system like this would provide an incentive for customers to report dirty cars (saving money on their ride) and to leave cars clean (avoid paying for car cleaning). Those incentives are hopefully at least as strong as the incentives in place now.
"Using superadvanced machine learning techniques we've trained our great Uber Customer Surv^D^D^D^DSatisfaction Improvement Module (TM) to predict customer vomitting with over 90% accuracy and will automatically deploy a <your name here> antiemetic for just $10 (free for Uber Prime customers). Tell us how we can improve Uber even more in the comments!(1)"
What keeps people from doing that on subways now? Nothing but the social contract, and so subways are generally somewhat dirtier than cabs, but still get you from point A to point B.
The difference is you have much more privacy in an Uber vehicle. Someone about to make a mess in the subway might be more inclined to wait until they get off vs a private vehicle with just them in there.
How do you have more privacy in an Uber vehicle that knows your identity, everything about your trip and can bill you for any damage? You're anonymous on most transit.
Go with the flow: Institute a highly profitable vomit fee, and offer a VR headset for the passengers to wear during the ride, to increase the chances that they'll throw up.
"I'm a vomit-bucket-half-full sort of guy. I don't think these cabbies are trying to charge you for puking, I think they are offering you the premium service of vomiting in their taxi."
"America was built on the idea of premium services. This is how the wealthy are able to have so much more fun than everyone else. They can behave however they want as long as they have the money to cover the premiums! The $70 currently in my wallet entitles me to a good three blocks on Michigan Avenue and a nice half-digested deep dish pizza projectile vomited all over the headrest."
"Yep Uber is pretty good here in Sydney as well. I had the outside of my car given the Jackson Pollock chunky rainbow look by a passenger. His friends thought it was a real laugh and afterwards I found they had also taken my giveaway chocolates as well. Sent the report in with photos and $95 receipt for the cleanup, and got $250 credited to my account the very next payment.
I have to say, in all my years of driving public vehicles Uber (with its hold on the rider's credit card detail) , have the best and quickest method to compensate drivers for these horrible incidents when they occur."
They can certainly have a video/image feed of the inside, to remote check if it is messy once the ride is over(can be automated to a good extent with image recognition), and retire it from service until its cleaned up.
Self-checking a car interior between rides should be one of the easiest computer vision tasks imaginable: fixed camera angles, fixed scene and any differences in outside lighting could be drowned with a strong internal flash (which could even double as a minor privacy improvement if the cameras were tailored to match the light intensity of that flash to be practically blind without).
Have cameras take pictures of the interior before and after each ride. You then have a verifiable record of what state the car is in, and you don't have a bunch of data-privacy issues around recording activities while someone is inside the car.
The only way I think this can work is if the customer who reports the incident describes exactly what they see in the report. Then the human who receives the dirty car and cleans it verifies that the car is not only dirty but meets the description of the reporter. How else could "dirty car" reports be audited?
In another case, a customer could actually throw garbage in the car and then report the car dirty - causing the previous customer to be fined. The only safety net against this I know is to record video inside the car while a customer is present.
Probably. Even if it can't be automated (which doesn't sound too complicated), it wouldn't take many people to evaluate a photo of every cab interior once after every ride.
You've already got someone to evaluate it for free: the next person to get in it. Then you only need people to evaluate the reports, and you offer the person reporting it either a free ride in the dirty car or a fresh car in a few minutes at a discounted price.
They describe this as a third-generation vehicle, but it still requires multiple human-driver interventions over a single journey. How many more generations to reach reliable true autonomy?
My opinion as a researcher in AI/ML but without any expertise in self-driving tech:
We are a long way from where I would be willing to trust my life to self-driving cars - as a passenger, as another driver on the streets, as a cyclist, or as a pedestrian. Much farther away than these companies press releases make it seem.
Here's why. These driving algorithms are successful in large part because of data. They train their systems, such as visual recognition (what are the objects in the world around me), on millions of miles of visual data collected on the roads, most of it in California in the sunny daytime.
This means they are very likely to perform well in the average case when everything goes according to plan. And if deployed there they might live up to the hype and save thousands of lives compared to human drivers.
But now say you're in a major city in the midwest or northeast, for instance. It may be night time. It might be raining. There might be two feet of snow on the ground, narrow lanes, road signs covered up and unreadable. There may be a pedestrian crossing in dark colors. The street lines may be faded or nonexistent. There may be a street that is marked one way on the GPS map but is currently detoured the opposite direction due to construction.
There may be a policeman directing traffic. The police might pull the car over and direct it to a parking lot. There might be a fire truck or ambulance coming at an unusual time.
A computerized system trained on data can only perform well in situations very similar to its training set. But its vision will have a hard time recognizing objects it hasn't seen before. Its language processing will not understand unusual or novel road signs. Even if it recognizes the objects around it correclty, it lacks the "true" intelligence to deal with unforeseen situations falling significantly outside its training set.
I believe that cars are quite likely to run into novel situations they haven't experienced before, and I don't trust their reactions or decisionmaking in these scenarios. So I think what we have are self-driving cars that perform very well in the common, easy case, as we have already seen in numerous press releases, but are in my opinion very unpredictable in the long, fat tail of situations.
I think you are missing the most important part here. These cars are always online and share data between them. They have a detailed map of every street and every road bump and every road pole/sign that can be used for navigation. Even if everything is in snow and the camera/lidar is frozen and can't see anything, these cars know exactly where they are and where the road is from predictive navigation based on speed, direction, road shape/bumps from previous data that was collected from 1000s of passes before that on that very same road. At first AI cars will probably avoid certain areas that have not been mapped. Each car will signal any unexpected road blocks, data will be sent realtime to a human operator who will script a walkthrough in seconds. Such as "ok, you are legitimately stuck in traffic right now, just wait" or "ok, there is a crashed car ahead of you so the right-turning lane is closed, move into the left lane and you can turn right from here as an exception". There will be humans like ATC in all cases.
Police and emergency services will just coordinate with the "ATC" to pre-script routes differently depending on the situation.
Also, communicating and negotiating with human drivers. You have to do this all the time in a city like London, mostly on two way roads where there's only room for one car, due to obstacles (roadworks) or parked cars.
In complex situations, flashing lights and honks are not enough, you need to verbally communicate with other drivers.
I would eat my hat if an AI could handle these kinds of situations.
So in short, I'll believe the hype when I see a video of a full auto drive through London at rush hour.
If they're referring to the typical classification of autonomous vehicles, Level 5 is complete autonomy. Level 3 refers to human intervention required for everything but highway driving in good conditions. Of course, they may be referring to the third generation of vehicles made by Uber, which is a relatively useless term as each manufacturer will require a different number of generations before reaching full autonomy. Not really sure as it's entirely possible this is both third generation and level 3 autonomy.
It's Uber's 3rd iteration of their prototypes. The first they tested in private for 18 months, the second had those ridiculous 22 camera 6 Lidar sensor suites and was unveiled in May, and this new configuration is brought to you by Anthony Levandowski, Kalanick's new top general, and the only real reason for the Otto acquisition. "My brother from another mother" said Travis.
The Otto deal was set around incentives, it's not actually a flat 600 million dollar deal. If Levandowsky can deliver a Robotaxi OS that carries Uber to greatness, his networth could be astronomical. He's probably the top candidate in the world right now to lead an autonomous driving project to commercialization. He's got a vicious Randian streak and he's been at the bleeding edge of driverless vehicles since the Darpa days.
Currently, self-driving AIs are kind of like chess AIs, or the Watson AI that won at Jeopardy: they have some inhuman strengths that mostly compensate for their sub-human understanding of inputs.
So, for example, chess AIs are better chess players than humans because they have longer lookahead, even though they're worse at analyzing a given board. The Watson AI didn't understand questions as well as a human, and had obvious comprehension failures such as the final jeopardy answer, but when its comprehension was good enough, it had a vast database that it could perfectly recall at very high speed, and those more than compensated for its comprehension problems.
Driving AIs are not as good at understanding what's happening around them, but they are constantly attentive in 360 degrees and have fast reactions, which (may, someday, but does not yet) compensates for their imperfect understanding of the world.
This AI is not generalizable to tons of other circumstances in which there is no obvious way to parlay the inhuman strengths of the AI into compensating for their weaknesses. As such, while there may be some other places where a driving-AI-like intelligence could be used profitably, there probably aren't many such places.
Yes, but there are lots of applications for driving capabilities, that are not getting people from a to b.
Like transporting cargo at an airport, dumpers at a pit mine, even long haul trucking (see Otto).
I ride a motorcycle in SF daily and frequently "split lanes". I saw one of the Ubers all decked out with spinning radar and wondered how it would react as I passed through at 10-15mph. There was a person at the steering wheel, I am not sure who was in control. Thankfully it was uneventful. I thought the car might brake or veer away from me.
I'd be curious as well, but after lane splitting in the bay area, my default assumption would be that a self-driving car would react in a safer way. In my experience most lane splitting close calls are due to inattentive drivers changing lanes without signalling or randomly shifting position within their own lane.
That seems like click bait to me. I believe they require a permit for a fully autonomous vehicle, but since there's a driver in the car monitoring the systems with the ability to override, it isn't considered fully autonomous. This is neither sketchy, surprising, or even skirting the rules. The fact that it's picking up passengers seems mostly like a PR move to me as the income generated by them probably pales in comparison to the cost of the car plus the overhead of a (not typical) driver plus engineer.
No permit as far as I can tell. They have drivers actively monitoring their vehicles at all times and are playing semantic games right now as that article says. The biggest disappointment to me is the way they're deliberately ignoring NHTSA's guidelines around collision and near-miss data sharing.
"The company doesn’t require a permit from the California DMV to operate in the state, it says, because the cars don’t qualify as fully autonomous as defined by state law because of the always present onboard safety operator."
Interesting. My guess is Uber is just planning on paying the inevitable fines as is their modus operandi. I wonder how much an infraction like this costs? I doubt there's any precedent.
Not sure why you're being downvoted. It's the MO of airbnb, uber and other SV startups. We've seen more than once that making a land grab for a new market brings more long-term success than any penalties dished out for not playing by the rules.
Because there is no reason to guess and make baseless claims, the article already addressed this topic. A permit is not required because a safety operator is on board at all times. Skirting regulations and paying fines has nothing to do with it.
That's only because of Uber's interpretation. From the FT article:
> California law defines an autonomous vehicle as one that is able to drive “without the active physical control or monitoring by a human operator”.
> Anthony Levandowski, a former Google engineer who leads Uber’s driverless car efforts, said the fleet would be “self-driving” but not “autonomous”, making an unusual semantic distinction.
A pretty tenuous way of avoiding regulations if you ask me.
My understanding of the law is that an autonomous vehicle does not require control or monitoring. However, Uber's car has an operator monitoring the car's activity, although not actively controlling the car. Because of this, it is not considered autonomous by California law.
I wonder if everything you say & do is recorded in these? The car has all those cameras covering the outside of the vehicle, it would be little effort to add another inside, and I doubt Uber will be able to resist.
Uber could claim they need to record passengers in order to spot damage or dirtying of the cab. (Otherwise passengers could blame any damage on the previous occupant.)
This is already occurring in almost every Taxi I've ever been in. They use it to protect the driver and passenger. This would be the first time all of that data is aggregated to one company for potential abuse but compared to what we see from gvt agencies, this is not my biggest concern.
What planet have you been living on? Of course they will do that if not regulated. Officially they can claim blurb like "improving customer experience and safety", "compliance with law enforcement", "protection of our property", "automatic detection of medical emergencies" while their main economic interest will be the further dismantling (not much left to dismantle anymore in 2016 anyway) of privacy for marketing purposes and buyer manipulation.
Most taxis already have cameras inside them, plus there's a driver there watching you. I'm a staunch privacy advocate but a taxi is not a private space and never has been.
> The car has all those cameras covering the outside of the vehicle
This raises a good point i never thought about. If in some day all of our cars have self driving or self aware features, then that means nearly all locations in the city are within reach of cameras and possibly microphones. I'm not "a paranoid" about privacy stuff, but that is quite impressive nonetheless.
The future is going to be quite interesting. I always thought we'd end up with cameras on every street corner monitoring everything. I never thought our own cars could become every present monitoring devices.
I didn't thought of it that way round! I was concerned with cameras & microphones inside the car, but as you say, it's likely that there will be microphones on the outside too.
Yes, they said that in the email they sent about self driving service:
> I heard there is a camera in the car. That’s correct. To learn more about how we can improve the self-driving experience for our riders, your trip may be recorded.
When picking up a ride they notify it's a self-driving Uber and let you cancel for free.
The biggest problem human drivers have in winter weather is simply driving too fast. Self-driving cars won't get impatient and will presumably never exceed safe speeds for given conditions.
Having lived nearly my whole life in Michigan, I think this grossly oversimplifies the problem. I'm sure my Canadian friends will agree. This assumes there's some clear algorithm for determining a "safe" speed. There's not.
Even in highway driving you can lose traction in an instant if you hit a patch of black ice. Autonomous vehicles will need to be able to recover from a complete loss of traction safely. This isn't trivial - in fact it's probably the most complicated bit of driving I tend to do. Once you are sliding and your steering wheel becomes more of a suggestion than a command, the entire act of driving becomes a process of trying to coax the car off the road using a combination of steering, brakes, and even occasionally gas. I think it's possible for a computer to do this - but you can't avoid all slides just by driving slowly.
Then there's the plethora of other winter fun you run into with a vehicle: getting stuck (happens all the time on city streets) and all the techniques to get unstuck, going too slow and losing your momentum (and thus traction), having every indication you have traction and then discovering you actually don't (it's very easy to be driving at a "safe" speed and still slide through an intersection), white out conditions where you are guessing where the lane is... etc...
To be clear, I believe most of these conditions could eventually be handled by computers. I also believe a lot of people drive too fast in/on snow. However, winter driving is in no way simple. It's a problem domain unto itself, and one I've seen relatively little work being done on.
You may be correct that relatively little work is being done in this area, however I think you overstate the difficulty.
Winter driving is really no different from any other kind of driving in which the driver exceeds the limits of the vehicle's available traction. The methods of recovery are mostly well-known; the problem is more often the driver's inability to implement them in a timely manner.
Constant input from wheel speed and accelerometer sensor arrays, coupled with the vehicle's ability to individually brake/slow wheels (which also gives the vehicle the ability to accelerate individual wheels independently!) means that it could be a far easier 'problem' for self-driving cars to solve than it is for humans.
Again, if they're working on it :) But there's already been decades of work put into ABS, traction/stability control, etc.
> Winter driving is really no different from any other kind of driving in which the driver exceeds the limits of the vehicle's available traction.
I actually agree with this statement. The problem is that, in good conditions, low traction events are rare, and often caused by catastrophic conditions. In winter driving it's practically the norm, once you leave well traveled roadways.
I think AI could be trained to drive a car that only occasionally has full traction, and probably more effectively than a person given enough time. But again - it's like you said - someone needs to be addressing this case directly.
I use to live in an area that got heavy snow and had an all wheel drive vehicle (a WRX). I have lost control in the snow, even driving slowly in 2nd gear, under 30kph. For a human at least, your reactions need to be pretty automatic: point the wheels where you want to go and .. and this only applies to AWD/4x4: don't break. You don't really want to accelerate much, but you don't want to lose power to the wheels. Keep calm and the car will straighten.
It comes down to balancing, and we know we can build machines that can balance. Segways, bipedal robots, etc. So we have some of the tech and algorithms to do this in other applications, but applying it to autonomous vehicles will be its own beast.
I think such acts of control are one of the things where computers are naturally better than humans. We have a lot of experience in automatically controlling dynamic systems. Winter conditions are more of a problem because they might interfere with the sensors (e.g. radars and snow don't play together very well).
On the other hand a computer-controlled vehicle, possibly electric with 4-wheel drive, would have millisecond control over all those inputs (steering/brakes/acceleration)
I don't disagree. But without exception every winter slide-off I've witnessed first hand has been caused by excessive speed. You're driving along on a snow-covered interstate highway, someone in a BMW or 4x4 flies past you because "I have all wheel drive" starts to fishtail, overcorrects, spins out, and ends up backwards and stuck on the median.
Respectfully, I think this amounts to bias more than anything. It's a lot easier to believe that the slide-offs are caused by bad driving - and most probably are. But some are unavoidable.
I have to agree as well. I've only had two winter driving accidents, both of them just me and one of them was unavoidable: I was travelling at very low speed down a plowed street and found a long streak of ice. I've been in plenty of winter skids and am decent and correcting them but I had 0 traction and essentially skated at 5mph into a telephone pole (which seemed safer than the cross street I was drifting toward).
I'm not sure how a self-driving car would have fixed that. I certainly believe one could but it would require a ton of learning beforehand and conditions vary widely in storms. I suppose with all-wheel drive and some selective application of the wheels in reverse it might have stopped the car before any damage was done, but how to handle situations where the AI can't stop the car before an incident and has to minimize the damage? That feels like a lawsuit waiting to happen in the US.
The biggest problem computer drivers have in winter weather is not being able to see lane markers. Humans are (for now) still better at figuring out where the lane is. Presumably this problem can be solved with improvements in maps, cameras, and image recognition algorithms.
Humans do an unbelievably bad job at this, too, like all other tasks behind the wheel. Humans essentially just follow the tracks of the last car, creating ad hoc lanes. Computers can do this just as well.
On the other hand, humans also have the common sense not to (e.g.) follow the tracks of a car that has spun off the road.
I'd expect logic at that level might be more difficult to train for than the relatively low-level visual task of identifying and following tracks. But I guess existing self-driving systems already need to work at that level to predict the actions of other cars and pedestrians, so presumably the current data-driven training techniques would also handle it OK? (Assuming a lack of pranksters with shovels...)
What happen when you are the first on the road with no tracks to follow from previous cars and suddenly it turns into a blizzard/zero visibility? I am guessing the car can pull over and turn on emergency flasher...I don't know but to me I believe a Human will always have to take over at some point because you never know what mother nature will throw at you, keep in mind I am talking about a future where we aim for 100% autonomous car.
Having lived through many such blizzards, there are two things that humans do:
The smart ones pull over and turn on the emergency flasher. Then they check to make sure they have lots of gas, blankets and water, just as they prepared.
The dumb ones slow down to 45 mph and stare into the white blankness looking for clues about curves.
Exactly! Every question about how a computer-driven car will handle exceptional situations encountered by humans shares the same answer: the computer won't have got itself into the situation in the first place. Humans might drive a car into a whiteout blizzard, but that only serves to demonstrate the urgency of cutting humans out of the loop.
Just to provide a counter point, in my experience, the biggest problem I face in winter weather is other drivers. I know when I hit a patch of black ice and have difficulty stopping that the oversized van behind me is going to have more difficulty stopping, and I have the instinct to quickly turn instead of waiting on the red light. I also know that cars have difficulty stopping on a steep hill and I may wait before going through an intersection. Both of these experiences have happened in the past year for me.
This may seem like a straw-man argument, I don't intend it to be. I think self-driving cars will be on the whole better drivers, but these are also situations that I see as being extremely difficult for a computer to identify to the level a human driver is capable of.
Agreed. I don't think the problem in winter driving is the speed. It's anticipating what the hazards are and how to avoid them.
You - and the other drivers on the road - have less control at all speeds and are always much closer to the limit of traction. Driving on winter roads is a lot like racing a car on a track with other drivers - everyone is near the limit of traction and a hazard can present itself very quickly. Having the right reaction at the right time helps, but planning ahead is more important. Daily driving in summer months is benign in comparison.
> Driving on winter roads is a lot like racing a car on a track with other drivers - everyone is near the limit of traction and a hazard can present itself very quickly
Bit of a conspiracy theory sort of question, but... is there any reason to believe that maybe self-driving car technology is being backed by the military-industrial complex as a way to run R&D for military automation and related technologies? That could explain why there was a sudden spike in interest in this technology several years ago, and it could also explain why enormous amounts of hype are continually being generated for a technology that is probably still quite far away from being approved for fully automated road use.
No, it's the other way around — the private industry's self-driving car technology is based on research funded and carried out by the military-industrial complex. See: Uber's takeover of the CMU robotics lab (NREC).
There's no conspiracy about it - that is exactly what happened, and it happened right out in the open, in the form of the DARPA Grand Challenge races in 2004-2007. The explicitly stated goal of those races was to motivate research into autonomous vehicles that would enable the US military to begin converting its ground vehicles to autonomous operation. The much-hyped commercial applications are a spin-off.
The DARPA challenge is somewhat of an orthogonal use case. The DARPA challenge was based off of an autonomous vehicle that doesn't have extensively pre-mapped roads; it is in rough terrain. The vehicle had to alter its route on-the-fly if it encountered impassable obstacles. It also didn't have to concern itself with traffic. I presume the use case is getting materiel to troops without risking human drivers. This is important considering IEDs were a major source of injury and death for coalition troops in the most recent wars.
Google/Waymo and its successors are using pre-mapped courses with many heuristics and edge case tweaks. Routes are generated from existing resources (Google Maps, etc). Much effort is devoted to avoiding other vehicles and pedestrians. Much of the rules are based off of U.S. traffic rules, such as speed limits, stop signs, traffic signals, and lane markings.
They both share technology (computer vision, momentum/traction control), but I conjecture the bulk of the work for commercial autonomous driving was not related to the DARPA challenge and wasn't paid by its grants.
That was true for the first two Grand Challenges, but once that prize was claimed DARPA simply cranked up the difficulty level; the 2007 race was all about navigation in urban environments, and compliance with traffic laws was a condition of success.
Sure, of course you're right that "the bulk of the work" is not related. But it's also no secret that the "sudden spike in interest in this technology" lucker referred to above happened because the US government paid for it to happen, as a means of advancing military vehicle automation technology.
> “a blade architecture, a whole bunch of CPUs and GPUs that we can swap out under there,” though he wouldn’t speak to who’s supplying those components specifically.
Does anyone have more info or speculation on the tech stack sitting in the trunk?
Is this a voluntary thing? Do I get a discount to sit in those? I would at the very least want to know in advance that an autonomous car is going to pick me up so if I want I can reject it if I want to and not waste time.
One thing that's still not clear to me is how autonomous these truly are. There's still a human in the driver's seat, but how much intervention is required from them?
Does anybody know what the chance of getting a self driving car is? I could try to use uberX instead of pool for a while, if there's a sufficient chance of getting one.
