Frozen Driverless Cars are Delaying San Francisco's Buses (wired.com) 188
There's a new problem with driverless test vehicles. Wired obtain records from San Francisco's public transit agency for about six months showing that driverless cars testing on city streets "resulted in at least 83 minutes of direct delays" for the city's "Muni" buses. And "that data likely doesn't reflect the true scale of the problem," Wired argues, since "a single delay can slow other lines, worsening the blow."
Some examples from the article: - On January 22, a Cruise at a green light wouldn't budge, preventing a San Francisco light-rail train from moving for nearly 16 minutes. As the train driver headed out to investigate, a passenger said, "Nobody in there, huh?" Over a span of 10 minutes, the driver chatted with passengers, checked with managers over the radio, and walked around the motionless Cruise vehicle. Someone wearing a reflective vest and holding a tablet eventually got into the Cruise and drove it away...
- On September 30, 2022, a Muni light-rail train, or streetcar, that was full of celebrating baseball fans began driving from a station into an intersection. An empty Cruise robotaxi at a stop sign to the train's left then also drove forward... It was seven minutes before the driverless car cleared the track and the train started again, drawing cheers from riders...
- On January 21, a Muni bus with a couple of riders aboard had lost six minutes because a Cruise was lingering across an intersection crowded by police and fire vehicles, video shows. While other cars maneuvered past, the Cruise did not. "I have one of those autonomous cars in front of me, so I'm stuck," the driver radioed. "I could make this turn on Sixth Avenue if this car wasn't in front of me...."
- In November, one light-rail passenger called it quits after waiting nearly six minutes for a Cruise driverless car in front to move. "There's nobody in the car," the driver told the person as they stepped off the train.
- [After a white Waymo SUV stopped in the middle of the road, Waymo spokesperson Sandy Karp] says one of the company's roadside assistance crews arrived within 11 minutes of being dispatched to drive the SUV, clearing the blockage about 15 minutes after it began. Karp declined to elaborate on why the remote responder's guidance failed but said engineers have since introduced an unspecified change that allows addressing "these rare situations faster and with more flexibility...."
Some examples from the article: - On January 22, a Cruise at a green light wouldn't budge, preventing a San Francisco light-rail train from moving for nearly 16 minutes. As the train driver headed out to investigate, a passenger said, "Nobody in there, huh?" Over a span of 10 minutes, the driver chatted with passengers, checked with managers over the radio, and walked around the motionless Cruise vehicle. Someone wearing a reflective vest and holding a tablet eventually got into the Cruise and drove it away...
- On September 30, 2022, a Muni light-rail train, or streetcar, that was full of celebrating baseball fans began driving from a station into an intersection. An empty Cruise robotaxi at a stop sign to the train's left then also drove forward... It was seven minutes before the driverless car cleared the track and the train started again, drawing cheers from riders...
- On January 21, a Muni bus with a couple of riders aboard had lost six minutes because a Cruise was lingering across an intersection crowded by police and fire vehicles, video shows. While other cars maneuvered past, the Cruise did not. "I have one of those autonomous cars in front of me, so I'm stuck," the driver radioed. "I could make this turn on Sixth Avenue if this car wasn't in front of me...."
- In November, one light-rail passenger called it quits after waiting nearly six minutes for a Cruise driverless car in front to move. "There's nobody in the car," the driver told the person as they stepped off the train.
- [After a white Waymo SUV stopped in the middle of the road, Waymo spokesperson Sandy Karp] says one of the company's roadside assistance crews arrived within 11 minutes of being dispatched to drive the SUV, clearing the blockage about 15 minutes after it began. Karp declined to elaborate on why the remote responder's guidance failed but said engineers have since introduced an unspecified change that allows addressing "these rare situations faster and with more flexibility...."
Follow me (Score:5, Interesting)
Police, firemen, and other public servants should have an app on their phones that generates a one-time QR code that they can hold in front of the car's camera and the car will slowly follow the phone until it is out of the problem area.
In these examples the bus driver or light rail driver could guide the car off the main thoroughfare or tracks and the stalled car could report the override back to the owning company with time and GPS location. Police and firemen could do the same thing during emergencies to get driverless cars out of the area.
Re: (Score:2)
Or the owning companies should be obliged to fit a remote control system and use that to drive them out of the way after 5 minutes of being unnecessarily stationary?
Re:Follow me (Score:5, Interesting)
That is a clever solution!
Or the owning companies should be obliged to fit a remote control system and use that to drive them out of the way after 5 minutes of being unnecessarily stationary?
Until someone with basic cryptography skills and a poor morals decides to hack it and generate their own QR codes for optimum car theft.
Re:Follow me (Score:5, Insightful)
That is a clever solution!
Or the owning companies should be obliged to fit a remote control system and use that to drive them out of the way after 5 minutes of being unnecessarily stationary?
Until someone with basic cryptography skills and a poor morals decides to hack it and generate their own QR codes for optimum car theft.
Right. Stealing a car with cameras and GPS location sounds like a pretty smart way to use those skills.
Re: (Score:2, Flamebait)
Until someone with basic cryptography skills and a poor morals decides to hack it and generate their own QR codes for optimum car theft.
The system could phone home and all the authentication would happen somewhere else. Or to work in an outage, it could generate short-lived OTPs and distribute them to both devices and vehicles on a schedule.
needs to work with no network / limited network (Score:2, Offtopic)
needs to work with no network / limited network.
Re:Follow me (Score:4, Insightful)
for optimum car theft.
I doubt anyone would consider the incredibly slow walking pace theft of a highly connected car surrounded by security cameras and location sensors which constantly calls home "optimum".
That would be right up there with breaking and entering a police station.
Re:Follow me (Score:5, Insightful)
Does not have be a validated process.
That QR code for example could be ignored unless the vehicle is in already in failed/I don't know what to do other than stop for safety state.
The code could just identify a specific employee at public transport/police/fire/etc. Which could then look up their device in a dataset not encoded in the QR code but out of any attackers control to discover that officers mobile device; than a one time use set of secrets is sent to both the car and mobile device and they can establish trust between each other using some CHAP over bluetooth mechanism or similar.
There are ways to do this pretty safely with the existing hardware and communication lines out there if anyone wanted to get together and agree on how, and bother to implement it.
Re: (Score:2)
That is a clever solution!
Assuming the car is still looking at the camera and responding to it.
(which seems unlikely if the car's frozen to the spot)
Maybe what it needs is a watchdog timer combined with a honk detector. If the wheels haven't turned for 2 minutes and it detects at least three different honk sounds around it then it should reset the computer.
Re: (Score:2)
Re: (Score:2)
Or the owning companies should be obliged to fit a remote control system and use that to drive them out of the way after 5 minutes of being unnecessarily stationary?
In many cases here we're only talking about a couple of minutes. They already send someone out to move them on short notice. 5 minutes to block traffic is already a huge burden that typically is only allowed with special permits from councils.
Re: Follow me (Score:3)
Or they ensure there is an employee in the car at all times. They can be on their laptop doing something else for all I care. If the car companies think that is too risky, then maybe they arenâ(TM)t ready for public roads?
Re: (Score:2)
Or we can aggressively issue fines for obstructing a public right of way, and add a stipulation that the fine doubles each time. By the 20th occurrence, you can almost guarantee they will have gotten the message.
Re: (Score:2)
Stationary driverless cars are just exacerbating an existing flaw with any kind of rail vehicle.
While normal road vehicles can drive around an obstacle, a vehicle running on rails cannot - anything blocking the track stops the vehicle until it can be cleared.
Normal road users have to put up with delays all day every day - pedestrian crossings, traffic lights, loading/unloading, defective vehicles, damaged vehicles, dropped goods, damaged roads etc.
Re: Follow me (Score:5, Interesting)
100 light rail tram *filled with passengers*
As tempting as it sounds, I feel like perhaps the better solution might well be to just get out, and tip the bloody thing on its back like we used to do at work when idiots would park cars in front of our truck depot trapping millions of dollars in shipments. We'd just walk over , everyone grabbed one said of the car and pushed it over onto its back to get it out the way. Oh it'd destroy the car, but with the big sign that promised fines of up to $100K for obstructing our business (and trust me, we'd be losing more than that), fixing a dinged car roof was the least of their problems. "Sure you can ask us to pay for your damaged car, but we'll just be deducting the $5K repair bill from the $100K you'll be oweing us. Or you could f**** off and consider it a hard lesson learned"
Re: (Score:3)
Re: (Score:2)
"Sure you can ask us to pay for your damaged car, but we'll just be deducting the $5K repair bill from the $100K you'll be oweing us. Or you could f**** off and consider it a hard lesson learned"
Unless you're a government, if they were blocking you in but still on a public road, you'd be out of pocket for more than $105k after you lose your court case.
Private enterprises don't have rights recover damage from the inaccessibility of public infrastructure, and no one has the right to damage someone else's property. And yes this has been tested in court many times over by people who think like you.
Re: (Score:2)
Private enterprises don't have rights recover damage from the inaccessibility of public infrastructure,
I can understand that a private enterprise can't sue the government for highway construction that delays them, or for a toll booth delay. But the scenario presented is a private enterprise suing an individual for blocking their truck depot. I don't think a person can just park a car and block a commercial enterprise and say "nyah nyah, the law says you can't stop me, hahaha!"
Re: Follow me (Score:4, Informative)
Private enterprises don't have rights recover damage from the inaccessibility of public infrastructure,
I can understand that a private enterprise can't sue the government for highway construction that delays them, or for a toll booth delay. But the scenario presented is a private enterprise suing an individual for blocking their truck depot. I don't think a person can just park a car and block a commercial enterprise and say "nyah nyah, the law says you can't stop me, hahaha!"
No, but their recourse consists of calling police/parking enforcement (if on a public street) or calling a tow truck (if on private property). Under no circumstances is it legal for the business to "fine" the parker "$100K" or do thousands of dollars of damage through vandalism.
Re: (Score:3)
and then the renter of the auto taxi will get that (Score:2)
and then the renter of the auto taxi will get that bill + admin fees and you will not get paided as there CC will auto reject an 100K bill.
Re: (Score:2)
Where are you at that the businesses just get to make up their own fines? Kinda sounds like bullshit to me.
Re: (Score:3)
Re:Follow me (Score:4, Funny)
I don't think I want to stand in front of a malfunctioning driverless automobile.
It's better than standing in front of a normal vehicle with a malfunctioning driver.
Re: (Score:2)
I don't think I want to stand in front of a malfunctioning driverless automobile.
It's better than standing in front of a normal vehicle with a malfunctioning driver.
Is it? Both seem equally stupid. It's not like this is a binary choice where you have to pick one.
Re:Follow me (Score:4, Interesting)
Police, firemen, and other public servants should have an app on their phones that generates a one-time QR code that they can hold in front of the car's camera and the car will slowly follow the phone until it is out of the problem area.
I have a much simpler solution: robotaxi should be keyed-alike, like police cars and fire engines are. When the damn thing refuses to move, the copper open the stubborn vehicle with his cruiser's key, climbs into the car, disables Robostupid and parks the car the old-fashioned way.
Re: (Score:2)
Given the increasing availability of bluetooth phone keys, it would be easier to implement than ever.
Re:Follow me (Score:5, Insightful)
Re: Follow me (Score:4, Interesting)
The problem with this whole situation is that the car shouldn't have been in the bus lane to start with. Because busses are crippled as viable transportation unless they have their own lanes. If they have to travel in the same lanes as the car traffic, guess what that means: it means it's impossible for them to be any faster than driving. It also means it will be impossible for them to run on a timetable, which is critical for public transportation because you have to be able to rely on the timetable to get to work on time, to plan when to depart and return, etc.. So who would ever take the bus if you just sit in the same traffic as you would if you drove? And who would take the bus if you can't rely on it showing up on time to bring you back? Answer: nobody who isn't desperate, which is basically the case in North America. But this is by design. They don't want public transportation to succeed, only do it as a halfhearted or malicious attempt to "prove it doesn't work" so nobody uses it, so that they can flow the money to their cronies in the road construction business where they want it to. Don't want Americans to get any ideas in their heads about postponing their next car purchase. The thousands of busses belching diesel and sitting in car traffic in North America are proof. I don't think it's possible at this point to support the theory that they are actually this incompetent anymore. It's malice. And posts like this prove that it works.
Re: (Score:3)
Sounds like at least some of them were blocking intersections. I've never seen a bus route no matter how dedicated avoid having intersections with other vehicles. Which is normally fine, if the lights are well integrated, they can cycle through in a way to be amenable to bus traffic, or at least have a well predicted small impact on the schedule, but if a self-driving car gets confused and parks itself in such an intersection, not much can be done.
It's true that it would be nice if mass transit has as muc
Re: (Score:2)
I live in one of your glorious public transportation based cities. It's fucking shit. In America you can drive to costco, get economy of scale on household goods like meat and veg, and then transport it home, all in less time that I spend on one leg of the bus trip to buy just what meager amount of groceries I can carry myself. And that's even before you take into account that a car gives you a climate controlled way to transport your food, rather than needing to stand in the sun and then try to fit onto a
Re: Follow me (Score:5, Insightful)
Bus lanes - like buses themselves - have their effiency entirely determined by density.
If you have a bus lane on which buses travel frequently, yes, they're increasing your throughput per unit area. But if you have bus lanes that spend most of their time empty, it's much more area-efficient to just mix buses into regular traffic and allocate that area to general traffic, even if it means somewhat reducing per-bus throughput.
Mass transit requires that first word: "mass". A mass of people wanting to travel on a specific route. It requires the existence of density and key areas of travel focus. If you have this, mass transit becomes very efficient. If you lack this, it becomes very inefficient, and trying to shoehorn everyone onto mass transit without the existence of density and key departure / arrival areas just slows everyone down and costs more money.
Re: (Score:2)
Waymo can in some instances take remote control of the car. Obviously it relies on there being a cellular signal, or maybe satellite.
My understanding is that the interface isn't a wheel and pedals, instead they can see all the cameras and tell the vehicle that it is safe to proceed along one of the paths it has found for itself, but is not sure of.
I guess if it is damaged they will have to send someone to drive it in person.
but will need to have an common keyset like elevat (Score:2)
but will need to have an common keyset like elevators as you don't want to have an fireman have to deal with 20 differnt apps + and 50+ differnt keys if each state has it's own and that may be based on where an car is registered vs where it is at.
Re: (Score:2)
Police, firemen, and other public servants should have an app on their phones that generates a one-time QR code that they can hold in front of the car's camera and the car will slowly follow the phone until it is out of the problem area.
Better still, the QR code should a) allow access to the car even if its network connection isn't working and b) allow the vehicle to be put into manual override mode and driven to an appropriate location. Putting city employees at risk while they're leading the car through traffic on a virtual leash is unacceptable.
Re: Follow me (Score:2)
Re: (Score:2)
You are barking up the wrong tree. TFA iw about confused self driving cars.
Re: (Score:2)
get off my lawn!
Re: (Score:2)
Electric cars should be abandoned at the first sign of problems!!
You're obviously being hyperbolic, but we should note that electric cars have nothing to do with autonomous cars. There exist both non-autonomous electric cars and autonomous combustion cars.
Immature electric cars should, perhaps, not be allowed out on the streets where all of us get to bear the costs of assisting the development of these things by various private companies. Let the companies build a test track that doesn't screw up life for the rest of us (and occasionally kill someone).
Re: (Score:2)
A clean energy economy, including electric vehicles, involves LESS [fosstodon.org] mining [fosstodon.org], by a sizable margin.
EV fires are also rarer than fossil-vehicle fires. Which literally involve having a large tank full of liquid chosen specifically for how flammable it is, with no isolation or fire suppression system, plumbed up through a hot engine past numerous potential ignition sources, and then with hot exhaust plumbed out the back through a catalytic converter that gets so hot it can ignite dry grass.
Re: (Score:3)
This is false. There are no oxidizers inside battery cells. A battery fire is by and large "burning electrolyte", the electrolyte generally being organic carbonates (petrochemical solvents, not that different from gasoline). But unlike gasoline, it's inside individual steel cans surrounded by fire retardant.
The difference in extinguishing is that cells each have stored energy and self-heat as they discharge, so you're not going to just put it out by snuffing off the oxygen (in a gasoline fire, the heat c
How much delay do shitty drivers cause? (Score:4, Insightful)
We haven't fixed human-driven cars in a century and they are killing us by the thousands.
It is long past time we phased out human-driven cars.
Re: How much delay do shitty drivers cause? (Score:2, Flamebait)
Says some kid who probably can't drive and is just suffering from a bad dose of envy of those who can.
Re: (Score:2)
Nice, because... You've had yours, who cares about anyone else.
Re: (Score:2)
How childish.
Re: (Score:2)
Nope. Someone who, a long time ago, when he was a dumb kid like you, had a nice sportscar but now cycles around town and has always realised the absolute mind-blowing stupidity of filling our cities with cars and roads.
Cool. Wait until you get a bit older still, and cycling isn't as easy as it is now, I guarantee you'll have further realisations about urban planning.
Re: (Score:3)
FFS, are people really that stupid.
Re: (Score:2)
So what are you other choices if you live somewhere with no public transport? Walk?
Moron.
Re: (Score:2)
So you don't have answer then. Kind of what I thought. People like you always resort to abuse when your cretinous ideology is shown up for the BS it is.
Re: (Score:2)
Also Vioil8: "People like you always resort to abuse when your cretinous ideology is shown up for the BS it is."
Moron.
Re: (Score:2)
"a dumb kid like you"
Seriously? This "dumb kid" has been driving since the 80s and people like you have been around ever since I first went on the road.
"now cycles around town"
Lucky you. Now cycle 50 miles to a business meeting or do some family (which you probably don't have) grocery shopping and fit it all on a push bike or carry it on a bus.
Re: (Score:3)
Can you find a single story about a human driver causing this kind of delay? Or are you just ranting inanely again about apples not being oranges? Remember, driverless cars still aren't approved for remotely tricky situations -- they have this kind of sad failure mode even in easy situations.
Re: (Score:3)
There have been plenty of incidents of human drivers causing catastrophic failures due to ignoring signals at level crossings.
For that matter, there have been plenty of incidents of cars stalling on tracks due to other failures of car systems, and cars stalling due to the driver falling asleep at the wheel while waiting for a light.
Re: (Score:3)
[citation needed]
Come on, dude, this isn't the 1970s any more. Modern cars are more reliable than they used to be. Regardless, most of the failure modes you mentioned aren't related to human vs automated driving; and if a driver falls asleep, it doesn't take 17 minutes to wake them up like the first example in this story.
Re: (Score:2)
I can offer one from personal experience. There was a car stopped on a roundabout, holding up traffic. A couple of people went around it, but as I got near I could see that the driver was unconscious. I put my hazard lights on, stopped and called 999. Didn't hang around much after the ambulance and police turned up, but they closed the road can created big delays as it came up to rush hour.
Re: (Score:2)
More reliable if properly maintained and driven correctly, but you still can't always rely on people.
Re: (Score:3)
Modern drivers aren't more reliable. If anything they are far less so.
Re: (Score:2)
Oh, hey, since I didn't provide an example, how about the famous 1995 Fox River Grove collision https://en.wikipedia.org/wiki/... [wikipedia.org], in which a critical contributing factor was that the driver of the bus did not realize that the back of the bus was fully clear of the tracks.
Re: (Score:2)
The National Transportation Safety Board (NTSB) investigation found that, while the bus driver was not aware that a portion of the bus was on the tracks as she should have been, the timing of signals was so insufficient that, even if she had identified the hazard as the train approached, she would have had to proceed against a red traffic signal into the highway intersection to have moved out of the train's path.
That critical factor?
The changes made in response to that crash were technical. The most human-centered one was adding stickers to tell substitute drivers the specific length of the bus they are driving.
Re: (Score:3)
Can you find a single story about a human driver causing this kind of delay?
While not exactly the same, there is this [fox5dc.com] and this [cleveland19.com]. Or the people who deliberately block an intersection [mercurynews.com] so they don't have to wait for the next green light.
Of course there are the multitude of stories of vehicles blocking train tracks [cbsnews.com] and being plowed into [usnews.com], delaying traffic there.
Re: (Score:2)
Self-driving cars aren't allowed to drive in those locations, are they? (They're not allowed to because people believe they'll make similar errors, perhaps even more often than humans.)
Re: (Score:2)
But why do you think causes the normal, unrelenting jams filling our cities ?
Yep - delays caused by shitty drivers in their shitty cars.
Re: (Score:2)
No facts to back up your ranting. I am shocked! Shocked! Okay, not that shocked.
Re: (Score:2)
Can you find a single story about a human driver causing this kind of delay?
kinda [thenewstribune.com]
driverless cars still aren't approved for remotely tricky situations -- they have this kind of sad failure mode even in easy situations.
I still don't get how they are failing to be able to remotely pilot them out of these situations. They're having basic communication failures? That should be even more embarrassing than the self-driving failures.
Re: (Score:2)
Are you familiar with the concept of Mean Time Between Failures? Hardware dies a lot quicker than people.
Re: (Score:2)
Can you find a single story about a human driver causing this kind of delay?
You're joking right? Are you asking if I can find a single story about an arsehole carelessly blocking a street and causing a traffic jam?
To help me answer your question are you limiting the time frame to just this week? It's only Monday but yeah I can already already share a story even in that timeframe. If you want stories from a larger time period you may need to request aggregated statistics, because humans are absolute selfish fuckwits with how they use the road.
Re: (Score:2)
No -- I want a story that is actually comparable to the cases here, where a human driver blocks a light-rail train or reserved-lane bus for 5 or 15 minutes because they have totally stopped responding to outside inputs.
The idiot I responded to pretends that autonomous vehicles are better at driving than humans, while ignoring that they are very limited as to what and when and where they are allowed to drive. The result is that he counts a lot of scenarios as showing that humans drive poorly where autonomou
Re: (Score:2)
I'm doubtful of that thesis as are you, but at least, mercifully, in this case we are talking about an appropriate failure state of autonomous cars, where they are just slowing down traffic instead of killing people. I guess GM/cruise is just a better system than Tesla's FSD.
Re: (Score:2)
I am not skeptical of the argument that we can more easily improve automated drivers than we can human drivers. I'm just fed up with bullshit lines like the OP's "It is long past time we phased out human-driven cars." -- especially when those arguments are based on such stupid evidence.
It's not clear that automated driving systems are really better than humans in the cases where they are approved for use already. It is pretty clear they're worse for a lot of more challenging situations, and I am not super
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
>We can fix the driverless cars.
Nope, there are more edge cases than dreamt of in the fail safe expert system in these cars. It's an AI hard problem and enslaving an AGI to drive a car has other problems.
The best solution is to allow full RC. Not just an instructor providing remote guidance, but allowing a remote driver to take complete control through VR.
Re: (Score:2)
For the most part, software gets better and more reliable. I can't remember the last time my PC had a Blue Screen in Windows, or a Kernel Panic in Linux. Or even having to compare your file against a Checksum, or run a disk check on your Computer?
New stuff fails, and we can either just go it is crap, and make new stuff that fails, or actually see if we can fix the problems. Yes sometimes a full redo is the correct option, especially if you had coded yourself in a corner where the fixes just make things w
Solution (Score:5, Insightful)
Declare the stalled, empty vehicles abandoned and/or illegally parked and have them towed for a fine.
I imagine Cruise/Waymo will correct, or prevent, their issues promptly.
Re: (Score:3)
How do you do that safely, if the car can at any moment suddenly decide to start moving? Do they have some button under a cover somewhere that first responders and towing agencies can use to immobilize it?
Re: (Score:2)
How do you do that safely, if the car can at any moment suddenly decide to start moving? Do they have some button under a cover somewhere that first responders and towing agencies can use to immobilize it?
With a great big grabber that snatches it off the ground in a tenth of a second.
A few scratches on the paintwork might get them to fix their cars.
Re:Solution (Score:4, Interesting)
Yep, and it doesn't even have to be a great big grabber, a light duty wrecker with a stinger could do exactly that just like they do for at-risk repo. A skilled operator can literally handle that in seconds, although it gets more complicated if it's an AWD/4WD vehicle.
Re: (Score:3)
Re: (Score:3)
It is already a traffic violation and should trigger a fine.
Yes, a really big juicy fine payable by the manufacturer. It's the only way they learn.
Re: (Score:2)
Yes, a really big juicy fine payable by the manufacturer. It's the only way they learn.
The city goes after the owner/operator of record. If they can recover costs from the mfgr then that's great. If it happens enough the city may also seek damages from the mfgr for their irresponsibility.
Re: (Score:2)
Yes, a really big juicy fine payable by the manufacturer. It's the only way they learn.
The city goes after the owner/operator of record. If they can recover costs from the mfgr then that's great. If it happens enough the city may also seek damages from the mfgr for their irresponsibility.
Making the manufacturer pay seems pretty fair to me. It's not the owner of the self-driving vehicle that causes an accident, it's not the owner of the vehicle who's driving, it's the manufacturers AI. Therefore it's not him/her who caused the accident or traffic interruption, it was the manufacturer. If a company sells a car as 'self driving' it had better damn well be 'self driving' in the most literal sense of that term and any accident is therefore due to the incompetence of that company and their faulty
Re: (Score:3)
I imagine they will correct their issues promptly regardless of what you do. They are already suffering monetary and reputational damage, just adding extra complication of having to pick up the cars as well won't make a difference.
You would however be wasting time for a tow-truck to address the issue of what averages to ... let's see... 27.6 seconds worth of delay somewhere within the entire bus network every day. Does that sound like a good use of someone's time?
83 minutes of delays over 6 months (Score:2, Insightful)
Re: (Score:3)
I mean seriously? Is that your reason for going apeshift insane? I guess for haters any pretext is good enough.
People have been shot for less than that in freedom-land.
Re: (Score:3)
I mean seriously? Is that your reason for going apeshift insane? I guess for haters any pretext is good enough.
People have been shot for less than that in freedom-land.
Yea, in some parts of the country it wouldn't be a delay, it'd be target practice time...
Re: (Score:2)
Re: (Score:2)
Geeeze what is the comparable statistic counting nothing but busses broke down in the road over the same period? It happens. We cope.
Re: (Score:2)
BTW, how's your tinfoil hat fitting?
Tightly!
annoying, but... (Score:2)
[After a white Waymo SUV stopped in the middle of the road.... roadside assistance crews arrived within 11 minutes of being dispatched to drive....
Why not remotely drive the car to a safe haven? They should make sure these cars can also be remotely driven. It shouldn't be a big problem with all the camera's in place, have the remotedriving be done using a VR headset so the remote driver has the same view as a normal driver. Also I wonder if they really have a decent tracking station which would alert immediately if a car is stopped.
But we're still in the early days
Re: (Score:2)
Why not remotely drive the car to a safe haven?
That might be illegal, or the car might have broken down such that remote access didn't work.
Or, quoted from the article:
Instead, the Waymo Driver, as the company calls its technology, alerted a remote “fleet response specialist” to help. Waymo spokesperson Sandy Karp says that this worker provided guidance to the car that “was not ideal under the circumstances” and made it challenging to resume driving.
forced windows updates strike again? (Score:5, Funny)
"Software updates are being installed, do not shut down your car. Estimated time to completion: 39 minutes."
will they need an fire Phase I / II like override? (Score:2)
will they need an fire Phase I / II like override? so that any emergency vehicles can do an quick and easy override of one?
two words (Score:2)
Two more words (Score:2)
Qualified Immunity.
Generally speaking, if the police have a vehicle towed, for any reason, they are not liable for any damages the owner encounters, and the owner is frequently liable for towing and storage charges.
People have tried to fight this, and rarely win. Even in the case where their vehicle was stolen, and later recovered by the police, the owner is generally charged towing and storage fees. The most likely outcome in an driverless car case will be that the company will just pay the fees as
Nobody wants driverless cars (Score:2)
Solution (Score:3)
Issue the "driver" a citation. Accumulate enough if these and your autonomous operators license is suspended*, pending completion of a remedial driving class.
*That means all of Waymo, or whoever operates the fleet.
Charge 'em (Score:2)
$10,000 per minute sounds about right. I bet there wouldn't be many offenders after that.
lesson (Score:2)
Main lesson: don't rely on some bureaucracy to get you from point A to point B.
Secondary lesson: SF has a new form of govt, whose main goal is to persecute and torture its citizens while taking ever larger gobs of their money.
Fine (Score:2)
The *owners and operators* of frozen driverless cars should be fined. Maybe $1000 for ever five minutes that they're blocking traffic, and $2000 if they're blocking public transit.
Inaction (Score:2)
"I may not do harm to human beings, or, through inaction, allow them to come to harm."
"I must obey all human commands, unless it conflicts with the first law."
"Get out of the way!"
"Hmmmm. An impatient bus driver sideswiped a robot car once. Therefore best not to move if a bus, possibly filled with humans, is honking its horn impatiently."