Pedestrian fatally crushed between two parked cars after one started remotely
gothamist.com
gothamist.com
I thought it was a Tesla that somehow completely failed to see a pedestrian and then to stop when it made contact, which would have terrible implications for the rest of the fleet.
Not really, this means there is a whole bunch of cars like this already in the wild and people being unaware of this.
I would guess this car has had some kind of aftermarket fiddling...
They pushed the car away from the kid, did they just let it roll back up on him? I doubt it, it's more likely the vehicle was in gear.
I don't mean to downplay the seriousness of this bug, I'm just saying it doesn't have as catastrophic implications as it would if this happened in a state-of-the-art Tesla.
There are approximately 250e6 registered vehicles in the US, so if all the Tesla-related fatalities are those "3 per year" (which seems unlikely to me; it assumes that there are no Teslas involved in fatal accidents when the automatic driving stuff is not engaged) to be disproportionate there would have to be only ~21000 Teslas on the roads. There are close to an order of magnitude more than that at this point.
This site seems to be an interesting aggregation of Tesla deaths: https://www.tesladeaths.com/ It claims 29 deaths in the US from Teslas in 2019; 3 are from accidents where auto pilot use was claimed.
No one would do that if they could though. No manufacturer would share it if they were asked either since it could “leak IP” and so the reality is that won’t happen and the manufacturer should probably be responsible.
After all, the operator deployed the autonomous vehicle and should be responsible for what it does. If they aren't familiar with its capabilities and risks, they should not deploy such a vehicle unless they are willing to bear the responsibility for whatever it might do. Until such a time autonomous vehicles have free will and can be held accountable for their actions, the operator who deployed it assumes that accountability.
This isn't even a new concept. If I let my dog loose in the park and it bites someone, I'm responsible for it. We don't hold dogs responsible for their actions, we hold the owner who "deployed" it responsible.
I'll cede that things are a little murkier when manufacturers are pushing OTA updates to vehicles that change their behavior. But there's no reason the operator and the manufacturer can't share responsibility at a split decided in court.
You forget that we live in a society where we need people to use cars. Nothing force you to have a dog, nothing force you to let him loose either in a park. I hope one day it will no longer be a necessity and public transit will replace that need, but sadly, in most cities, it's pretty far from being able to.
It seems like people don't take their responsibilities as drivers very seriously while pedestrians, motorcyclists, and cyclists are being killed and maimed as a result. All without much consequence to the drivers who create such an unsafe condition. Then when you ask motorists to drive responsibly they rage out at you trying to shift blame to someone else.
I don't see how you could interpret my comment as believing that was the subject.
I'm not arguing about people not taking responsibility either. I'm arguing that one day, we will no longer even have enough control over a car to be able to be responsible for it.
In the current state of "self-driving car" (which is closer to a fancy cruise control) I believe you are entirely responsible for it.
> You forget that we live in a society where we need people to use cars.
Even if this is completely true, it does not absolve the driver of any responsibility, so I don't really see the point of bringing it up.
> One day we won't be able to drive at all, it will be all self driving, and at that point, you'll no longer be able to do anything to avoid an accident.
When this day comes, who the operator is may change, but there is still an operator somewhere deploying and managing these vehicles to hold accountable. If the taxi you're riding in is involved in a collision, you're not responsible.
It does if your only choice is using it or not.
Being responsible is about making choice to reduce risk. If you are required to use a car to sustains the current society and you have no control over the risk of the car, how can you be the one responsible for what it does?
> When this day comes, who the operator is may change, but there is still an operator somewhere deploying and managing these vehicles to hold accountable.
My point is that this "operator" will be the manufacturer.
> If the taxi you're riding in is involved in a collision, you're not responsible.
Thanks for the perfect example to support my case. At one point owning a car will be just like going into a taxi. You wouldn't be responsible for an accident caused by the taxi driver, why would you be if it was a virtual one? Sure if you still had a wheel, but that's certainly not the future we are going toward.
However if self driving cars aren't individually owned and are run by a transit agency, then the responsibility shifts to the agency who turned on the car. Just like when a train or buss hits a car.
We also don't need to use cars even in cities with poor transit. Most people's car trips are biking distances. Everyone can also own a moped and enjoy 100mpg, easy parking, and experience less congestion than what the 20 foot car introduces to a city. Driving a car is a choice.
An emergency brake isn't the end of all sadly.
> laws require you to be present and focused to drive on the road
Sure, I'm not arguing the current state of self driving car (which I wouldn't really call self driving car, but more like fancy cruise control, on the Volt they called it adaptive cruise control, which make more sense). I'm arguing about the future without a wheel to drive. It's not happening right now, that's for sure, but Google is working on it and there's no
> You still get a dui on a horse even if you are passed out on top and the horse is walking itself home.
You still can act on a horse though, what made you unable to control it is your consumption (which is your responsibility not to). In a self driving car without a wheel, you no longer have any control, even if you wanted to. It may have an emergency brake, but that's can only do so much. In a world where going on a horse would cause you to be inebriated, you wouldn't get a DUI for riding one ;).
> We also don't need to use cars even in cities with poor transit. Most people's car trips are biking distances. Everyone can also own a moped and enjoy 100mpg, easy parking, and experience less congestion than what the 20 foot car introduces to a city.
Where do you live? Have you lived outside of there? In most cities you still need a car. Sorry for the accusation but theses sentences made it seems like you never lived outside of there. I'm close to Montreal in Quebec, even my car yesterday had trouble in some streets, believe me, even though we are lucky to have some amazing bike lane here, biking distance get thrown out of the window for the next 4 months. Walking is much slower right now, a moped would be actually illegal. I traded 2 hours of transit a day to 30 minutes by buying a car. I know so many people that currently do more than 2 hours a day, and that's in a city with a relatively good public transit in a pretty dense area.
Inside most cities, you still need a car. Without cars, you couldn't support the current society. You are free to disagree that our current way of living is alright, but sadly, we can't change that easily and it will certainly not happen before self driving car become the norm.
I live in LA but grew up on the east coast. I'm not sure about Canada, but 50 percent of trips in the U.S. are 4 miles or less (1). 4 miles at a comfortable biking pace translates to 20 minutes, which is pretty doable on a bike (not to mention good cardio). On a bike you don't really experience traffic either, in fact it's almost better to bike when cars aren't moving during rush hour.
Even if people just biked while the weather was nice, it would make a huge difference for congestion and the environment, and would drive public support for better bike infrastructure to make it even easier. It's damn cheap to paint a bike lane, compared to other transit plans after all.
1. https://www.fhwa.dot.gov/policyinformation/pubs/pl08021/fig4...
I don’t want personal responsibility. I want to not get killed by an autonomous vehicle, period, and I say hold the party responsible whose responsibility more likely prevents my death. For me, that leans toward the manufacturer. If that bankrupts manufacturers because their devices kill people, fantastic, that just opens the market wide for autonomous vehicles that don’t kill people.
Don't we though? At least in the past, if a dog bit someone, they were sometimes euthanized.
This isn't about AI, it's a simple dumb case "guy presses button that makes metal move. Metal crushes pedestrian". There should be no doubt about who's wrong, and it's not a moral question.
Naturally if the starting was a failure of the remote start system it's the manufacturer, but still not really a moral discussion, just plain old product failure.
However, it’s functionality is greatly reduced compared to the US version, and requires the operator to be standing nearby with a line of sight to the vehicle.
for me it just shows how far off autonomous driving really is, I swear a roomba could navigate a parking lot better and quicker. currently the driving features are best confined to limited access roads; interstates
https://www.reddit.com/r/teslamotors/comments/dtx14k/tesla_s...
(Per parent post, this is not applicable in Europe)
(I'm very much against wiping the insides of the windows as the smearing can make driving at night very dazzling)
What if you open the driver's side window at first?
1. Emissions, catalytic converters and such will not reach full operating temp, so it dumps a lot more emissions than when it's up to operating temp. Diesels were even worse in this regard cause the particulate filter plugs up when idling and then needs to trigger fuel wasting regen cycles.
2. Engine wear/damage. Oil pressures usually don't come up enough to properly lubricate components on the engine while cold idling for long periods of time. (Like, after the rpm's settle back down) This adds unnecessary wear. Should really only run an engine long enough to bring up initial oil pressure and then take it easy in it until it's up to full operating temp. (This is because parts expand at different rates, so pushing a cold engine hard will oftentimes blow seals)
For some reason that doesn’t sound so crazy for Germany though. I’m not sure why.
I didn't say illegal, I said bad.
I haven't seen this in practice. In fact, I see the opposite. My 2003 350z has an oil pressure gauge, and when it's idling while cold (say, < 40 degrees F), the pressure is actually higher than it is when cruising on the highway once it's warmed up.
I attribute this to the oil being "thicker" when cold.
On a really cold start, the pressure goes well beyond 60 PSI until the RPMs taper back down from around 1500 to the normal cold idle of ~1000, at which time it drops a bit.
The other major issue that specifically effects cold idling is that the ECM is going to increase the amount of fuel dumped into each cylinder, past what gets completely burned. This leads to problem called 'quench' that leaves gasoline behind on the walls of the cylinder after combustion. This will mix with the oil on these areas and actually dissolve it over time.
Anyways, don't take it just from me. No manufacturer specifies over 30 seconds of idle time for their vehicles: https://info.ornl.gov/sites/publications/files/Pub61263.pdf
Performance will still be worse and wear higher than a warm engine. But there’s no reason you can’t let it warm up on the way. Growing up in Canada it was always body rust not the engine that ultimately did the cars in anyways.
They also advance the timing when cold which increases emissions. I forgot to mention that as well.
> And multi viscosity oils ensure lubrication even when cold.
Improved, still not ideal.
See "Driving in winter" [1]
[1] https://finland.fi/life-society/how-to-survive-winter-in-fin...
I don't even see people letting their cars idle while scraping, which was the norm 15 years ago. I think the info has spread that it is bad for your engine and the environment. And newer cars don't need it at all.
For the windows to fog up, you need to sit in the car right? How long do you need to sit in your car until they fog up? Why are you sitting in your non-moving car for so long? And how often does that happen?
In our cars it takes about 10 minutes in cold winter before the fog comes up when around 2 people sit in it, unless it is raining and we just jumped in wet a minute ago. And it's gone a minute after starting the AC again. And one is a small car 2 week short of being 20 years old.
From this viewpoint I cannot possibly understand your reasoning? Are you trying to find an excuse to disable Start/Stop?
False. In autumn, when temperatures are low and winter comes (minuses) the humidity is really high. I can't tell you exactly what must happen, but I'v had fogged windows in the morning.
Throughout the years living here, I have learned that following every law by the book, is more a stereotype foreigners have about Germany than how things actually work.
The argumentation in the manual is that the car won't reach operational temperatures for a longer time while idling. New efficient motors apparently idle very efficiently (surprise?), which generates not enough heat.
I work in that business.
Source please?
I live in Europe and first time I hear such a thing. Yeah, I'v got my drivers license a while ago already, but no-one has mentioned not even whispered that such a law exists.
...and with even more force, since the car now had a bit more distance to build up some speed. One wonders whether he would've survived if they had left him there, and it was the second blow that killed him, since the fact that they were able to push it back suggests it was probably an automatic left in neutral with nothing more than the idle torque. If it was a manual, stopping it or pushing it backwards would've already stalled the engine. I suppose another lesson here is that if you ever have to push a car out of the way, chock the wheels afterwards, or otherwise try to prevent it from moving again.
I'm also surprised that was enough to kill.
Edit: downvoters, explain yourselves please.
Then the person pushing the car back gets tired (since they are pushing against an automatic cars 'creep' torque converter), and the car comes back at you again, this time crushing your head.
I can see how that's deadly.
Depending on the height of the bumpers involved, I would suspect several possible problems: one or both femoral arteries severed; or both femurs fractured; or the pelvis fractured; or some combination of these.
Even if the arteries were untouched, fractures of both femurs or the pelvis cause life threatening blood loss as these are heavy bones which have major vascularisation (to allow blood cells generated in the marrow to enter the blood stream). A closed femur fracture is a 1000 mL to 1500 mL loss of blood (per femur). An open femur fracture is double that. A pelvic fracture can go to 5000 mL (approximately the total volume of blood in the average adult male).
In the field, major lower limb arterial bleeding can be controlled with tourniquets; femur fractures can be somewhat reduced with traction splints; pelvic fractures can be somewhat stablised with pelvic compression devices... However all of these scenarios are bad news for the patient and require rapid evacuation and surgical intervention (ideally in less than 1 hour from the time of injury) to have a chance of saving the person's life. RIP and condolences to patient's family.
Looks like it would pin you just above the knees if you were her height.
(Edit: typos.)
You're accusing good samaritans of killing someone - with no evidence to indicate that this was the case?
A manual could start in neutral without the clutch, but if it somehow slipped into gear (don't see how that's possible without someone forcing the stick), it would stall immediately.
(Also, yes, the 2002 IS300 was available in stick.)
The handbrake can start slipping when the brakes cool and the car could roll, so both are necessary, especially on a hill.
Maybe it is not current anymore, but I have always been instilled a certain distrust of handbrakes. :)
It's a non-issue for the gears. You can turn a 4-cylinder crank with your bare hands. There's very little load, you're just overcoming the static friction of bearings, piston rings, and compressing valve springs. At the static to creeping rate we're talking, the cylinder fluids leak right through the piston ring gaps.
If you want to worry about something with regard to statically loading everything, it's a lack of lubrication that should worry you.
The engine isn't maintaining oil pressure when stopped. So you're leaving the crank bearings loaded in one spot while oil slowly squeezes out. "Frictionless" oil film bearings are made of relatively soft metals, so this isn't exactly good for the engine.
Also the gearbox isn't dispersing oil as it does when driving. So the gears are slowly drying up while loaded. I'd be more concerned about the engine bearings though.
I never leave my manual vehicles loaded in gear when off. It's pointlessly hard on the bearings. Starting is already harsh enough in that brief moment before oil pressure builds. Why add insult to injury?
There are oil pressure accumulators on the market in part for maintaining oil pressure when off for friendlier starting of high compression race engines:
An off engine is not unable to turn. Its resistance to turning is not constant either. All piston rings leak, even when new. So while you may have observed the engine in first gear holding the vehicle, minutes later the leaking piston rings may have allowed the crank to turn, and vehicle to roll.
The primary holding mechanism should always be the brake, and you need to load it to verify it's functioning and that you've pulled hard enough on the parking brake.
Putting it in gear is just a weak and largely ineffective form of redundancy.
Chocking turned wheels against the curb is much better redundancy (and legally required when parking on hills around here).
this happened in my 2005 mazda 3 and the habit carried over to my 2017 Golf (not sure if the brake design is similar or not, maybe it's not necessary on the vw)
So was it in gear or was it in neutral?
Handbrake should be used to keep it effective (so that the cable moves regularly and does not get stuck). And, if the gear is released accidentally, you still have the brake, and vice versa.
(Myself, I have always parked with handbrake on + gear in 1st or reverse; nowadays with my VW it's pretty automatic because it's DSG and you have to put the selector to P to take the key out, and electric parking brake activates automatically.)
> (unless on grade) to avoid the handbrake becoming less
> effective over time as the cable stretches.
Your justification for not using the handbrake for its intended purpose it that it may become less effective in the future? Seems counter intuitive unless I'm missing something...
If you believe your handbrake cable will stretch/break, just check it regularly. It really isn't that expensive to replace (compared to other car maintenance) and there's no harm in being extra cautious in this case.
The only emergency it is useful for is when you really have to pull a 180 right now and tight, as the automated systems do not expect this situation and will relax the brakes while you're deliberately trying to spin. I wouldn't trust even an above average drive with pulling that off - it's not being practiced. Usually it would be a desperation move when there's no space and a head on collision is likely.
Also depending on weather and temperature the handbrake can fail.
I’ve never seen anyone successfully start an engine from stand-still with it in gear (though I’ve seen people try). You get it rolling in neutral and only then engage the gear to start the engine without a starter or battery.
Your engine will lose compression over time so it's plausible that it'll roll a bit without the brake engaged.
And parking brakes by themselves are some of the most fail prone systems in the cars.
If you've ever changed tires on a fwd car and kept it in gear, you will notice you can torque the front tires far harder than the rear, as the front will be held by the engine and the rear wheels can spin surprisingly easily by your own power even when held by the hand brake.
We didn't know it until recently, which is how we ended up with blocked rear wheels after the car stayed parked for a couple of days, one of which was rainy. The process of unlocking the wheels damaged the brakes, so we also ended up with two visits to repair shop (one to hot-fix them before a trip home, another to fix them properly).
We drove the car straight to the mechanic after that.
Just be careful, and make sure your electrical cord is in good condition, and you should be fine. Bonus points in that you get an excuse to nab the Mrs./Ms. hair dryer for a good laugh when someone inevitably sees you giving your car a good blow down.
I'd be incredibly leery of trying to force anything through engine output/mechanical vibration alone. Always try undoing the environmental complication first. It's much less prone to causing costly repairs, and easier to do than it seems. A car cover that goes to the ground, and a beefy enough space heater underneath the cover in the morning might do the trick too. Although winds can make the heat gun/hair dryer + point heating work better.
Brake fluid doesn't freeze, but any water it manages to have pulled out of the air will. In fact if you have too much water in your brake lines from moisture absorption from the atmosphere because you were fiddling with the master cylinder, or using fluid that had a chance to absorb significant moisture, you could get large enough concentrations of water to get some crystallization started. Normally though, hydraulic brake systems in good condition shouldn't have a problem with that. In the unlikely event you do, the amount of ice shouldn't threaten the physical integrity of the system, and warming the brake lines would resolve it. Note that having enough water for that to happen gas probably made your braking sponge though, and you should really flush your brakes at that point when you get a good dry day.
They were just stuck on too hard. At that age I didn’t know any better, and my dad who was the one encouraging me to gun it was never that mechanically minded.
Over 10 years later I now do all maintenance and servicing on my cars myself, so yeah if it happened again I would not have been doing that!
Not sure how this applies to electric hand brakes, but I would hope they have a sensor to detect if the motor has engaged the brake or if it's just stalled out.
Learned the hard way when I saw my car in the middle of the street one day. Left it in gear every time after that.
https://c8.alamy.com/comp/B79ANK/san-francisco-california-st...
If I'm worried about the car rolling, I don't particularly want it turning the engine backwards.
In terms of ratios I don't give it much thought. On pretty much every car I've driven first gear is shorter than reverse (typically first is > 4:1 vs reverse at < 4:1).
> "I've never heard of this happening before," a police spokesperson told Gothamist.
> Multiple inquiries to Lexus were not returned.
I hope a software bug that missed checking for neutral in an avoidable edge-case isn't what's responsible for the young man's life. Lexus should respond to add clarity as to what really happened.
Edit: completely forgot about really cold winters...
If this isn't the epitome of a first world problem... Imagine the man power, resources and money spend so that we can avoid being cold for 3 minutes.
Sometimes it's even challenging to open the doors, because they are all frozen.
I get that not everyone gets the white snow christmas, but some do :)
No more iced or snow-covered windows.
> An owner's manual cautions: “NEVER remotely start the engine if you are unaware of the circumstances surrounding the vehicle as it may cause a life-threatening situation for those located in the vehicle’s vicinity.”
I predict that if self driving cars do get widely deployed, the number of 'hit and runs' they do because they couldn't detect they had collided will be very high, even if the overall accident rate is low.
Whatever gives you the idea no one cares?