Dash cams are becoming increasingly appealing. Obviously if someone pulls in front of you that is a different case than hitting someone you are already following behind.
Dash cams certainly would help make it very clear what actually happened.
If you did have someone jump in front of you, find the centre of the lane and then brake suddenly, a dash cam will show there was not enough time to re-establish a safe following distance.
Now there is a question.... How long is a reasonable time to drop back to a safe distance when someone cuts in front? Too short and you would have to jump on the brakes which could cause an accident. Too long and you're leaving yourself open. Either way, the traffic behind you won't like your caution if you make it too obvious.
If they still had loads of false object detections, enabling it may have caused loads of rear end collisions as well.
This. Not only do pedestrians need to be predictable, but EVERYTHING needs to be predictable to eliminate collisions.
Utter nonsense. You have to be prepared for the driver in front of you braking for apparently no good reason at all times. Because of that in most countries you are at fault when you drive your car into the rear of the car in front of you: you can't see what the driver in front of you is seeing.
I agree you need to be prepared for the driver ahead of you to brake. But the thing being missed is that is actually quite likely and therefore predictable and more importantly you don't expect them to reach zero speed instantly. The car ahead didn't drop a brick wall that you have to avoid hitting. A "safe following distance" is such that you can react to the predictable situations such as the car ahead braking. They have a stopping distance similar as you, so what you are reacting to is their action of braking. This is predictably something a car does and in this case if you have "a safe following distance" you both brake and you both decelerate together and don't collide. "A safe following distance" is different if the car ahead of you is a Lamborghini and you are in a Suburban, but these are things you can work out on the fly and don't violate the rules. A semi-truck won't magically have the stopping distance of the Lamborghini. .
A story. Once, I was driving in the third lane out of a 4 lane carriageway - just before it splits into 2+2 for a junction. Ahead of me was a White Van. I was following it leaving a safe distance, when...
...the van engine seized. Almighty cloud of blue smoke from the back end of the van. A mixture of diesel, rubber, and bits of engine (but not lubricating oil. Clearly no lubricating oil). The thing just stopped in the middle of the road with traffic passing either side.
I didn't hit it - I had left plenty of room to stop in an emergency. I always do now. Always. Even if it annoys the heck out of the Perfect Drivers.
Three true stories...
My first car was, um, interesting. If you coasted up to a roundabout and then put your foot down to accelerate into the gap in traffic - it would drop out of gear leaving you stranded. Oh the joys of a car where the automatic gearbox transmission oil is shared with the dirty engine oil
My second car was also interesting, but in a nicer way. Except when pulling back in after overtaking on the A11 in Thetford forest - the nearside wheel fell off and buried itself in the hedge (fortunately!). I got the RAC out, and we decided there was no damage to the car, so I simply took one nut off each of the other wheels and bolted the wheel back on.
Or, when travelling 2s behind another car at 50mph, a female deer decided to run across in that gap.
Annnd, another proof that autopilots just can't autopilot: if the human fails, the autopilot fails.
Looks like either the police and Uber have decided to scapegoat the driver, or the police and Uber tried to justify the collision in the first instant and now have to back pedal due to the NTSB investigation's findings.
A police report released Thursday would seem to imply that the driver was looking at her mobile phone and only looked up 0.5s before the crash. Wouldn't be surprised if the driver faces charges of vehicle manslaughter. It's illegal in the UK to drive whilst using a mobile phone and it's probably illegal in a lot of other countries. Probably illegal in the state of Arizona as well, but people still do it.
https://eu.azcentral.com/story/news/local/tempe-breaking/2018/06/21/uber-self-driving-car-crash-tempe-police-elaine-herzberg/724344002/
Well colour me surprised. Who on earth would have ever conceived something like that would happen? Certainly not a software engineer, since they infamously respond to any problem with a car by turning it off and expecting it to have disappeared when they turn it back on.
There are two "safe" and one "dangerous" designs.
Safe: no automation, driver knows they are in control
Safe: full automation, driver can "switch off" and relax
Dangerous: partial automation with shared responsibility and quick responses required. If the driver can't relax, why have the automation in the first place.
Military maxim: what is everybodys' responsibility is nobodyd' responsibility.
A police report released Thursday would seem to imply that the driver was looking at her mobile phone and only looked up 0.5s before the crash. Wouldn't be surprised if the driver faces charges of vehicle manslaughter. It's illegal in the UK to drive whilst using a mobile phone and it's probably illegal in a lot of other countries. Probably illegal in the state of Arizona as well, but people still do it.
https://eu.azcentral.com/story/news/local/tempe-breaking/2018/06/21/uber-self-driving-car-crash-tempe-police-elaine-herzberg/724344002/
For those particularly interested, here's the police report:
https://www.documentcloud.org/documents/4551043-18-32694-Redacted-Report.htmlIt even mentions that if the driver had paid attention, she would've been able to stop 42.61 feet before the accident. It also says her eyes were off the road for 6 minutes and 47.2 seconds total of 21 minutes and 48 second drive, or about 3.67 of 11.8 total miles.
I do think, though, that it's a bit silly to put humans in a task like this, though. I'm not sure how you expect someone to just watch the road go by for hours upon hours day after day. Not that I'm trying to downplay the driver's responsibility, but I think we still need to focus more on Uber's responsibility, especially getting rid of the second person in the vehicle and disabling emergency breaking without even giving any sort of notification.
It even mentions that if the driver had paid attention, she would've been able to stop 42.61 feet before the accident.
How did they arrive at that exact number? Amazing precision.
By checking the car interior video, which supposedly was better quality than the dodgy one from the dash camera they released.
A police report released Thursday would seem to imply that the driver was looking at her mobile phone and only looked up 0.5s before the crash. Wouldn't be surprised if the driver faces charges of vehicle manslaughter. It's illegal in the UK to drive whilst using a mobile phone and it's probably illegal in a lot of other countries. Probably illegal in the state of Arizona as well, but people still do it.
https://eu.azcentral.com/story/news/local/tempe-breaking/2018/06/21/uber-self-driving-car-crash-tempe-police-elaine-herzberg/724344002/
She wasn't driving a conventional car under conventional licence conditions. I wonder how well they tied down legal responsibilities and liabilities when they issued the special licences to operate partially tested autonomous cars on the Pheonix roads?
Not very clever on her part, then, I'd guess she knew she was being recorded, didn't she?
It's illegal in the UK to drive whilst using a mobile phone
Only if the phone is handheld. Although any driver can be prosecuted for an instance of bad driving which is ostensibly not simple human error but due to engaging in some kind of distraction at the wheel.
https://www.gov.uk/using-mobile-phones-when-driving-the-law
It's illegal in the UK to drive whilst using a mobile phone
Only if the phone is handheld. Although any driver can be prosecuted for an instance of bad driving which is ostensibly not simple human error but due to engaging in some kind of distraction at the wheel.
https://www.gov.uk/using-mobile-phones-when-driving-the-law
The same is true in the US, despite numerous studies showing that it's the conversation that is distracting, not the act of holding onto the phone.
Personally I don't understand why it's so hard to just put the phone away while driving. It wasn't that long ago that the phone was attached to the wall at home and nobody expected everyone to be immediately reachable at all times.
Personally I don't understand why it's so hard to just put the phone away while driving. It wasn't that long ago that the phone was attached to the wall at home and nobody expected everyone to be immediately reachable at all times.
Times change - and so do expectations.
The daughter of a friend has her world so tied up with instant communication that it comes back to bite her all too often. Whenever others want to communicate with her she can't get "time off". If she doesn't respond within the
expected time frame, she gets harangued for ignoring them.
I hate to say it, but she made that rod for her own back.
I don't run that way. Not in the slightest.
The same is true in the US, despite numerous studies showing that it's the conversation that is distracting, not the act of holding onto the phone.
You need to consider how things
fail, not just how they
work.
It is difficult to safely change gear with a phone in one hand.
It is difficult to take quick avoiding action with only one hand on the wheel.
Driving conditions in the US are very different to those in Europe.
The same is true in the US, despite numerous studies showing that it's the conversation that is distracting, not the act of holding onto the phone.
You need to consider how things fail, not just how they work.
It is difficult to safely change gear with a phone in one hand.
Americans don't change gears.
(and (b), nor do Europeans in emergency situations)
It is difficult to take quick avoiding action with only one hand on the wheel.
It doesn't stop them from stomping on the double-wide brake pedal though.
Driving conditions in the US are very different to those in Europe.
So are the cars.
I just noticed that Uber sold off its autonomous driving business to Aurora at the end of last year -
https://www.theguardian.com/technology/2020/dec/08/uber-self-driving-car-aurora They seem to have got $4B for it, so Aurora seems to think Uber's work has real value. Most people are backing away from their crazy claims that autonomous driving is nearly here. I hope this doesn't suck all the financing out of this sector. With more realistic goals and sane time scales I do think this in an important area of R&D.
The same is true in the US, despite numerous studies showing that it's the conversation that is distracting, not the act of holding onto the phone.
You need to consider how things fail, not just how they work.
It is difficult to safely change gear with a phone in one hand.
It is difficult to take quick avoiding action with only one hand on the wheel.
Driving conditions in the US are very different to those in Europe.
There was a pretty bad crash near our house - in a 20mph zone, a driver managed to plow his BMW into a telephone pole (car was a writeoff). The sheepish looking driver said to me "the snow on the side of the road caught the wheel and steered me into the pole" - utter BS, I thought, looking at the tracks in the snow from the wheels... they led in a straight line into the pole!
Pretty sure the (male) driver was not looking where he was going just that little bit too long...
I just noticed that Uber sold off its autonomous driving business to Aurora at the end of last year - https://www.theguardian.com/technology/2020/dec/08/uber-self-driving-car-aurora They seem to have got $4B for it, so Aurora seems to think Uber's work has real value. Most people are backing away from their crazy claims that autonomous driving is nearly here. I hope this doesn't suck all the financing out of this sector. With more realistic goals and sane time scales I do think this in an important area of R&D.
Weird because I thought the whole long-game point of Uber was to get to the goal of full autonomous driving.
Weird because I thought the whole long-game point of Uber was to get to the goal of full autonomous driving.
They don't need to develop their own system if all the major car manufacturers are already doing it for them.
(plus it's less liability if somebody else does it)