First self driving (reported) fatality

[glasgowcyclist]

Here's some video of the street where the collision occurred, taken at night and showing just how well lit the area is. Remember that the human eye will have an even clearer view than the camera's sensor.
(Rather worryingly, it appears to be handheld device by driver!)
RTC location is about 33 seconds in.


View: https://www.youtube.com/watch?v=CRW0q8i3u6E&feature=youtu.be

 

[mjr]

They're already calling for cyclists to be forced to wear a beacon so they can be detected, and pedestrians will be next.
That idea alone makes it clear to me that AVs are nowhere near ready to be trialled on public roads.

(Oh, and what do you do when you're miles from home and your beacon goes flat?)

You have to get off and walk. AVs can cope with people walking.... oh, wait...

I wonder whether the beacons will have a light on them so you know it's working... and how much the light will shorten the beacon's battery life... and whether there will be cheap fake beacons on ebay that only have the light... it's not really a workable idea, is it? Why is it not being laughed out of existence?

 

[mjr]

The stupid decisions are when pedestrians suddenly cross a road without looking or making any signs of body language which suggests that they are about to cross the road.
Do or can self driving cars read body language ?

Some do, because one got confused by a trackstand.

The person killed didn't exactly "suddenly cross a road without looking or making any signs of body language" - they were most of the way across the road before being struck and you can't have much clearer body language than having almost crossed.

This looks like a catastrophic fundamental system failure and all vehicles using that system must be kept off road until it's debugged, plus I feel the test vehicles should have something like the Driver Safety Device and Driver Vigilance Device used on trains to increase the probability that the human driver is alive and attentive.

 

[mjr]

[QUOTE 5190767, member: 45"]I was physically threatened in an IKEA car park a few weeks ago because I was making my way to our car and in a driver's way. This attitude stinks and it's performed on the public highway day in, day out.[/QUOTE]
Was that near you? Motorists seem particularly aggressive there. I saw someone walking deliberately (albeit gently, thankfully) run into in Worle Sainsbury's car park seemingly because they didn't get out of the driver's way after being indicated at and honked at. Several of us in the car park told the driver what we thought of them but the victim didn't want to call the police... was it that she didn't want to waste their time, given how Avon & Somerset Police back then didn't seem interested in road rage, or maybe it's part of widespread public acceptance that if you are walking not on a footway then you're fair game for motorists

 

[humboldt]

I really loathe it when someone lays on the horn if they perceive you as walking slightly too slowly and holding them up, especially if they're actually going to be going nowhere anyway. I was midway across Holloway Road the other day when the green man started flashing and a couple of plumbers started revving their Transit and honking at me. I made sure to stop dead and give them the finger for a solid five seconds just so they could actually have something to legitimately be annoyed about. I'm nice like that sometimes.

 

[Illaveago]

[QUOTE 5190767, member: 45"]Dramatics aside, I think that driverless vehicles could bring a positive if it takes us back to the place where pedestrians can rightly walk in the road without fear of being killed.

I was physically threatened in an IKEA car park a few weeks ago because I was making my way to our car and in a driver's way. This attitude stinks and it's performed on the public highway day in, day out.[/QUOTE]
I am just being realistic. These vehicles seem to be designed where Jaywalking is illegal and the driverless car has right of way over anyone that happens to be in it's way. As in a previous post above which includes an American assessment of Uber incidents so far, it implies that the vehicles do not recognise the difference between lanes of stationary vehicles and an open road and will speed along regardless of any vehicle which might cross it's path. It could be the reason why the car failed to react going by what was written in that post.

 

[mjr]

[QUOTE 5190903, member: 9609"]It should make no difference whether jaywalking rules apply or not, the car should not crash into any obstacle that it encounters, I'm sure it would be illegal to leave a 20 ton block of concrete in the middle of the road, whats the car going to do with that, drive into it 40mph then say it shouldn't have been there.[/QUOTE]
Does this mean we can get rid of self-driving cars simply by installing car traps on their roads and expecting humans to drive around one side of them? After all, if they can't see a bicycle in front of them, what chance a car trap:

http://cycle.st/p17296

 

[jarlrmai]

I am just being realistic. These vehicles seem to be designed where Jaywalking is illegal and the driverless car has right of way over anyone that happens to be in it's way. As in a previous post above which includes an American assessment of Uber incidents so far, it implies that the vehicles do not recognise the difference between lanes of stationary vehicles and an open road and will speed along regardless of any vehicle which might cross it's path. It could be the reason why the car failed to react going by what was written in that post.

So this is either the terrifying reality of the jaywalking law written in to a robot terminator style or the robot car was just not able to detect an obstruction and stop the car which would appear to be the most basic expectation of a robot car.

 

[SkipdiverJohn]

Obviously yes in this instance they were clearly distracted by a phone as well which is even worse but it demonstrates they either need to be really actually self driving or not on the roads at all.

The problem is that humans quickly become bored by things they are not interested in. When I'm driving a vehicle I have no choice but to pay attention in order to be able to control it. Therefore I don't get bored. When I'm a passenger in a vehicle that someone else is driving in a smooth manner (and not throwing me around), my attention soon wanders from what is going on ahead to all sorts of other things. In reality, you cannot expect a person sitting in a car who is not actually driving it, to maintain a high level of alertness to road hazards. I am NOT a believer in autonomous vehicles in any shape or form being allowed on to public roads where the environment is always going to be unpredictable and often chaotic. The roads in places like Arizona aren't even a difficult test of this technology. If you want to try to prove autonomous vehicles are ACTUALLY safe then run the trials in the middle of a major city in a third-world country teeming with suicidal truck, bus & car drivers, motorcyclists, pedal cyclists, animals, and pedestrians and see how many times the autonomous car has an accident in that scenario.

 

[jefmcg]

I think that is a demonstration of the big problem with them. None of the cars makes any adjustments for the humans, because the people are moving with the same precision as the cars. If a pedestrian stopped to look at their phone, or jumped forwards or back because they didn't trust the car not to hit them, then there would be a huge pile up.

The vehicles are all moving like a giant clock work engine. Put a normally behaving human in there, and the effect would be like dropping a mouse into the workings of big ben**

**yes, I know Ben is the bell, not the clock.

 

[mjr]

[QUOTE 5191040, member: 45"]I think it's more complex than that, and that there are levels of consciousness and alert. I doubt there are many drivers who are conscious that they're changing gear. Speed and engine noise monitoring, and processing of which gear is needed are all subconscious, as is the resultant gear change. We don't consciously steer either. Or indicate.[/QUOTE]
I can accept that the listening or glancing at the rev counter needle/bar and "SHIFT v^" indicator (if you have it) may be routine almost to the point of subconscious but I'm surprised that gear changing (in a manual car, anyway) is subconscious for anyone.

Similarly steering - I often find myself thinking about steering and how much I want to turn the wheel which way when to move the car around in a certain way.

Maybe that sort of thing is why I don't really like driving Cycling seems much more intuitive, especially now half my bikes only have one gear lever.

 

[mjr]

[QUOTE 5191068, member: 45"]Maybe we're both assuming that everyone else is the same as we are. Our car doesn't have a rev counter and I don't look at the shift indicator, and I'm never aware that I'm changing gear, nor steering.[/QUOTE]
Blimey. Next you'll be saying it doesn't have a head-up display or any way to fit one. Is it a vintage car?

 

[glasgowcyclist]

In the past I've occasionally had a car journey where, on arrival at my destination and thinking back to the trip, I've little or no recollection of events along the way. Almost like I've been in a trance or on autopilot, driving without absorbing my surroundings.

That hasn't happened for a very long time, possibly because it worried me and the fear of it happening again has made me more attentive.

One thing I have noticed is that this has never happened to me at the end of a cycling journey. (In fact I can replay the full trip in my head.)
Is this because when cycling we are so much more engaged with what's around us, since we are so vulnerable?

 

[jefmcg]

Almost like I've been in a trance or on autopilot, driving without absorbing my surroundings.

I know that feeling. I think you probably were paying attention, but because your amygdala is not excited, so nothing got written to short term memory. If something - god forbid - exciting had happened on your journey, you'd have found yourself well aware of what had immediately preceded it. But as nothing did happen, 15 seconds later your brain says "I am not going to replace the last memory of our 12 birthday with 1/4 of a mile of the M8"

 

[glasgowcyclist]

because your amygdala is not excited

Does that mean my amygdala is excited when I'm cycling? Not that I know what an amygdala is ... <goes off to google>